CN115182747B - Automatic tunnel crack repairing method, device and system and readable storage medium - Google Patents
Automatic tunnel crack repairing method, device and system and readable storage medium Download PDFInfo
- Publication number
- CN115182747B CN115182747B CN202211106955.6A CN202211106955A CN115182747B CN 115182747 B CN115182747 B CN 115182747B CN 202211106955 A CN202211106955 A CN 202211106955A CN 115182747 B CN115182747 B CN 115182747B
- Authority
- CN
- China
- Prior art keywords
- tunnel
- crack
- data
- repaired
- map
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 148
- 238000003860 storage Methods 0.000 title claims abstract description 22
- 230000008569 process Effects 0.000 claims abstract description 52
- 230000015556 catabolic process Effects 0.000 claims abstract description 16
- 238000006731 degradation reaction Methods 0.000 claims abstract description 16
- 230000010412 perfusion Effects 0.000 claims abstract description 11
- 238000001179 sorption measurement Methods 0.000 claims abstract description 11
- 230000009194 climbing Effects 0.000 claims description 42
- 239000002245 particle Substances 0.000 claims description 38
- 239000003822 epoxy resin Substances 0.000 claims description 35
- 229920000647 polyepoxide Polymers 0.000 claims description 35
- 230000008439 repair process Effects 0.000 claims description 33
- 230000006870 function Effects 0.000 claims description 27
- 230000033001 locomotion Effects 0.000 claims description 27
- 239000000463 material Substances 0.000 claims description 27
- 238000012545 processing Methods 0.000 claims description 23
- 238000005259 measurement Methods 0.000 claims description 21
- 239000002131 composite material Substances 0.000 claims description 18
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N Silicium dioxide Chemical compound O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 claims description 17
- 238000006243 chemical reaction Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 12
- 238000010276 construction Methods 0.000 claims description 10
- 238000001914 filtration Methods 0.000 claims description 7
- 239000000835 fiber Substances 0.000 claims description 6
- 230000005484 gravity Effects 0.000 claims description 6
- 239000003094 microcapsule Substances 0.000 claims description 6
- 238000002156 mixing Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 239000004372 Polyvinyl alcohol Substances 0.000 claims description 5
- 230000001133 acceleration Effects 0.000 claims description 5
- 229920002451 polyvinyl alcohol Polymers 0.000 claims description 5
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 4
- 229910021389 graphene Inorganic materials 0.000 claims description 4
- FPYJFEHAWHCUMM-UHFFFAOYSA-N maleic anhydride Chemical compound O=C1OC(=O)C=C1 FPYJFEHAWHCUMM-UHFFFAOYSA-N 0.000 claims description 4
- 239000005543 nano-size silicon particle Substances 0.000 claims description 3
- 235000012239 silicon dioxide Nutrition 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 6
- 238000004422 calculation algorithm Methods 0.000 description 20
- 239000004567 concrete Substances 0.000 description 20
- 238000010586 diagram Methods 0.000 description 13
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 13
- 230000000694 effects Effects 0.000 description 12
- 239000013598 vector Substances 0.000 description 11
- 239000011159 matrix material Substances 0.000 description 10
- 239000004568 cement Substances 0.000 description 7
- 238000013507 mapping Methods 0.000 description 7
- 239000000243 solution Substances 0.000 description 7
- 238000013461 design Methods 0.000 description 6
- 239000002002 slurry Substances 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 230000006378 damage Effects 0.000 description 5
- 238000001514 detection method Methods 0.000 description 5
- 230000000149 penetrating effect Effects 0.000 description 5
- 238000005070 sampling Methods 0.000 description 5
- 239000000377 silicon dioxide Substances 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 4
- 238000013527 convolutional neural network Methods 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 4
- 229920005989 resin Polymers 0.000 description 4
- 239000011347 resin Substances 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 238000013459 approach Methods 0.000 description 3
- 239000002585 base Substances 0.000 description 3
- 238000005187 foaming Methods 0.000 description 3
- 230000004927 fusion Effects 0.000 description 3
- 238000002347 injection Methods 0.000 description 3
- 239000007924 injection Substances 0.000 description 3
- 208000037805 labour Diseases 0.000 description 3
- 230000004807 localization Effects 0.000 description 3
- 239000002184 metal Substances 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 239000004814 polyurethane Substances 0.000 description 3
- 229920002635 polyurethane Polymers 0.000 description 3
- 239000013643 reference control Substances 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000001680 brushing effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000005336 cracking Methods 0.000 description 2
- 238000006073 displacement reaction Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 239000002105 nanoparticle Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011176 pooling Methods 0.000 description 2
- 239000011148 porous material Substances 0.000 description 2
- 239000000047 product Substances 0.000 description 2
- 238000007789 sealing Methods 0.000 description 2
- 239000010959 steel Substances 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 238000009755 vacuum infusion Methods 0.000 description 2
- VEXZGXHMUGYJMC-UHFFFAOYSA-M Chloride anion Chemical compound [Cl-] VEXZGXHMUGYJMC-UHFFFAOYSA-M 0.000 description 1
- 239000004593 Epoxy Substances 0.000 description 1
- 238000000342 Monte Carlo simulation Methods 0.000 description 1
- 239000011398 Portland cement Substances 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 239000002253 acid Substances 0.000 description 1
- 239000013543 active substance Substances 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 239000003513 alkali Substances 0.000 description 1
- XLJMAIOERFSOGZ-UHFFFAOYSA-N anhydrous cyanic acid Natural products OC#N XLJMAIOERFSOGZ-UHFFFAOYSA-N 0.000 description 1
- 150000001540 azides Chemical class 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 229910002092 carbon dioxide Inorganic materials 0.000 description 1
- 239000001569 carbon dioxide Substances 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000007596 consolidation process Methods 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 230000008602 contraction Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 239000013078 crystal Substances 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000005553 drilling Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000010006 flight Effects 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 230000009746 freeze damage Effects 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 239000011440 grout Substances 0.000 description 1
- 230000012010 growth Effects 0.000 description 1
- 239000013003 healing agent Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000010220 ion permeability Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- -1 maleic anhydride modified graphene Chemical class 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 239000003960 organic solvent Substances 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 239000003973 paint Substances 0.000 description 1
- 238000010422 painting Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 239000012466 permeate Substances 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 229920001084 poly(chloroprene) Polymers 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 230000001681 protective effect Effects 0.000 description 1
- 238000004080 punching Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 230000003014 reinforcing effect Effects 0.000 description 1
- 239000012779 reinforcing material Substances 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
- 238000007569 slipcasting Methods 0.000 description 1
- ORFSSYGWXNGVFB-UHFFFAOYSA-N sodium 4-amino-6-[[4-[4-[(8-amino-1-hydroxy-5,7-disulfonaphthalen-2-yl)diazenyl]-3-methoxyphenyl]-2-methoxyphenyl]diazenyl]-5-hydroxynaphthalene-1,3-disulfonic acid Chemical compound COC1=C(C=CC(=C1)C2=CC(=C(C=C2)N=NC3=C(C4=C(C=C3)C(=CC(=C4N)S(=O)(=O)O)S(=O)(=O)O)O)OC)N=NC5=C(C6=C(C=C5)C(=CC(=C6N)S(=O)(=O)O)S(=O)(=O)O)O.[Na+] ORFSSYGWXNGVFB-UHFFFAOYSA-N 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000009736 wetting Methods 0.000 description 1
Images
Classifications
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH OR ROCK DRILLING; MINING
- E21D—SHAFTS; TUNNELS; GALLERIES; LARGE UNDERGROUND CHAMBERS
- E21D11/00—Lining tunnels, galleries or other underground cavities, e.g. large underground chambers; Linings therefor; Making such linings in situ, e.g. by assembling
- E21D11/003—Linings or provisions thereon, specially adapted for traffic tunnels, e.g. with built-in cleaning devices
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH OR ROCK DRILLING; MINING
- E21F—SAFETY DEVICES, TRANSPORT, FILLING-UP, RESCUE, VENTILATION, OR DRAINING IN OR OF MINES OR TUNNELS
- E21F17/00—Methods or devices for use in mines or tunnels, not covered elsewhere
- E21F17/18—Special adaptations of signalling or alarm devices
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/50—Depth or shape recovery
- G06T7/521—Depth or shape recovery from laser ranging, e.g. using interferometry; from the projection of structured light
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Engineering & Computer Science (AREA)
- Mining & Mineral Resources (AREA)
- Physics & Mathematics (AREA)
- Geochemistry & Mineralogy (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geology (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Optics & Photonics (AREA)
- Architecture (AREA)
- Civil Engineering (AREA)
- Structural Engineering (AREA)
- Lining And Supports For Tunnels (AREA)
Abstract
The application discloses a method, a device and a system for automatically repairing a tunnel crack and a readable storage medium, which are applied to engineering safety technology. The method comprises the steps that an unmanned aerial vehicle carrying a plurality of sensors for collecting image data and positioning data is used for collecting multi-mode data based on an autonomous planning path in the flight process of a tunnel to be detected without a light source or a GPS (global positioning system) so as to construct a three-dimensional tunnel map; in the three-dimensional tunnel map building process, degradation characteristics of the surface of the tunnel and the interior of the tunnel are detected at the same time, and a three-dimensional tunnel cloud map for identifying position information of an area to be repaired is generated based on the three-dimensional tunnel map and the degradation characteristics; the method comprises the steps of positioning each crack to be repaired of a tunnel to be detected by using a wall-climbing robot based on a three-dimensional tunnel cloud chart, and moving the crack to be repaired to a target to perform negative pressure adsorption and perfusion, so that the crack of the tunnel can be safely, efficiently and effectively repaired, and the stability and the safety of a tunnel structure are effectively improved.
Description
Technical Field
The present application relates to the field of engineering safety technologies, and in particular, to a method, an apparatus, a system, and a readable storage medium for automatically repairing a tunnel crack.
Background
The lining cracks such as cracks, water freeze damage, structural deformation and the like are main diseases threatening the safety of the tunnel under the comprehensive influence of geological environment, adjacent construction, self structural load and the like. The tunnel cracks threaten the stability of the tunnel structure, and the serious deformation caused by the cracks can reduce the bearing capacity of the lining structure on surrounding rocks, so that the lining falls off or even collapses, and major accidents are caused. In order to ensure the stability of the tunnel structure, it is necessary to repair the tunnel crack. The cracks are generally divided into three types, namely type I cracks with the crack width smaller than 0.05mm, type II cracks with the crack width between 0.05mm and 0.2mm, and type III cracks with the crack width larger than 0.2 mm. Type I fractures are generally not repaired, type II fractures are generally surface closed, and type III fractures are generally chemically grouted.
The related art carries out crack repair through manual work and even requires high-altitude work for grouting and punching, such as high-pressure epoxy resinPouring, pinhole method high-pressure grouting, slurry brushing repair and the like. The high-pressure epoxy resin pouring mode is to directly inject epoxy resin materials into the cracks through a high-pressure injection gun (the pressure value is generally 0-60 MPa). The epoxy resin slurry can be cured at normal temperature, and the cured product has high compressive strength and tensile strength, high bonding force, small shrinkage and high acid and alkali corrosion resistance. However, in the case of narrow cracks, the grout is difficult to pour. And the epoxy resin has large brittleness, cannot resist the expansion caused by heat and contraction caused by cold due to stable environment, and is very easy to cause crack deformation. In addition, the slurry has high viscosity, poor injectability and poor bonding capability with wet cracks, and the high content of organic solvent in the slurry has certain damage to the environment and the health of constructors. The process of the high-pressure grouting by the pinhole method comprises the following steps: determining a leakage point → cleaning a leakage base surface → drilling → cleaning a hole → grouting → embedding a needle head → sealing a seam → grouting under high pressure by slurry → dismantling a grouting nozzle → repairing a pinhole → painting surface waterproof paint. Compared with the method, the method has better effect by adopting the high-pressure grouting machine, and the grouting material is neoprene, epoxy resin or polyurethane plugging material, so that the hydrophilicity is good, carbon dioxide gas is generated, and a high-strength elastic consolidation body can be generated. However, the surface of the structure must be damaged, which complicates the construction. The injection pressure is not high enough and the slurry cannot enter the deep part of the crack. When the polyurethane material is adopted for repairing, the plug is expanded at a water leakage position through chemical reaction with water, the principle is that isocyanic acid radical in the material is freely foamed when meeting water, the strength of the foaming body is low, the bonding strength with concrete is low, the foaming body is easy to shrink, partial water leakage can be generated when the foaming body is soaked in water for a long time, and therefore water corrodes steel bars, and the steel bars expand to further cause concrete cracking. The brushing pulp repairing mode is as follows: after the traditional cement paste is combined with concrete, the traditional cement paste can leak into the concrete, crystals insoluble in water are formed in the concrete, capillary channels are blocked, the incompletely hydrated components are crystallized, and cracks formed in the later period are closed. The repairing agent consists of Portland cement, silica sand and a plurality of active substances. By utilizing the chemical property and porosity of the concrete and the permeation action of water, the concrete micro-pores and capillary tubes are permeated and filled to catalyze concrete cement particles and unhydrated components, so that the concreteThe cement rehydrates to form water insoluble crystalline particles. However, the concrete interface is weak in connection, and the old concrete interface forms a water film, so that the local water-cement ratio is too high, and the interface is positionedAFtAnd Ca (OH) 2 The number is increased, and the preferred orientation of the page can greatly reduce the strength of the page; the surface of the concrete matrix is uneven, the coarse aggregate is formed on the surface of old concrete, cement paste cannot permeate into the concrete, and cavities are formed on the interface of the new concrete and the old concrete due to the lack of the cement paste.
To sum up, although manual work can reduce the probability that the tunnel takes place danger, but the manual injection method, the length of man-hour, construction quality is difficult to guarantee, and efficiency is not high to the manual work need go on discerning earlier the slip casting in the crack, expend a large amount of labours, and need scramble the scaffold frame, the safety problem can't be ensured. In addition, the crack repairing material adopted by the method, such as the traditional epoxy resin or polyurethane material, has high brittleness, low toughness and easy cracking, and the like, so that the crack repairing effect is poor, and the stability of the tunnel structure cannot be guaranteed.
Therefore, how to ensure the stability of the tunnel structure through a safe, efficient and effective crack repairing mode is a technical problem to be solved by technical personnel in the field.
Disclosure of Invention
The application provides a method, a device and a system for automatically repairing a tunnel crack and a readable storage medium, which can safely, efficiently and effectively repair the tunnel crack and effectively improve the stability and the safety of a tunnel structure.
In order to solve the above technical problem, the embodiments of the present invention provide the following technical solutions:
an embodiment of the present invention provides an automatic repairing method for a tunnel crack, including:
collecting multi-mode data to construct a three-dimensional tunnel map by using an unmanned aerial vehicle carrying a plurality of sensors for collecting image data and positioning data and based on an autonomous planning path in the flight process of a light-source-free GPS-free tunnel to be measured;
in the three-dimensional tunnel map building process, degradation features on the surface of a tunnel and in the tunnel are detected simultaneously, and a three-dimensional tunnel cloud map for identifying position information of an area to be repaired is generated based on the three-dimensional tunnel map and the degradation features;
and positioning each crack to be repaired of the tunnel to be repaired by using a wall climbing robot based on the three-dimensional tunnel cloud picture, and moving the tunnel to the target crack to be repaired for negative pressure adsorption and perfusion.
Optionally, the moving to the target crack to be repaired for negative pressure adsorption perfusion includes:
mixing nano silicon dioxide and epoxy resin in advance according to a first preset proportion, or mixing maleic anhydride, a graphene sheet material and epoxy resin according to a second preset proportion to generate composite epoxy resin; adding polyvinyl alcohol fiber or double-wall microcapsule material into the composite epoxy resin to modify the composite epoxy resin;
in the advancing process of each preset repairing path of the crack to be repaired according to the path, when the crack to be repaired moves to the current crack to be repaired, the wall-climbing robot pours the modified composite epoxy resin into the current crack to be repaired under the negative pressure environment.
Optionally, the unmanned aerial vehicle that utilizes the sensor of carrying on the multiclass and gathering image data and location data gathers multimode data at the flight in-process that does not have the tunnel that awaits measuring of light source no GPS based on independently planning the route to construct three-dimensional tunnel map, include:
according to unscented Kalman filtering, generating a three-dimensional tunnel map and determining the position of the unmanned aerial vehicle in real time according to data processing results of laser scanning data of a laser radar, ranging data of an ultrasonic range finder and measuring data of an inertia measuring unit and image data acquired by infrared image acquisition equipment;
and based on the three-dimensional tunnel map and the measured tunnel wall depth value of the tunnel to be measured, automatically determining the flight path at the next moment by combining an unmanned aerial vehicle motion model according to the mode that the unmanned aerial vehicle moves towards the estimated three-dimensional point on the tunnel crankshaft and moves in the tunnel to be measured step by step.
Optionally, according to unscented kalman filter, to laser scanning data of laser radar, the range data of ultrasonic distance meter and the data processing result of the measured data of inertial measurement unit, combine the image data that infrared image acquisition equipment gathered, generate three-dimensional tunnel map and confirm unmanned aerial vehicle's position in real time, include:
constructing a three-dimensional tunnel map according to the aligned laser scanning data;
determining the flight speed of the unmanned aerial vehicle according to the laser scanning data of the unmanned aerial vehicle at different moments;
determining absolute position information of the unmanned aerial vehicle in a flight area based on each laser scanning data and the latest available map state;
and fusing the flight speed of the unmanned aerial vehicle, the gravity adjustment acceleration of the inertial measurement unit and the absolute position information.
Optionally, before the constructing the three-dimensional tunnel map according to the aligned laser scanning data, the method further includes:
establishing a corresponding relation between all points of laser scanning data at different moments;
and determining the conversion information of the laser scanning data aligned to the current point of the laser scanning data at the current moment through a minimized error function, and applying the conversion information to each point of the aligned laser scanning data until the alignment condition of the laser scanning points is met.
Optionally, the positioning, by using the wall-climbing robot, each crack to be repaired of the tunnel to be repaired based on the three-dimensional tunnel cloud map includes:
the method comprises the steps of representing by using particles in advance based on position uncertainty of the wall-climbing robot, simultaneously expanding uncertainty of each position by using RRT, representing uncertainty propagation in a tunnel environment constructed by various types of terrains by using the particles, and determining a path planning rule of the wall-climbing robot;
generating a wall-climbing path of the wall-climbing robot based on the three-dimensional tunnel cloud picture and the path planning rule;
and calling a crack identification model to identify whether the crack to be repaired exists at the current position or not in the process that the wall climbing robot travels according to the wall climbing path.
Optionally, before the invoking of the crack recognition model identifies whether the crack to be repaired exists at the current position, the method further includes:
training a crack recognition model in advance based on a mode that information of each loss function is reflected to a network respectively to update a weight;
the crack identification model comprises an input layer, a convolutional layer, a plurality of dense blocks and an output layer; the output feature map of each dense block is input to the next dense block and the output layer simultaneously.
Another aspect of the embodiments of the present invention provides an automatic repairing apparatus for a tunnel crack, including:
the data acquisition module is used for acquiring multi-mode data based on an autonomous planning path in the flight process of the light source-free GPS-free tunnel to be detected by utilizing an unmanned aerial vehicle carrying multiple types of sensors for acquiring image data and positioning data so as to construct a three-dimensional tunnel map;
the initial positioning module is used for simultaneously detecting degradation characteristics of the surface of the tunnel and the interior of the tunnel in the construction process of the three-dimensional tunnel map and generating a three-dimensional tunnel cloud map for identifying the position information of the area to be repaired based on the three-dimensional tunnel map and the degradation characteristics;
and the grouting repair module is used for positioning each crack to be repaired of the tunnel to be repaired by using the wall-climbing robot based on the three-dimensional tunnel cloud chart, and moving the crack to be repaired to a target crack to be repaired for negative pressure adsorption and perfusion.
The embodiment of the invention also provides an automatic tunnel crack repairing system, which comprises an unmanned aerial vehicle carrying a plurality of sensors for acquiring image data and positioning data, a wall-climbing robot for deploying a vacuum perfusion instrument and electronic equipment, wherein the wall-climbing robot comprises a plurality of sensors for acquiring image data and positioning data;
the electronic device comprises a processor and a memory, wherein the processor is used for implementing the steps of the automatic tunnel crack repairing method when executing the computer program stored in the memory.
Finally, an embodiment of the present invention provides a readable storage medium, where a computer program is stored on the readable storage medium, and when the computer program is executed by a processor, the computer program implements the steps of the automatic tunnel crack repairing method according to any of the foregoing methods.
The technical scheme that this application provided's advantage lies in, unmanned aerial vehicle is based on the cloud picture of the approximate region that the sensor collection of carrying was gathered to the data generation sign tunnel needs to restore the crack, and wall climbing robot utilizes this cloud picture to confirm roughly the route, and rethread self sensor detects the discernment crack, moves to crack initial position, carries out detail detection and restoration. The whole crack repairing process is convenient and quick, a large number of maintenance personnel are not needed for operation, labor cost is saved, and the risk of harm to the maintenance personnel caused by unfavorable environment can be avoided. Wall climbing robot fills the restoration based on the vacuum, with air escape, forms the negative pressure cavity, fills the effect better, can go deep into the crack, and need not punch the operation, further promotes crack repair efficiency to can be safe, high-efficient, effectively repair the tunnel crack, and then promote tunnel structure's stability and security.
In addition, the embodiment of the invention also provides a corresponding implementation device, a corresponding system and a corresponding readable storage medium for the automatic tunnel crack repairing method, so that the method has higher practicability, and the device, the system and the readable storage medium have corresponding advantages.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the related arts, the drawings used in the description of the embodiments or the related arts will be briefly described below, it is obvious that the drawings in the description below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of an automatic tunnel crack repairing method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an exemplary three-dimensional tunnel model provided by an embodiment of the present invention;
FIG. 3 is a schematic view of an exemplary vacuum grouting provided by an embodiment of the present invention;
FIG. 4 is a schematic diagram of a point cloud construction process according to an embodiment of the present invention;
fig. 5 is a schematic view of an exemplary drone navigation provided by an embodiment of the present invention;
FIG. 6 is a schematic diagram of an exemplary wall-climbing robot according to an embodiment of the present invention, traveling over various types of terrain;
FIG. 7 is a schematic diagram of an exemplary fracture identification model according to an embodiment of the present invention;
FIG. 8 is a schematic diagram illustrating an exemplary training of a fracture identification model according to an embodiment of the present invention;
fig. 9 is a schematic diagram of a trajectory error of an exemplary wall-climbing robot in a path tracking process according to an embodiment of the present invention;
fig. 10 is a structural diagram of an embodiment of an automatic tunnel crack repairing apparatus according to an embodiment of the present invention;
fig. 11 is a structural diagram of an embodiment of an automatic tunnel crack repairing system according to an embodiment of the present invention;
fig. 12 is a block diagram of an embodiment of an electronic device according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the disclosure, the invention will be described in further detail with reference to the accompanying drawings and specific embodiments. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The terms "first," "second," "third," "fourth," and the like in the description and claims of this application and in the above-described drawings are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "comprising" and "having," as well as any variations of the two, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements but may include other steps or elements not expressly listed.
Having described the technical solutions of the embodiments of the present invention, various non-limiting embodiments of the present application are described in detail below.
Referring to fig. 1, fig. 1 is a schematic flow chart of an automatic repairing method for a tunnel crack according to an embodiment of the present invention, where the embodiment of the present invention may include the following:
s101: the method comprises the steps that an unmanned aerial vehicle carrying a plurality of sensors for collecting image data and positioning data is utilized, and multi-mode data are collected based on an autonomous planning path in the flight process of a non-light source GPS-free tunnel to be detected, so that a three-dimensional tunnel map is constructed.
The unmanned aerial vehicle of this embodiment carries on multiclass sensor, these sensors are used for gathering image data and are used for gathering relevant locating data, utilize these sensors to acquire and understand environmental information data, can include laser radar like SLAM laser radar, ground penetrating radar or infrared camera, distance measuring instrument like ultrasonic ranging appearance or laser range finder, inertial measurement unit IMU, light stream sensor, flight time distance sensor, of course, unmanned aerial vehicle still can carry other types's sensor, this all does not influence the realization of this application. The multi-modal data is all the raw sensor data acquired by the airborne sensor, or data obtained by processing the raw sensor data, and a three-dimensional tunnel map is generated by cooperatively processing the multi-modal data such as a camera, a laser radar, an ultrasonic range finder, and an IMU, as shown in fig. 2. The sensor data collected by each hardware facility carried by the unmanned aerial vehicle is modal data, various modal data are subjected to fusion processing, various different information interactions exist among different modules, the redundancy among the modalities is eliminated by utilizing the complementarity among the multiple modalities, different modalities are embedded into the same feature space to learn common expression, the feature difference is considered, and the commonality and the personality among the different modalities are found through a cross-modality shared feature transfer algorithm, so that better feature representation is learned, and more accurate tunnel structure features are obtained.
The tunnel environment that this application is suitable for is the dark environment that does not have GPS, and unmanned aerial vehicle realizes data acquisition at tunnel flight in-process, and this embodiment accessible unmanned aerial vehicle's airborne sensor relies on to the measured light processing of sensor to guide unmanned aerial vehicle to travel along the tunnel axis like SLAM laser radar and ground penetrating radar (or infrared camera), realizes realizing independently no collision navigation in unknown tunnel environment. Optionally, an Unscented Kalman Filter (UKF) and a particle filter are combined to process IMU (Inertial Measurement Unit) and distance Measurement, so that the unmanned aerial vehicle performs local mapping through an algorithm to obtain a point cloud model, thereby performing positioning. The UKF (Unscented Kalman Filter) is used to provide a 6 degree of freedom estimate of the unmanned aerial vehicle attitude by fusing the data from the IMU, two range sensors and four cameras with a pre-known 3D occupancy grid map. Based on the SLAM method, a cylindrical model is fitted and applied to point clouds obtained from heterogeneous sensors by combining a range estimator and a vision estimator, and a map is obtained by utilizing preliminary flight. The tunnel axis is then estimated from the local map and the drone position is determined along the tunnel axis and then used to guide the drone. This application is guiding unmanned aerial vehicle in-process and considering system dynamics to more be applicable to under the tunnel dark surrounds, and need not light stream information, can effectively guarantee that unmanned aerial vehicle is in the tunnel center all the time. This is completely different from the related art, such as an indoor unmanned aerial vehicle multi-sensor combined navigation method based on unscented kalman filtering, which cannot realize the guided flight of the unmanned aerial vehicle in a dark tunnel environment without light flow. When collecting images for inspection, a drone with a rotating camera may be used to minimize field of view (FOV) obstructions. The measurements from the laser ranging sensor array are used for localization to estimate the position and heading of the drone in the tunnel and to know its geometry in advance. A method of estimating the tunnel axis using sensor array measurements is based on maintaining the local position of the drone in the center of the tunnel. Unmanned aerial vehicle tunnel flight mainly passes through the axis guide mode. Further, optical flow sensors are used with time-of-flight distance sensors to estimate the distance traveled along the tunnel axis.
S102: in the process of building the three-dimensional tunnel map, degradation characteristics of the surface of the tunnel and the interior of the tunnel are detected at the same time, and the three-dimensional tunnel cloud map for identifying the position information of the area to be repaired is generated based on the three-dimensional tunnel map and the degradation characteristics.
In the step, degradation characteristics such as stripping, cracks, cavities and the like of the surface and the interior of the concrete tunnel are detected simultaneously in the process of constructing the three-dimensional tunnel map. Data collection is characterized by radar, such as SLAM lidar, and ground penetrating radar results are stated by analyzing surface and near-surface features. Short-time fourier transform (STFT) is used to detect free water within the lining, which in this case produces ground penetrating radar scattering and energy absorption phenomena. The results of SLAM lidar supplemented the GPR data. The location where water is detected is a pure blue color and the red line represents the detection threshold. The method comprises the steps of forming a combined frame through synchronous positioning, mapping laser imaging technology (SLAM LIDAR) and Ground Penetrating Radar (GPR), ranging, underground moisture and detected degradation characteristics of the concrete tunnel through cooperative processing, constructing a three-dimensional graph, framing and marking the crack at the spatial position of the three-dimensional cloud graph, and roughly determining the position and trend of the crack.
S103: and positioning each crack to be repaired of the tunnel to be detected by using the wall-climbing robot based on the three-dimensional tunnel cloud picture, and moving the crack to be repaired to a target crack to be repaired for negative pressure adsorption and perfusion.
The utility model provides a wall climbing robot deploys a plurality of sensors and vacuum perfusion instrument, also the wall climbing robot can be for full negative pressure wall climbing robot, compares traditional wall climbing robot and like roller type wall climbing robot, and vacuum adsorption can form the negative cavity, is more difficult to drop, is favorable to the smooth execution of task. Compare roller type wall climbing robot, the wall climbing of vacuum adsorption design is restoreed robot and corresponding vacuum repair means, also detects the robot collaborative means that restores the integration or detect and deal with the integration and compare in jointly patrolling and examining the mode, and crack repair effect and efficiency are all higher. The wall-climbing robot acquires and understands environmental information data through the sensors, and compares and learns the detected data and the collected data in the actual execution process, so that the distribution situation of the surrounding cracks is detected. The wall climbing robot of this embodiment is in same three-dimensional tunnel cloud picture with unmanned aerial vehicle, and coordinate position information is the same, utilizes the produced three-dimensional tunnel cloud picture of unmanned aerial vehicle, and the wall climbing robot can be from the barrier, and the automatic planning route climbs in tunnel lining surface. In the generated three-dimensional cloud pictures, the unmanned aerial vehicle is mainly used for roughly identifying cracks, and the wall climbing robot carries out detail processing, precisely positioning the cracks and grouting in the roughly identified cloud pictures. The operation mode of establishing the three-dimensional space cloud picture and sharing the positioning information by the multiple robots is completely different from the mode of repairing the crack executed by the unmanned aerial vehicle in the related technology, and the effect is better. In the field of automatic driving, path planning refers to planning an effective path which has no collision and can safely reach a target place according to a certain condition or a certain performance index. Two steps may be included: the method comprises the steps of obtaining an environment map through environment perception, and searching for feasible paths in the environment map by using a search algorithm. The wall-climbing robot identifies and detects the crack, so that the wall-climbing robot always walks along the center of the path and identifies the crack after reaching the position area of the crack. The wall climbing robot has the advantages that the wall climbing robot is adsorbed on the wall surface by adsorbing and moving two materials and is not limited by the materials of the wall surface, so that the materials can be poured in a negative pressure mode, and grouting can be stopped on the surface under the action of a vacuum environment, so that the wall climbing robot can go deep into the deep part of a crack and does not leave a large number of pores. As shown in fig. 3, 31 is the inlet of the repair material, 32 is a spiral tube wrapped with a peeling layer, 33 is a sealing band, 34 is a vacuum outlet, 35 is a T-shaped port, 36 is a peeling layer, 37 is a filter protective sleeve, 38 is a mold, and 39 is a reinforcing layer, which is pumped to vacuum by a self-contained vacuum pump, and then the material is poured into and cracks. The process uses atmospheric pressure to drive the resin through the dry reinforcement stack under vacuum. The resin simply flows into and through the fiber mass, wetting as it passes. The fibres are under a vacuum bag so that everything is well consolidated and pressed against the mould. The crack is repaired to a vacuum environment similar to a vacuum bag device. A vacuum infusion process: preparing a mold, constructing a gel coat surface, paving a reinforcing material, paving a vacuum auxiliary material, vacuumizing, preparing resin, introducing the resin, and demolding. The air is discharged by adopting vacuum infusion repair to form a negative pressure cavity, the repair material is filled in the crack under the negative pressure environment, the crack can be deeply penetrated, the repair effect is better, and the concrete surface is not required to be punched. Not only easy operation punches moreover and firstly causes the destruction of structure, secondly the outward appearance is pleasing to the eye, thirdly needs a large amount of manpowers to operate, especially is located tunnel lining's top.
Further, after the wall climbing robot completes repairing of all to-be-repaired cracks of the to-be-repaired tunnel, the unmanned aerial vehicle flies to the tunnel again, the steps S101 and S102 are executed, the modification condition is evaluated according to the generated three-dimensional tunnel cloud picture, and if the cracks which are not repaired still exist, the wall climbing robot executes the step S103 to repair the cracks, so that the safety and the stability of the tunnel structure are ensured.
In the technical scheme provided by the embodiment of the invention, the unmanned aerial vehicle generates a cloud picture for identifying the approximate region of the tunnel where the crack needs to be repaired based on the data acquired by the carried sensor, the wall-climbing robot determines the approximate path by using the cloud picture, detects and identifies the crack through the sensor of the wall-climbing robot, moves to the initial position of the crack, and performs detail detection and repair. The whole crack repairing process is convenient and quick, a large number of maintenance personnel are not needed for operation, labor cost is saved, and the risk of harm to the maintenance personnel caused by unfavorable environment can be avoided. Wall climbing robot fills the restoration based on the vacuum, discharges the air, forms the negative pressure cavity, and it is better to fill the effect, can go deep into the crack, and need not punch the operation, further promotes crack repair efficiency to can be safe, high-efficient, effectively repair the tunnel crack, and then promote tunnel structure's stability and security.
It should be noted that, in the present application, there is no strict sequential execution order among the steps, and as long as the logical order is met, the steps may be executed simultaneously or according to a certain preset order, and fig. 1 is only an exemplary manner, and does not represent that only the execution order is the order.
In order to improve tunnel structure's stability, avoid the low easy problem that leads to crack repair effect poor such as the filling material toughness, based on above-mentioned embodiment, this application has still given the optional filling material that adopts under the negative pressure absorption filling mode, can include:
mixing nano silicon dioxide and epoxy resin in advance according to a first preset proportion, or mixing maleic anhydride, a graphene sheet material and epoxy resin according to a second preset proportion to generate composite epoxy resin; adding polyvinyl alcohol fiber or double-wall microcapsule material into the composite epoxy resin to modify the composite epoxy resin;
and in the advancing process of each preset repairing path of the crack to be repaired according to the path, when the crack to be repaired moves to the current crack to be repaired, the wall climbing robot pours the modified composite epoxy resin into the current crack to be repaired under the negative pressure environment.
According to the embodiment, the epoxy resin is modified by the nano-silica, so that the toughness of the epoxy resin is improved, the nano-silica is doped, the thermal expansion coefficient is reduced, the Young modulus is increased, the critical stress intensity factor is increased, the nano-particles can achieve a good toughening effect and need to be uniformly dispersed in the epoxy resin, and the nano-particles can act on the surface of the silica through the photosensitive azide. Ultraviolet radiation is controlled to control the properties of the interface. The interface effect with epoxy resin can be increased by the maleic anhydride modified graphene sheet material, so that the strength is improved, and the thermal stability coefficient is reduced. Furthermore, PVA fiber is adopted to reinforce concrete repair, enhance the surface bonding strength and simultaneously improve the chloride ion permeability resistance and the fracture toughness. The double-wall microcapsule is added into the composite epoxy resin, and self-repairing can be realized in later-stage damage, namely when the composite film formed in the current stage is damaged, the healing agent flows out, and cracks are bonded after curing.
In the foregoing embodiment, how to perform step S101 is not limited, and an optional implementation manner of this step may include the following:
and according to unscented Kalman filtering, generating a three-dimensional tunnel map and determining the position of the unmanned aerial vehicle in real time by combining the data processing results of laser scanning data of the laser radar, ranging data of the ultrasonic range finder and measurement data of the inertial measurement unit with image data acquired by the infrared image acquisition equipment. Optionally, as shown in fig. 4, laser scanning data is obtained, and a three-dimensional tunnel map is constructed according to the aligned laser scanning data; determining the flight speed of the unmanned aerial vehicle according to the laser scanning data of the unmanned aerial vehicle at different moments; determining absolute position information of the unmanned aerial vehicle in a flight area based on each laser scanning data and the latest available map state; and fusing the flight speed of the unmanned aerial vehicle, the gravity adjustment acceleration of the inertial measurement unit and the absolute position information.
Based on the three-dimensional tunnel map and the measured value of the depth of the tunnel wall of the tunnel to be measured, the flight path at the next moment is automatically determined by combining the unmanned aerial vehicle motion model according to the mode that the unmanned aerial vehicle moves towards the estimated three-dimensional point on the tunnel crankshaft and moves in the tunnel to be measured step by step.
The embodiment also provides an alignment mode, and firstly, corresponding relations are established among all points of the laser scanning data at different moments; and determining the conversion information of the laser scanning data aligned to the current point of the laser scanning data at the current moment through a minimized error function, and applying the conversion information to each point of the aligned laser scanning data until the alignment condition of the laser scanning points is met.
The unmanned aerial vehicle of the embodiment autonomously plans a path in a dark environment and performs SLAM modeling, and the whole process can include the following contents:
in calculating the speed of the drone, the present embodiment does not directly integrate the relative displacement, but uses the known timestamp of each laser scan data to calculate, i.e., may useIs calculated, whereinIs the speed at the time of the t-time,the difference in displacement between the two laser scan data aligned during the sequence matching process,the time difference between two laser scan data that are aligned in the sequence matching process. At the same time, the laser scanning data at the time tAlso aligned by global matching with a global map, which is built up step by step in a mapping module to obtain a global position estimate. Two estimated valuesAndfusing in Kalman filter to generate final drift-free position estimateI.e. the position estimate at time t.
In the estimationAndbefore, eachA pre-treatment is performed. The processed scan being a subsetWherein,the original laser light is scanned by the laser light,is the subset after the scanning process and,the closing point is removed from the subset of,the background-removed subset is then selected,the noise-removed subset of the noise is,the external processing removes the subset. The polar coordinates are converted into rectangular coordinates, and the attitude and the height of the unmanned aerial vehicle are converted. The threshold is set by setting h (height).
In autonomous planning of a flight path of a drone, it is necessary to be able to reliably estimate the time of the drone in any environment with obstacles(complete set of positive real numbers)Andthe difference between. FindingCorresponding to the search respectively inAndlaser scanning pattern of position shootingAnda rigid two-dimensional transformation between the representations,,is composed of,The position of the moment of time is located,,is composed of,Scanning laser pattern at time instant.
Wherein argmin is a function that minimizes the target, the argument function beingThe value of the compound is within the range,is laser scanningThe point (i) of (2) is,is laser scanningJ (th) point of. OperatorIs the number of points in the scan Z,is corresponding toIndependent variableIs thatFinding a point distanceNearest and squared the distance.
B. The transformation of the alignment laser scan is calculated by minimizing the error function.
To estimate the error magnitude for the error accumulation value,as in the above-mentioned meaning, it is,is composed ofAn initialized identity matrix.
D. And evaluating the alignment quality, and iterating all the steps until a stop condition is met, namely a laser scanning point alignment condition is met.
Whether the stop condition is met is checked by the difference in FRMSD (a new distance measurement method-fractional mean squared distance) between the current iteration and the last iteration. Since the error generally decreases exponentially, the difference between successive errors is compared to a constant close to zero, and if the difference is small, the algorithm has converged, and if too long, convergence is considered as well.
Global matching is used to estimate the absolute position of the drone within the flight area. To achieve this goal, a current pre-processing scan is requiredAnd the latest available map statusTo locate the drone. The module outputs a conversion matrixAnd the obtained absolute position estimation value is transmitted to a Kalman filter to be fused with other measurement values. The map of the present embodiment may be stored in an octree data structure with a resolution limited to r =0.2m. When the current state of the map is queried, the module generates a point cloud representation M. The mapping module also retains a dense point cloud comprised of all successfully aligned scan results, which can be output as a side of the drone performing a task in an unknown environment. The outputs of the sequential and global matches are fused in a linear kalman filter as measurements of velocity and position, respectively. Due to the position of the unmanned planeAre independent of each other, so a decoupled estimate can be made for each axis, which reduces the size of the state space description. This approach uses another tri-state model for estimating altitude for a two-dimensional drone(As a height direction vector, from the height direction positionSpeed ofAcceleration ofConstituent(s) of the filter. The gravity measured by the IMU adjusts the acceleration driving the prediction step of the filter and the measurements from the downward laser rangefinder are fused into the filter as a correction.
The path autonomous planning of the unmanned aerial vehicle of the embodiment can process the movement in the tunnel type environment in a reactive mode, the shape and the direction of the environment are changed in three dimensions, and the unmanned aerial vehicle autonomous planning is different from any existing two-dimensional reactive method and limits the movement of the unmanned aerial vehicle to a certain fixed height. In contrast to the existing two-dimensional planning-based method, the embodiment can provide a solution with a small calculation amount for the autonomous navigation problem. The motion decision is mainly used for guiding the unmanned aerial vehicle based on the existing measurement value of the airborne sensor, does not need accurate positioning and is suitable for the environment without a GPS. Optionally, the unmanned aerial vehicle is moved towards the estimated three-dimensional point on the tunnel crankshaft, and the unmanned aerial vehicle is moved gradually in the tunnel. These points are interpreted from existing depth measurements of the tunnel wall, and may be represented in a sensor-fixed coordinate frame, for example, in the form of a three-dimensional point cloud. This is suitable for an environment with few obstacles. However, by combining this with reactive obstacle avoidance using behavior-based control methods, the method can be extended to obstacle-filled environments. The general kinematic model is used for development, so that the applicability of the kinematic model is expanded. Noise measurements are also taken into account, using different sensor configurations and sensing algorithms. The autonomous navigation process of the drone may include:
firstly, establishing an unmanned aerial vehicle motion model,is a three-dimensional vector of rectangular coordinates defined by the drone in the world (inertial) frame of reference, the motion of which is described by equations.
(1),Is a linear velocity vector, the vector variable V (t) is the control input,is the linear velocity of the drone in the world (inertial) reference frame.
(2)Control input V (t) at discrete times 0, delta, 2 delta, 3 delta 8230whereIs the sampling period.
What is considered in this embodiment is a fairly common three-dimensional problem, namely that drones navigate autonomously and avoid collisions in an unknown tunnel-like environment. The aim is to design the navigation method for the drone so that it safely passes through the deformed cylindrical or annular tunnel T.
Is a given constant, d (T) represents the distance between the coordinates c (T) of the drone and the wall of the deformed cylindrical or toroidal tunnel T,in order to set the safety distance,is at the positionThe curvature coordinate of (2). If (3) and (4) are satisfied, that is,And are andtime of flightThen one unmanned navigation method is called navigation through the deformed cylindrical tunnel T safely. Further, if (3) holds, and for anyIn the presence of a sequenceUnmanned navigation is said to safely pass through a deformed annular tunnel T, Q being a collection of lengths L, where L is>0 is the length of the closed-axis curve C, so that. Equation (4) means that in the case of a deformed cylindrical tunnel the drone will travel to infinity within the tunnel, and equation (5) means that in the case of a deformed annular tunnel the drone will make an infinite number of cycles within the tunnel.
In the following assumptions, the deformation tunnel may be cylindrical or annular. Assuming existence of a constant,,,So that. Is introduced fromGo out toVector. In addition, letRepresenting from point a to the connectionAndthe distance of the straight line of (a).,Being the centre of gravity of the set of tunnel wall points. a is an arbitrary point and F is a certain vector.
Introducing vectorsSo that the vectorAndis equal to an angle between, and. As can be seen from the construction, it is,. Now, the following navigation rules and the following rules are introduced:
(6)
for k =1,2, \ 8943. M1 is a cylindrical ring path, M2 is a deformed cylindrical ring path, is a one-dimensional real number set,is an infinitely small number. Two different vector representations are introduced for two different tunnels.
Can be proved by a series of mathematical inference theories withThe unmanned aerial vehicles (2) and (6) can safely navigate in the deformation tunnel T.
With the navigation method provided by the embodiment, the unmanned aerial vehicle may sometimes deviate from a short section when the boundary direction of the moving tunnel passing through changes sharply. However, it still manages to keep a safe distance from the wall, see the drone navigation diagram shown in fig. 5.
In a specific implementation, the Robot Operating System (ROS) framework of the drone is used to implement the entire navigation software stack as connected nodes (i.e., processes running simultaneously), each node handling a specific task. An unmanned aerial vehicle control node implements the trajectory tracking controller described in the section to generate thrust and attitude commands for the low-level attitude controller. Transmitted to the flight controller unit via the body-fixed frame (B). The visual odometer received from the camera is also sent to the FCU (File Control Unit), fused with IMU data through an extended kalman filter. In addition, the camera nodes are used to process the 3D point clouds received from the depth camera so that they are available to other nodes, with an update rate of 30 hz. These algorithms run at an update rate of 2-10 Hz and are implemented in C + + using useful tools in the Point Cloud Library (PCL) to handle point cloud processing in a computationally efficient manner. A downsampling filter using PCL VoxelGrid is applied to the 3D point cloud to reduce the computational burden and, in conjunction with some other filtering process, the obtained integrals G1 and G2 are used from the reference trajectories of the previous algorithm to generate the reference trajectories to be sent to the drone control nodes.
The embodiment provides a new collision-free autonomous navigation method for an unmanned aerial vehicle flying in an unknown 3D tunnel dark environment. The method can be used in the absence of GPS signals and has a small amount of calculation. These environments change shape and orientation in 3D in a reactive way, unlike existing 2D reactive methods that limit drone motion to a certain fixed height. And the available surrounding environment map is utilized, so that the calculation efficiency is improved. When the unmanned aerial vehicle tunnel flies and collects crack data, the camera, the laser radar, the ultrasonic ranging and IMU collection parameters are respectively a module, and collected data are coupled through a multi-mode fusion algorithm, so that the unmanned aerial vehicle positioning and the three-dimensional cloud picture construction are realized. Additionally, SLAM-based methods, in conjunction with range and vision-based estimators, utilize maps obtained from preliminary flights. An algorithm is proposed to perform local mapping, where a cylindrical model fit is applied to the point cloud obtained from heterogeneous sensors. The tunnel axis is then estimated from the local map and the drone position is determined along the tunnel axis and then used to guide the drone.
In the embodiment, the wall-climbing robot roughly determines the range of the crack by using the initially generated three-dimensional tunnel cloud picture, selects the safest path, avoids the crack from falling due to the collision with a large obstacle, and further observes and identifies the crack through the self-sensor. Thereby grouting and repairing. And the approximate position of the crack is preliminarily judged by utilizing the three-dimensional cloud picture, and the wall-climbing robot is controlled to be adsorbed to the ground, so that the crack is locally amplified and tracked. One popular form of motion planning that takes uncertainty into account is an opportunistic constraint approach that keeps the collision probability of a path below a given threshold. The opportunity constrained approach has been combined with RRT and Markov Decision Processes (MDP) to achieve robustness against collisions. Optionally, the uncertainty of each position may be expressed by using particles in advance based on the uncertainty of the position of the wall-climbing robot, and the uncertainty of each position may be extended by using the RRT, and the path planning rule of the wall-climbing robot may be determined by using the particles to express uncertainty propagation in a tunnel environment constructed by various types of terrain; generating a wall climbing path of the wall climbing robot based on the three-dimensional tunnel cloud picture and the path planning rule; and calling a crack identification model to identify whether a crack to be repaired exists at the current position in the process that the wall-climbing robot travels according to the wall-climbing path. The obstacle avoidance function and the path planning of the wall-climbing robot realized through the opportunity constraint algorithm are different from the existing obstacle and path planning technology, and compared with the existing obstacle avoidance method and the existing path planning method, the obstacle avoidance method has the advantages that the obstacle avoidance effect is better, and the finally planned path is more excellent.
In the embodiment, data collected by the sensors deployed at the bottom layer of the wall climbing robot is processed through four steps for subsequent path planning. The four steps are respectively as follows: a dual stream network, a shared system, a shared feature transmission network, and add modules. Double-flow network: the feature extractor obtains features of two modalities. The sharing system carries out uniform feature representation on the extracted features. The pairwise affinity model is built using commonality and characteristics with the goal of correlating each sample direction within and between modalities. A uniform large square shape can be spliced, two opposite angles are characteristics, and the two opposite angles have the same property. Shared feature transport network: intra-modality and inter-modality similarities are determined and shared and specific features are propagated among different modalities to offset the lack of specific information and enhance shared features. Two project confrontation reconstruction blocks and a mode adaptation module are added on the module of the feature extractor to obtain distinctive and complementary shared features and specific features. In determining the approximate location of the fracture to be repaired, the process of pinpointing may include: and estimating the attitude of the wall-climbing robot by adopting a VLP-SLAM (Vision phase Pre-training-Simultaneous Localization and Mapping) fusion method, and performing laser radar obstacle detection and map data path planning. ROS (Robot OS Robot operating system) mobile bases are used to achieve autonomous navigation. The method comprises positioning information fused by multiple sensors and provides a moving command for the moving base of the wall-climbing robot so that the wall-climbing robot can safely move in the environment. Low drift motion estimation can be obtained by multi-sensor fusion, in particular lidar scanning compensation. The defects of independent sensors can be overcome by adopting multi-sensor fusion, and the local navigation is more stable. The interaction between VLP (Vision navigation Pre-tracking) and LiDAR-SLAM (radar-Simultaneous Localization and Mapping) provides reliable and accurate pose estimation. The output three-dimensional coordinates coincide with the three-dimensional coordinates of the unmanned aerial vehicle modeling diagram.
The wall climbing path of the wall climbing robot is considered in a two-dimensional coordinate system, and the center of the coordinate system is the current position of the robot. Commanding the robot to travel straight aheadExpressed as a vectorAnd T is a transposed matrix. Position of the robot after driving in a coordinate systemAnd (4) showing. Responding to command distanceUncertainty of motion ofIs represented as follows:
The motion uncertainty model depends only on the type of terrain and is expressed as follows:
t represents a type of terrain that is to be imaged,representing a terrain T by a specified distanceThe uncertainty of (a) is determined,the percentage of the distribution of the uncertainty,variance-variance of uncertainty distribution-oblique variance matrix.,Represents a longitudinal direction andthe average value in the transverse direction is,,representing the longitudinal and lateral standard deviations.Also interpreted as the average slip rate. Typically, the actual distance of the drive distance is less than the command distance, and therefore,Is set to 0.
The RRT (Rapid-expanding Random Trees) expands the graph to Random sampling points. The map expands rapidly due to random sampling, but position uncertainty easily increases due to rotation in terrain where motion uncertainty is large. To solve this problem, the proposed method extends a graph from nodes with small position uncertainty. The position uncertainty of the node is represented by a particle. The same number of particles is assigned to each node. Uncertainty propagation is described by node extension moving particles. The pseudo code MU-RRT of the proposed RRT considering Motion Uncertainty (Motion Uncertainty-rapid-exploiting Random Trees, fast search Random Trees considering Motion Uncertainty) is shown as algorithm 1 shown in table 1 below, and the implementation process of the algorithm is:
1: the graph is initialized with a starting point s and starting positions are assigned to all particles of the starting point.
2: to randomly sampled point p, n nearest neighbor points are selected from the graph,,…,. If the number of all nodes is less than n, all nodes are selected.
3: will be provided with,,…Node in,,…,At a command distance of the particleMoving downwards in the direction of p to obtain expanded candidate nodes,…,And their particles,…,。,…,Are respectively in the positions of,…,Average position of the particles. Computing position uncertainty of candidate nodes,…,。
4: selecting a candidate node with a minimum position uncertaintyIf, ifIf not determined to be a collision, willAdded to the figure. RetentionThe position of the particle. If a collision occurs, go back to 2.
5: if it is usedAnd meeting the requirement of reaching the target area, stopping the algorithm and returning the graph. Otherwise, go back to 2.
TABLE 1 Algorithm 1
The method of propagating uncertainty using particles during step 3 in algorithm 1. In moving the particles, two cases should be considered. One is that the particles move in the same type of terrain, and the other is that the particles move in multiple types of terrain.
First, consider that the particles are in the same type of terrain T fromMoveThe situation comes. The movement of the particle in the global coordinate system, as defined by the motion uncertainty, can be written as follows:
wherein,indicating the commanded direction from the x-axis,andrespectively representAndthe position of the kth particle.To representThe rotation matrix of (a) is set,represents a unit vector pointing in the x-axis direction,representing a terrain T by a specified distanceUncertainty of (2).
Next, a case where one particle moves on various types of terrain is considered. As shown in fig. 7, one particle is in turn on...,And (4) moving upwards. When the particles move straight in the direction of the command, on the terrainExpressed as actual driving distance in. Under this assumption, the motion of the particles can be expressed as:
wherein,wherein is the topography of the particleIn the middle toCommanded distance while moving. Actual driving distanceLess than instruction distanceDue to average slip. (14) Means thatIs equal to。(15)Representing the uncertainty of the specified distance. The motions represented by equations (13) - (15) are an approximation. Equations (13) - (15) assume that the particles are straight, but in reality the robot will move meanderingly. However, straight-driving is the most likely case, and therefore, the motion of the particles is approximated as straight-driving, and the calculation time is reduced compared to the monte carlo simulation.
First, calculateVariance-covariance matrix of medium particle positionsIt is a true infinite number of things, that is,diagonalizable, so that diagonalization is calculatedAnd introducing an orthogonal matrixAnd diagonal matrix。Position uncertainty ofByWriting into:
If it is usedAndthe line segment in between touches an obstacle, then a collision is observed. Otherwise, ifIs observed when the number of particles in the obstacle region exceeds a threshold valueIs a collision. To be exact, in judging a collision, consideration should be given toAndall line segments between a pair of particles in (a). However, when the length of the branch is short, the collision can be approximately judged in this way. Otherwise, if outside the target areaThe number of particles is less than the threshold, then observed as reaching the target region.
Algorithm 1 is an MU-RRT algorithm that reduces collisions and accumulated position uncertainty, with the algorithm parameters as shown above. Distance to instructionUnlike that used in the experiment, any command distance is available because the motion uncertainty parameter is defined as a percentage. Target sampling rateIs the probability of selecting a point from a given target region at random sampling. Based on the proposed MU-RRT, algorithm 1 can generate a path with small position uncertainty. The planned path does not include large angular rotations of regions of greater uncertainty. When the rotation is performed, the angle is small and smooth.
The embodiment is used for considering the motion uncertainty of the wall climbing robot in the rough environment. The method proposed by the present embodiment uses particles to express uncertainty propagation in complex environments built from various types of terrain. Meanwhile, RRT (fast-exploration random tree) is extended based on the uncertainty of each node to prevent increasing the accumulated position uncertainty. Thus, the generated path reduces the time for path tracking and re-planning based on inaccurate positioning information. The method reduces the position uncertainty, keeps the probability of avoiding collision and reaching the target area, and reduces the falling risk of the wall-climbing robot to the maximum extent.
In order to further improve the crack repairing accuracy, an indispensable step is an accurate crack identification, the above embodiment does not set any limit to the identification model used for identifying the crack, and this embodiment further provides an optional implementation manner, which includes the following steps:
training a crack recognition model in advance based on a mode that information of each loss function is respectively reflected to a network to update a weight;
the crack identification model comprises an input layer, a convolution layer, a plurality of Dense blocks (namely, a Dense Block) and an output layer; the output profile of each Dense Block is input to the next Dense Block and output layer simultaneously.
The fracture identification model of the embodiment is based on an image segmentation technology, and can improve the efficiency of a new deep neural network by utilizing a hierarchical convolutional neural network. The present embodiment may effectively improve the crack detection performance by training the deep neural network using the multi-loss update method, instead of using the conventional method of updating the weights by summing the loss values, by using a method of updating the weights by reflecting information of each loss function to the deep neural network, respectively. To overcome the speed disadvantages of pixel-based segmentation and improve accuracy, a lightweight segmentation network is employed with an encoder structure that uses a hierarchical convolutional neural network CNN.
In the embodiment, the crack identification model deletes the decoder part of the traditional network structure, so that the operation speed can be improved. As shown in fig. 8, the model uses hierarchical CNNs, performs learning based on several sub-outputs of each layer, and uses the sub-outputs and the label image to derive loss values and update weights on that basis without diversifying the encoder structure. A simple bottleneck network structure is adopted, and only one convolution layer with the kernel size of 1 is adopted, so that the channel size of the characteristics is reduced to 2, the calculation speed is increased, and the identification performance is improved. The final size of the encoded image used in the present embodiment is 1, 0.5, 0.25, and 0.125 times the size of the input image. To test the performance of the trained network, an additional process is required to obtain the results. Each sub-output is normalized using the softmax function. After summing the results obtained from each step, a process of obtaining an average value is performed. As a result, the value of each pixel is between 0 and 1, and the pixel value of 0.5 is set as a threshold value for determining the presence of a concrete crack. Therefore, pixels having a value greater than 0.5 are identified as cracks and displayed in the segmentation result, being divided into cracked and non-cracked portions. The fracture identification model is based on DenseNet-121 (dense connection convolution network), and four original conditions of the fracture identification model are modified, and details of the architecture are shown in Table 2.
TABLE 2 parameter Table for crack recognition model architecture
First, the pooling process of the initial input images in the network is eliminated. The size of the image is reduced by half by the pooling process. Therefore, more features can be learned through the transition layer and the Dense Block, resulting in an image that is twice as large as when using the original network. Second, the composition of the dense mass is changed. The dense block of the original DenseNet has a bottleneck layer connected to the BN-corrected linear unit (ReLU) - (1 × 1) convolution (conv) -BN-ReLU- (3 × 3) conv. The number of repetitions per dense block of the layer is modified from (6, 12, 24, 16) to (4, 3,2, 1). By reducing the number of repetitions, the amount of computation and the inference time are reduced. Further, the (1 × 1) convolution layer conv of the bottleneck layer is modified to the (3 × 3) convolution layer conv. This is configured by adjusting the larger kernel size, including not only the features of each pixel, but also the relationship to neighboring pixels for learning. Third, (1 × 1) conv in the transition layer is converted into (3 × 3) conv. Finally, the growth rate is modified from 32 to 4, which results in a significant reduction in computational effort. As shown in fig. 10, a normal neural network is trained based on the principle of multiplying the value of each pixel of the input image by the value of the weight (i.e., the circle). Therefore, the change of the output image depends on the weight value multiplied by the input image. The goal of deep learning is to make the output image identical to the labeled image; to achieve this, the values of the weights are changed. This behavior is defined as learning. In this case, the loss function acts to update the weight value so that the output image and the annotation image are the same. Briefly, the effect of this loss function is that if the output image is the same as the annotation image, the loss function value becomes zero, so the greater the difference between the two images, the greater the loss function value.
After the wall climbing robot completes crack identification based on the above embodiment, path tracking of a wall climbing path can be realized by adopting a decoupling control method, the method relies on continuous state transformation, decoupling is controlled by using homogeneous finite time, control design is simplified, and the implementation process of the decoupling control method can include:
a partial conversion brings the system into a chain form. Later, the error tracking dynamics were transformed into a triangle-like form. In this way, the system can be divided into two interconnected subsystems. This embodiment also proposes a homogenous finite time controller to stabilize the rectangular coordinate x, thus, influencing x to decouple from the residual dynamics. A method similar to recursive back-stepping then stabilizes the remaining coordinates. The problem to be solved by the decoupled control method is the finite time stability of one of the cartesian coordinates. The present embodiment employs a more straightforward design that improves performance with faster convergence rates, ensuring gradual convergence to the desired trajectory. A so-called chain model (16) is used, as follows:
In which, as shown in figure 11,,are in the form of cartesian co-ordinates,is the angle of orientation with respect to the X-axis,is the angle of the turn-around angle,andis a control input to the control unit,in correspondence with the line speed,is the speed of the steering wheel, and,is the distance between the front and rear wheels. Taking into account trajectory limitations and tracking error.Represents the desired coordinates, i = 1.., 4;representing a reference control input. Definition ofIn order to track the coordinate errors,for outputting the error of the result, the following relation 17-20 is given:(17)、(18)、(19) And(20). The corresponding desired trajectory converges toAnd a series of differential processing is performed to rewrite the equation by a new function z:
a reference control input is shown that is referenced to,,speed error, reference control signalPositive at all instants in time, that is,. The above assumption means that the desired wall climbing robot speed is different from zero and the wall climbing robot moves forward. Starting from the subroutine (24), a continuous homogeneous controller, i.e. the relation (25), is proposed:
wherein,,is a symbolic function. Proposing the function V of Lyapunov 4 =|z 4 L is obtained by differentiating along the track of the subsystem (24)(26). Therefore, the temperature of the molten metal is controlled,is limitedTime of day ofIt is given.
According to a recursive programmingTo stabilize subsystems (21) - (24). In this case, the variablesAndis considered to be a function of time variation. Thus, whenThe result of subsystems (21) - (24) is a strict feedback control structure. Therefore, the temperature of the molten metal is controlled,the design of (c) will involve a method similar to feedback.
Wherein, the Lyapunov function, V is the Lyapunov function,,,is three different variables,,Standard deviation.
The trajectories are integrated and utilized virtually, for a given formula (26), anThrough a series of mathematical changes, it can be derived:
. Therefore, the temperature of the molten metal is controlled,to a. Thus, it is possible to provide. It is straightforward to say that. And forThe same may be true.
The embodiment of the invention also provides a corresponding device for the automatic tunnel crack repairing method, so that the method has higher practicability. Wherein the means can be described separately from the functional module point of view and the hardware point of view. In the following, the automatic tunnel crack repairing device provided by the embodiment of the present invention is introduced, and the automatic tunnel crack repairing device described below and the automatic tunnel crack repairing method described above may be referred to in correspondence with each other.
Based on the angle of the functional module, referring to fig. 10, fig. 10 is a structural diagram of an automatic repairing apparatus for tunnel cracks according to an embodiment of the present invention, in a specific implementation manner, the apparatus may include:
the data acquisition module 101 is used for acquiring multi-mode data in the flight process of the light source-free GPS-free tunnel to be detected based on the autonomous planning path by using an unmanned aerial vehicle carrying multiple types of sensors for acquiring image data and positioning data so as to construct a three-dimensional tunnel map;
the initial positioning module 102 is configured to simultaneously detect degradation features on the surface of a tunnel and in the tunnel in a three-dimensional tunnel map construction process, and generate a three-dimensional tunnel cloud map for identifying position information of an area to be repaired based on the three-dimensional tunnel map and the degradation features;
and the grouting repair module 103 is used for positioning each crack to be repaired of the tunnel to be repaired based on the three-dimensional tunnel cloud picture by using the wall-climbing robot, and moving the crack to be repaired to the target crack to be repaired for negative pressure adsorption and grouting.
Optionally, in some embodiments of this embodiment, the apparatus may further include a material preparation module, configured to mix, in advance, the nano silica and the epoxy resin according to a first preset ratio, or mix, in advance, the maleic anhydride, the graphene sheet material, and the epoxy resin according to a second preset ratio, so as to generate a composite epoxy resin; adding polyvinyl alcohol fiber or double-wall microcapsule material into the composite epoxy resin to modify the composite epoxy resin; correspondingly, the grouting repair module 103 is further configured to, in the advancing process according to the preset repair path of each to-be-repaired crack, when the to-be-repaired crack moves to the current to-be-repaired crack, fill the modified composite epoxy resin into the current to-be-repaired crack in the negative pressure environment by the wall climbing robot.
Optionally, in other embodiments of this embodiment, the data acquisition module 101 may be configured to: according to unscented Kalman filtering, generating a three-dimensional tunnel map and determining the position of the unmanned aerial vehicle in real time according to data processing results of laser scanning data of a laser radar, ranging data of an ultrasonic range finder and measuring data of an inertia measuring unit and image data acquired by infrared image acquisition equipment; based on the three-dimensional tunnel map and the tunnel wall depth measured value of the tunnel to be measured, the flight path at the next moment is automatically determined by combining the unmanned aerial vehicle motion model according to the mode that the unmanned aerial vehicle moves towards the estimated three-dimensional point on the tunnel crankshaft and moves in the tunnel to be measured step by step.
As an optional implementation manner of the foregoing embodiment, the data acquisition module 101 may be further configured to: constructing a three-dimensional tunnel map according to the aligned laser scanning data; determining the flight speed of the unmanned aerial vehicle according to the laser scanning data of the unmanned aerial vehicle at different moments; determining absolute position information of the unmanned aerial vehicle in a flight area based on each laser scanning data and the latest available map state; and fusing the flight speed of the unmanned aerial vehicle, the gravity adjustment acceleration of the inertial measurement unit and the absolute position information.
As another optional implementation manner of the foregoing embodiment, the data acquisition module 101 may be further configured to: establishing a corresponding relation between all points of laser scanning data at different moments; and determining the conversion information of the laser scanning data aligned to the current point of the laser scanning data at the current moment through a minimized error function, and applying the conversion information to each point of the aligned laser scanning data until the alignment condition of the laser scanning points is met.
Optionally, in still other embodiments of this embodiment, the grouting repair module 103 may further be configured to: the method comprises the steps that particle representation is used in advance based on position uncertainty of the wall climbing robot, meanwhile, RRT is used for expanding uncertainty of each position, uncertainty propagation in a tunnel environment constructed by various types of terrains is expressed by the particles, and a path planning rule of the wall climbing robot is determined; generating a wall climbing path of the wall climbing robot based on the three-dimensional tunnel cloud picture and the path planning rule; and calling a crack identification model to identify whether a crack to be repaired exists at the current position in the process that the wall-climbing robot travels according to the wall-climbing path.
Optionally, in some other embodiments of this embodiment, the grouting repair module 103 may further include a model training module, configured to train a crack identification model in advance based on a manner that information of each loss function is respectively reflected to a network to update a weight; the crack identification model comprises an input layer, a convolution layer, a plurality of dense blocks and an output layer; the output feature map of each dense block is input to the next dense block and output layer simultaneously.
The functions of the functional modules of the automatic tunnel crack repairing device according to the embodiments of the present invention may be specifically implemented according to the method in the embodiments of the method, and the specific implementation process may refer to the related description of the embodiments of the method, which is not described herein again.
Therefore, the embodiment of the invention can safely, efficiently and effectively repair the tunnel crack and effectively improve the stability and the safety of the tunnel structure.
The above-mentioned automatic tunnel crack repairing device is described from the perspective of a functional module, and further, the present application also provides an automatic tunnel crack repairing system, which is described from the perspective of hardware. Fig. 11 is a schematic structural diagram of an automatic tunnel crack repairing system provided in an embodiment of the present application in an implementation manner. As shown in fig. 11, the automatic tunnel crack repairing system may include an unmanned aerial vehicle 111 carrying a plurality of sensors for collecting image data and positioning data, a wall climbing robot 112 for deploying a vacuum perfusion apparatus, and an electronic device 113.
The unmanned aerial vehicle 111 autonomously plans a path in a dark environment, performs instant positioning and map construction, inputs a constructed three-dimensional tunnel cloud picture into the wall-climbing robot 112, the wall-climbing robot 112 autonomously moves to an area to be repaired based on position information of a diseased area, performs detailed identification on a crack by using a proposed crack identification model, combines an opportunity constraint method with an MU-RRT path planning mode, autonomously selects an optimal path, performs grouting repair along the crack, and can adopt a microcapsule material for grouting repair, so that secondary repair can be performed when the crack is re-cracked.
As shown in fig. 12, the electronic device 113 includes a memory 1130 for storing a computer program; a processor 1131, configured to execute the computer program to implement the steps of the automatic tunnel crack repairing method according to any of the embodiments.
The processor 1131 may include one or more processing cores, such as a 4-core processor, an 8-core processor, and the processor 1131 may also be a controller, a microcontroller, a microprocessor or other data processing chip, and the like. The processor 1131 may be implemented in at least one hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), and a PLA (Programmable Logic Array). The processor 1131 may also include a main processor and a coprocessor, where the main processor is a processor for Processing data in an awake state, and is also called a Central Processing Unit (CPU); a coprocessor is a low power processor for processing data in a standby state. In some embodiments, the processor 1131 may be integrated with a GPU (Graphics Processing Unit) that is responsible for rendering and drawing the content that the display screen needs to display. In some embodiments, the processor 1131 may further include an AI (Artificial Intelligence) processor for processing computing operations related to machine learning.
The memory 1130 may include one or more computer-readable storage media, which may be non-transitory. The memory 1130 may also include high-speed random access memory as well as non-volatile memory, such as one or more magnetic disk storage devices, flash memory storage devices. The memory 1130 may be, in some embodiments, an internal storage unit of the electronic device 113, such as a hard disk of a server. The memory 1130 may also be an external storage device of the electronic device 113 in other embodiments, such as a plug-in hard disk provided on a server, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like. Further, the memory 1130 may also include both internal storage units and external storage devices of the electronic device 113. The memory 1130 may be used for storing various data and application software installed in the electronic device 113, such as: the code of the program or the like in the course of executing the tunnel crack automatic repair method may also be used to temporarily store data that has been output or is to be output. In this embodiment, the memory 1130 is at least used for storing the following computer program 11301, wherein after being loaded and executed by the processor 1131, the computer program can implement the relevant steps of the automatic tunnel crack repairing method disclosed in any one of the foregoing embodiments. In addition, the resources stored in the memory 1130 may also include an operating system 11302 and data 11303, which may be stored in a transient or persistent manner. Operating system 11302 may include Windows, unix, linux, and the like. Data 11303 can include, but is not limited to, data corresponding to the results of automatic repair of tunnel fractures, and the like.
In some embodiments, the electronic device 113 may further include a display 1132, an input/output interface 1133, a communication interface 1134 or network interface, a power supply 1135, and a communication bus 1136. The display 1132 and the input/output interface 1133 such as a Keyboard (Keyboard) belong to a user interface, and the selectable user interface may further include a standard wired interface, a wireless interface, and the like. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, or the like. The display, which may also be referred to as a display screen or display unit, as appropriate, is used for displaying information processed in the electronic device 113 and for displaying a visualized user interface. The communication interface 1134 may optionally include a wired interface and/or a wireless interface, such as a WI-FI interface, a bluetooth interface, etc., typically used to establish a communication connection between the electronic device 113 and other electronic devices. The communication bus 1136 may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown in FIG. 12, but this is not intended to represent only one bus or type of bus.
Those skilled in the art will appreciate that the configuration shown in FIG. 12 is not intended to be limiting of the electronic device 113, and may include more or fewer components than those shown, such as a sensor 1137 that performs various functions.
The functions of the functional modules of the automatic tunnel crack repairing system according to the embodiment of the present invention may be specifically implemented according to the method in the embodiment of the method, and the specific implementation process may refer to the related description of the embodiment of the method, which is not described herein again.
Therefore, the embodiment of the invention can safely, efficiently and effectively repair the tunnel crack and effectively improve the stability and the safety of the tunnel structure.
It is understood that, if the automatic tunnel crack repairing method in the above embodiments is implemented in the form of a software functional unit and sold or used as a stand-alone product, it may be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, which are essential or part of the prior art, or all or part of the technical solutions may be embodied in the form of a software product, which is stored in a storage medium and executes all or part of the steps of the methods of the embodiments of the present application. And the aforementioned storage medium includes: a U disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), an electrically erasable programmable ROM, a register, a hard disk, a multimedia card, a card type Memory (e.g., SD or DX Memory, etc.), a magnetic Memory, a removable magnetic disk, a CD-ROM, a magnetic or optical disk, and other various media capable of storing program codes.
Based on this, an embodiment of the present invention further provides a readable storage medium, which stores a computer program, where the computer program is executed by a processor, and the steps of the method for automatically repairing a tunnel crack according to any one of the above embodiments are provided.
In the present specification, the embodiments are described in a progressive manner, and each embodiment focuses on differences from other embodiments, and the same or similar parts between the embodiments are referred to each other. For the hardware including the device and the automatic tunnel crack repairing system disclosed by the embodiment, the description is relatively simple because the method corresponds to the method disclosed by the embodiment, and the relevant points can be obtained by referring to the description of the method.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the technical solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The method, the device, the system and the readable storage medium for automatically repairing the tunnel crack provided by the application are described in detail above. The principles and embodiments of the present invention have been described herein using specific examples, which are presented only to assist in understanding the method and its core concepts of the present invention. It should be noted that, for those skilled in the art, it is possible to make various improvements and modifications to the present invention without departing from the principle of the present invention, and those improvements and modifications also fall within the scope of the claims of the present application.
Claims (9)
1. An automatic repairing method for a tunnel crack is characterized by comprising the following steps:
collecting multi-mode data to construct a three-dimensional tunnel map by using an unmanned aerial vehicle carrying a plurality of sensors for collecting image data and positioning data and based on an autonomous planning path in the flight process of a light-source-free GPS-free tunnel to be detected;
in the process of constructing the three-dimensional tunnel map, degradation features on the surface of the tunnel and in the tunnel are detected at the same time, and a three-dimensional tunnel cloud map for identifying position information of an area to be repaired is generated based on the three-dimensional tunnel map and the degradation features;
positioning each crack to be repaired of the tunnel to be repaired by using a wall climbing robot based on the three-dimensional tunnel cloud picture, and moving the crack to be repaired to a target crack to be repaired for negative pressure adsorption and perfusion;
wherein, the unmanned aerial vehicle that utilizes the sensor of carrying on the multiclass and gathering image data and location data gathers multimode data at the flight in-process that does not have the tunnel that awaits measuring of light source no GPS based on independently planning the route to construct three-dimensional tunnel map, include:
according to unscented Kalman filtering, generating a three-dimensional tunnel map and determining the position of the unmanned aerial vehicle in real time according to data processing results of laser scanning data of a laser radar, ranging data of an ultrasonic range finder and measuring data of an inertia measuring unit and image data acquired by infrared image acquisition equipment;
and based on the three-dimensional tunnel map and the measured value of the depth of the tunnel wall of the tunnel to be measured, automatically determining the flight path at the next moment by combining the unmanned aerial vehicle motion model according to the mode that the unmanned aerial vehicle moves towards the estimated three-dimensional point on the tunnel crankshaft and moves in the tunnel to be measured step by step.
2. The automatic repairing method for the tunnel crack according to claim 1, wherein the moving to the target crack to be repaired is performed with negative pressure adsorption pouring, and the method comprises the following steps:
mixing nano silicon dioxide and epoxy resin in advance according to a first preset proportion, or mixing maleic anhydride, graphene sheet material and epoxy resin according to a second preset proportion to generate composite epoxy resin; adding polyvinyl alcohol fiber or double-wall microcapsule material into the composite epoxy resin to modify the composite epoxy resin;
in the advancing process of each preset repairing path of the crack to be repaired according to the path, when the crack to be repaired moves to the current crack to be repaired, the wall climbing robot pours the modified composite epoxy resin into the current crack to be repaired under the negative pressure environment.
3. The method for automatically repairing a tunnel crack according to claim 1, wherein the step of generating a three-dimensional tunnel map and determining the position of the unmanned aerial vehicle in real time according to the data processing result of the laser scanning data of the laser radar, the ranging data of the ultrasonic range finder and the measurement data of the inertial measurement unit and the image data acquired by the infrared image acquisition device according to the unscented kalman filter comprises the steps of:
constructing a three-dimensional tunnel map according to the aligned laser scanning data;
determining the flight speed of the unmanned aerial vehicle according to the laser scanning data of the unmanned aerial vehicle at different moments;
determining absolute position information of the unmanned aerial vehicle in a flight area based on each laser scanning data and the latest available map state;
and fusing the flight speed of the unmanned aerial vehicle, the gravity adjustment acceleration of the inertial measurement unit and the absolute position information.
4. The automatic tunnel crack repairing method according to claim 3, wherein before the building of the three-dimensional tunnel map according to the aligned laser scanning data, the method further comprises:
establishing a corresponding relation between all points of laser scanning data at different moments;
and determining the conversion information of the laser scanning data aligned to the current point of the laser scanning data at the current moment through a minimized error function, and applying the conversion information to each point of the aligned laser scanning data until the alignment condition of the laser scanning points is met.
5. The method for automatically repairing the tunnel crack according to any one of claims 1 to 4, wherein the positioning of each crack to be repaired of the tunnel to be repaired based on the three-dimensional tunnel cloud map by using the wall-climbing robot comprises:
the method comprises the steps of representing by using particles in advance based on position uncertainty of the wall-climbing robot, simultaneously expanding uncertainty of each position by using RRT, representing uncertainty propagation in a tunnel environment constructed by various types of terrains by using the particles, and determining a path planning rule of the wall-climbing robot;
generating a wall-climbing path of the wall-climbing robot based on the three-dimensional tunnel cloud picture and the path planning rule;
and calling a crack identification model to identify whether a crack to be repaired exists at the current position in the process that the wall climbing robot travels according to the wall climbing path.
6. The automatic tunnel crack repairing method according to claim 5, wherein before the invoking of the crack recognition model identifies whether there is a crack to be repaired at the current position, the method further comprises:
training a crack recognition model in advance based on a mode that information of each loss function is respectively reflected to a network to update a weight;
the crack identification model comprises an input layer, a convolutional layer, a plurality of dense blocks and an output layer; the output feature map of each dense block is input to the next dense block and the output layer simultaneously.
7. An automatic tunnel crack repairing device is characterized by comprising:
the data acquisition module is used for acquiring multi-mode data based on an autonomous planning path in the flight process of the light source-free GPS-free tunnel to be detected by utilizing an unmanned aerial vehicle carrying a plurality of sensors for acquiring image data and positioning data so as to construct a three-dimensional tunnel map;
the initial positioning module is used for simultaneously detecting degradation characteristics of the surface of the tunnel and the interior of the tunnel in the three-dimensional tunnel map construction process, and generating a three-dimensional tunnel cloud map for identifying the position information of the area to be repaired based on the three-dimensional tunnel map and the degradation characteristics;
the grouting repair module is used for positioning each crack to be repaired of the tunnel to be repaired by using a wall-climbing robot based on the three-dimensional tunnel cloud chart, and moving the crack to be repaired to a target crack to be repaired for negative pressure adsorption and perfusion;
wherein the data acquisition module is further configured to: according to unscented Kalman filtering, generating a three-dimensional tunnel map and determining the position of the unmanned aerial vehicle in real time according to data processing results of laser scanning data of a laser radar, ranging data of an ultrasonic range finder and measuring data of an inertia measuring unit and image data acquired by infrared image acquisition equipment; and based on the three-dimensional tunnel map and the measured tunnel wall depth value of the tunnel to be measured, automatically determining the flight path at the next moment by combining an unmanned aerial vehicle motion model according to the mode that the unmanned aerial vehicle moves towards the estimated three-dimensional point on the tunnel crankshaft and moves in the tunnel to be measured step by step.
8. A readable storage medium, characterized in that the readable storage medium has stored thereon a computer program which, when being executed by a processor, realizes the steps of the automatic tunnel crack repairing method according to any one of claims 1 to 6.
9. An automatic tunnel crack repairing system is characterized by comprising an unmanned aerial vehicle carrying a plurality of sensors for collecting image data and positioning data, a wall-climbing robot for deploying a vacuum perfusion instrument and electronic equipment;
the electronic device comprises a processor and a memory, the processor being configured to implement the steps of the automatic tunnel crack repair method according to any one of claims 1 to 6 when executing the computer program stored in the memory.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211106955.6A CN115182747B (en) | 2022-09-13 | 2022-09-13 | Automatic tunnel crack repairing method, device and system and readable storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211106955.6A CN115182747B (en) | 2022-09-13 | 2022-09-13 | Automatic tunnel crack repairing method, device and system and readable storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115182747A CN115182747A (en) | 2022-10-14 |
CN115182747B true CN115182747B (en) | 2023-02-03 |
Family
ID=83524577
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211106955.6A Active CN115182747B (en) | 2022-09-13 | 2022-09-13 | Automatic tunnel crack repairing method, device and system and readable storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115182747B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116539635A (en) * | 2023-05-17 | 2023-08-04 | 广州大学 | Unmanned aerial vehicle shield tunnel inspection method based on calculation vision and SLAM |
CN118038284B (en) * | 2024-04-15 | 2024-06-14 | 贵州黔通工程技术有限公司 | Intelligent detection method and system for expressway diseases |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CA2354226A1 (en) * | 2001-01-31 | 2002-07-31 | Cal Holland | Robotic apparatus and method for non-destructive maintenance of intersecting conduits |
CN104843176A (en) * | 2015-04-28 | 2015-08-19 | 武汉大学 | Unmanned-gyroplane system used for automatic-inspection of bridges and tunnels and navigation method |
CN110763234A (en) * | 2019-10-15 | 2020-02-07 | 哈尔滨工程大学 | Submarine topography matching navigation path planning method for underwater robot |
CN113087869A (en) * | 2021-05-14 | 2021-07-09 | 重庆工程职业技术学院 | Coal mine grouting reinforcement material and preparation method thereof |
CN113358665A (en) * | 2021-05-25 | 2021-09-07 | 同济大学 | Unmanned aerial vehicle tunnel defect detection method and system |
CN114293481A (en) * | 2021-12-31 | 2022-04-08 | 山东省交通工程监理咨询有限公司 | Underwater overhaul robot, system and method for bridge pier |
-
2022
- 2022-09-13 CN CN202211106955.6A patent/CN115182747B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CA2354226A1 (en) * | 2001-01-31 | 2002-07-31 | Cal Holland | Robotic apparatus and method for non-destructive maintenance of intersecting conduits |
CN104843176A (en) * | 2015-04-28 | 2015-08-19 | 武汉大学 | Unmanned-gyroplane system used for automatic-inspection of bridges and tunnels and navigation method |
CN110763234A (en) * | 2019-10-15 | 2020-02-07 | 哈尔滨工程大学 | Submarine topography matching navigation path planning method for underwater robot |
CN113087869A (en) * | 2021-05-14 | 2021-07-09 | 重庆工程职业技术学院 | Coal mine grouting reinforcement material and preparation method thereof |
CN113358665A (en) * | 2021-05-25 | 2021-09-07 | 同济大学 | Unmanned aerial vehicle tunnel defect detection method and system |
CN114293481A (en) * | 2021-12-31 | 2022-04-08 | 山东省交通工程监理咨询有限公司 | Underwater overhaul robot, system and method for bridge pier |
Also Published As
Publication number | Publication date |
---|---|
CN115182747A (en) | 2022-10-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN115182747B (en) | Automatic tunnel crack repairing method, device and system and readable storage medium | |
CN112347840B (en) | Vision sensor laser radar integrated unmanned aerial vehicle positioning and image building device and method | |
US11530924B2 (en) | Apparatus and method for updating high definition map for autonomous driving | |
CN113945206A (en) | Positioning method and device based on multi-sensor fusion | |
KR101956447B1 (en) | Method and apparatus for position estimation of unmanned vehicle based on graph structure | |
CN104236548A (en) | Indoor autonomous navigation method for micro unmanned aerial vehicle | |
CN113587930B (en) | Indoor and outdoor navigation method and device of autonomous mobile robot based on multi-sensor fusion | |
CN110929402A (en) | Probabilistic terrain estimation method based on uncertain analysis | |
Cho et al. | Autonomous ship deck landing of a quadrotor UAV using feed-forward image-based visual servoing | |
Li et al. | Hybrid filtering framework based robust localization for industrial vehicles | |
CN112068152A (en) | Method and system for simultaneous 2D localization and 2D map creation using a 3D scanner | |
Lee et al. | Survey of robotics technologies for civil infrastructure inspection | |
Stumm et al. | Tensor-voting-based navigation for robotic inspection of 3D surfaces using lidar point clouds | |
CN113821044A (en) | Bridge detection unmanned aerial vehicle autonomous navigation and stability control method based on reinforcement learning | |
CN113096190A (en) | Omnidirectional mobile robot navigation method based on visual map building | |
KR102288609B1 (en) | Method and system for position estimation of unmanned aerial vehicle using graph structure based on multi module | |
Song et al. | BIM-aided scanning path planning for autonomous surveillance UAVs with LiDAR | |
Akremi et al. | Visual navigation of uavs in indoor corridor environments using deep learning | |
CN117234203A (en) | Multi-source mileage fusion SLAM downhole navigation method | |
Emter et al. | Stochastic cloning for robust fusion of multiple relative and absolute measurements | |
Wang et al. | Research on mobile robot localization and mapping method for underground long-narrow tunnels | |
Deusch et al. | Improving localization in digital maps with grid maps | |
Park et al. | Horizontal-vertical guidance of quadrotor for obstacle shape mapping | |
JP7396618B2 (en) | Mobile position estimation system | |
Zhou et al. | Localization for unmanned vehicle |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |