CN109634308A - Based on intelligent navigation method under dynamic (dynamical) rate pattern auxiliary water - Google Patents
Based on intelligent navigation method under dynamic (dynamical) rate pattern auxiliary water Download PDFInfo
- Publication number
- CN109634308A CN109634308A CN201910038062.4A CN201910038062A CN109634308A CN 109634308 A CN109634308 A CN 109634308A CN 201910038062 A CN201910038062 A CN 201910038062A CN 109634308 A CN109634308 A CN 109634308A
- Authority
- CN
- China
- Prior art keywords
- speed
- rudder
- model
- angle
- dvl
- 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.)
- Granted
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 41
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000012549 training Methods 0.000 claims abstract description 22
- 239000013598 vector Substances 0.000 claims description 17
- 239000011159 matrix material Substances 0.000 claims description 15
- 210000002569 neuron Anatomy 0.000 claims description 8
- 230000005484 gravity Effects 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 238000013016 damping Methods 0.000 claims description 5
- 230000001133 acceleration Effects 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 239000005871 repellent Substances 0.000 claims description 2
- 101000881206 Homo sapiens Spermine synthase Proteins 0.000 claims 2
- 102100037616 Spermine synthase Human genes 0.000 claims 2
- 230000008569 process Effects 0.000 abstract description 7
- 238000011161 development Methods 0.000 description 6
- 206010033799 Paralysis Diseases 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 241000282414 Homo sapiens Species 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000251468 Actinopterygii Species 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000013138 pruning Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000008093 supporting effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/10—Simultaneous control of position or course in three dimensions
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
- G05B13/027—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks only
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B9/00—Safety arrangements
- G05B9/02—Safety arrangements electric
- G05B9/03—Safety arrangements electric with multiple-channel loop, i.e. redundant control systems
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/0055—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots with safety arrangements
- G05D1/0077—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots with safety arrangements using redundant signals or controls
Landscapes
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Aviation & Aerospace Engineering (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Aerodynamic Tests, Hydrodynamic Tests, Wind Tunnels, And Water Tanks (AREA)
Abstract
The present invention discloses one kind based on intelligent navigation method under dynamic (dynamical) rate pattern auxiliary water, comprising steps of (1) periodically acquisition airborne sensor information；(2) input variable and output variable of rate pattern to be established are determined；(3) it constructs rate pattern and it is trained；(4) when DVL is detected failure or data failure, the output speed based on institute's training pattern, which is used as, carries out navigation solution, and then implementation model assisting navigation to low speed degree.This programme uses for reference kinetic model thought, it is proposed a kind of new AUV rate pattern, and consider the influence of the factors such as rudder piece rudder angle, course angle, in operational process, as long as the fault data of DVL is not detected, training set can all be increased at any time and carry out model training, available model output speed is substituted after detecting DVL fault data, navigation error increases caused by effectively avoiding the problem that because of DVL failure or failure, under the premise of not increasing hardware cost, the redundancy approach of velocity sensor is provided, system robustness is good, and can ensure high-precision navigation.
Description
Technical Field
The invention relates to the field of AUV intelligent navigation, in particular to a speed model-assisted underwater intelligent navigation method based on dynamics.
Background
The ocean is the second big space behind the relay land of the four tactical spaces developed by human beings, is a strategic development base of energy, biological resources, metal resources and water resources, and has a great supporting effect on the development of the economic society. The extensive and deep exploration and development of oceans become one of the development subjects of the 21 st century, the nation also refers to the unprecedented strategic heights of 'concerning oceans, knowing oceans and slightly oceans', plans such as 'bivalve and one sea', 'transparent oceans' and 'healthy oceans' are deeply developed, and a '21 st century silk-on-sea' ocean environment safety guarantee system is constructed.
As an important assistant for exploring and developing oceans by human beings, Autonomous Underwater Vehicles (AUVs) play no better role in ocean development than artificial satellites in space exploration, and the AUVs with high-performance underwater operation capability become comprehensive manifestation of national ocean competitiveness. The AUV is an important means in civil fields of marine science research, resource investigation, emergency search and rescue and the like, is regarded as a multiplier of modern naval force, can finish predetermined tasks such as marine science research, resource investigation, emergency search and rescue and the like without leaving underwater navigation technology, and the accuracy of navigation positioning information is a bottleneck problem for determining the development and application of the AUV.
The precise positioning of the AUV is very important for the AUV to execute a task, and for a standard AUV which is only provided with sensors such as an Attitude and Heading Reference System (AHRS), a Doppler Velocimeter (DVL), a depth gauge (INS), and the like, only a single navigation-related sensor is provided, and a redundant sensor is absent, so that when a single sensor fails, especially an easily interfered DVL, random or sudden noise, a fish school, or a ditch suddenly meeting over a detection range, no speed is output, and isolating or using a value at the last stable time causes a great deviation to the navigation positioning, and even leads to the paralysis of the AUV navigation System.
Disclosure of Invention
The invention provides a dynamic-based speed model auxiliary underwater intelligent navigation method, which aims at the defect that when a single sensor fails, particularly when a DVL which is easy to be interfered has data failure, even when the sensor has no speed output due to unexpected failure, the isolation or the use of a value at the last stable moment causes great deviation on navigation positioning, possibly causes error increase and even system paralysis of an AUV navigation system.
The invention is realized by adopting the following technical scheme: a dynamics-based speed model assisted underwater intelligent navigation method comprises the following steps:
a, AUV, launching a sailing ship, and periodically collecting information of an onboard sensor, wherein the information of the onboard sensor comprises speed information of DVL, acceleration, angular velocity and angle information of AHRS, rotating speed information of a propeller and rudder angle information of a rudder sheet;
b, determining an input variable and an output variable of a speed model to be established based on the sensor data acquired in the step A:
the input variables comprise rudder angle of a rudder, rudder angle of a vertical rudder, rotating speed of a propeller, roll angle, pitch angle and course angle, and the output variables are the base speed of the DVL;
step C, according to the determined input variable and output variable of the speed model, constructing the speed model and training the speed model, and continuously enriching a training set when fault data are not detected by the DVL;
and D, when the DVL is detected to be in fault or the data is invalid, performing navigation analysis on the base speed based on the speed output by the model trained in the step C, and further realizing model-assisted navigation.
Further, in the step B, when determining the input variable, the following method is specifically adopted:
step B1, based on the concept of the dynamic model, preliminarily determining a rudder angle of a rudder, a rudder angle of a vertical rudder, a rotating speed of a propeller, a roll angle and a pitch angle as input variables according to the influence of variables in the dynamic model on the speed of the water-repellent flow:
because the dynamic model does not consider the influence of ocean current factors, the output is the water velocity, a three-degree-of-freedom dynamic model is established, and the motion equation is as follows:
wherein V is [ u, V, r ═ V]The generalized velocity is the generalized velocity of the AUV under an airborne coordinate system, u and v respectively represent the forward velocity and the right velocity of the AUV, and r represents the angular velocity in the direction around the normal line; tau isRB＝[X,Y,N]Representing external forces and external moments, wherein X, Y represents forward and right forces, respectively, through the origin of the AUV coordinate system, N represents the moment of the heading angle, MRBAnd CRBThe inertia matrix and the rigid coriolis centripetal force matrix are respectively expressed as follows:
wherein m is the weight of AUV in water, IZFor the rigid body moment of inertia generated around the AUV normal vector (z axis), the force and moment vectors of the above equation of motion are solved, firstly, the vector of motion is decomposed into the following five influence terms, and a specific expression of each term is obtained respectively:
wherein the formula of the additional mass inertia term is expressed as follows:
wherein,are all hydrodynamic coefficients;
the equation for the damping term is expressed as follows:
wherein, Xvr、Xvv、Xrr、X|u|u、Y|v|v、Y|r|r、Yuv、Yur、Nuv、Nur、N|v|v、N|r|rAre also hydrodynamic coefficients;
the formula for the restoring force and moment is as follows:
the restoring force and its moment are mainly generated by gravity and buoyancy, wherein W is gravity, B is buoyancy, theta andpitch and roll angles, X, respectivelyb、Yb、NbAre all fixed deviations;
the equations for the effects of the upper and lower rudders and the left and right vertical rudder are shown below:
wherein, Xδδuu、Yδuu、NδuuAre all hydrodynamic coefficients；δrt、δrbIs the deflection angle delta e of the upper and lower rudder bladesp、δesYaw angles for port and starboard;
the equation for the force and moment generated by the propeller is as follows:
wherein, X|n|n、X|n|uThe hydrodynamic coefficient is solved, n represents the rotating speed of the propeller, w represents the wake fraction, and the hydrodynamic coefficient is introduced into a rudder angle of a rudder, a rudder angle of a vertical rudder, a pitch angle, a roll angle and a rotating speed variable of the propeller to obtain the water velocity according to the formula, namely the input variables of the used velocity model are preliminarily determined to comprise the rudder angle of the rudder, the rudder angle of the vertical rudder, the pitch angle, the course angle and the rotating speed of the propeller;
and step B2, considering that the DVL equipped by the standard AUV only has the function of outputting the base speed, the base speed is obtained by summing the water flow speed and the water speed, considering that the speed measured by the DVL is the base speed and the water flow speed is unchanged in direction in a short time, on the basis of the step B1, increasing the influence variables of the water flow speed part, including the course angle, the rudder angle of the direction rudder and the angle of the vertical rudder, and adding the course angle as an input variable.
Further, in the step C, the step of constructing the speed model specifically includes the following steps:
step C1, SLFN construction by ELM:
constructing a training data set X { (X)i,yi|xi∈Rn,yi∈Rm,i＝1,2,…,M)}，xiAnd yiRespectively representing N-dimensional input vectors and m-dimensional output vectors, and constructing an SLFN network with N hidden layer nodes by adopting an ELM algorithm, wherein the mathematical expression of an output function is as follows:
β thereiniIs a parameter, ω, connecting the ith hidden layer node and the output nodeiAnd biThe node parameters of the hidden layer are simplified into the following vector form by the formula (1):
Hβ＝Y
wherein
H is the output matrix of the hidden layer node of the neural network, the ith column HiIs the output of the ith hidden layer node to the input, randomly initializing the parameter omega of the hidden layer nodeiAnd biWherein i ═ 1,2, …, N; calculating an output matrix H of the hidden layer node; according toSolving for parameters β, whereinMoore-Penrose transpose of H;
step C2, carrying out importance sequencing based on MRSR:
regression matrix H based on multiple response sparse regressionTEach row of H is added into the speed model one by one, and the importance ordering of the hidden layer nodes is realized;
step C3, determining the number of hidden layer neurons based on LOO, and further obtaining an optimized velocity model:
according to the importance ranking provided by MRSR, cut off partial non-unions using leave-one-outImportant hidden layer nodes, i.e. based on formulasEvaluating LOO error and neuron quantity, and finally determining the number of the reserved hidden layer neurons;
step C4, continuously enriching a training set of the speed model, and taking the average value output by the multiple training models as the final output speed; in the AUV launching process, as long as fault data are not detected by the DVL, training data sets can be enriched continuously, after an optimal speed model is obtained through training, an ideal current speed can be obtained by inputting corresponding variable values, and for the accuracy of output speed, the average value of 10 times of model output can be used as the finally obtained speed.
Further, in the step D, when the AUV navigates on the water surface, navigation and positioning are performed according to GPS, DVL, AHRS, and IPS data, where the GPS is used as a correction amount; and when the AUV navigates underwater, calculating the position of the AUV based on a Kalman filtering or particle filtering algorithm or an optimization algorithm based on graph optimization according to the numerical values of DVL, AHRS and IPS, and after detecting the DVL data fault, continuing navigation and positioning by using the speed output by the speed model based on dynamics.
Compared with the prior art, the invention has the advantages and positive effects that:
the invention provides a new AUV speed model by referring to the idea of a dynamic model, the decomposition of the water flow speed in different directions in the model is automatically added to the speed output, and the influence of factors such as rudder angle and course angle of a rudder piece is considered; in the operation process, as long as the fault data of the DVL are not detected, a training set can be added at any time for model training, the model output speed can be used for replacing after the fault data of the DVL are detected, on the premise of not increasing the hardware cost, a redundant method of a speed sensor is provided, effective speed output is provided when the DVL is detected to have a fault, high-precision navigation is kept, and the method is mainly used for a standard AUV which only comprises one set of navigation system and sails in a shallow water area;
the parameters of the speed model are convenient to modify, the calculation is simpler than that of the traditional hydrodynamic vehicle model, the replacement of an AUV counterweight, a rudder sheet or a propeller does not influence the model, and the corresponding speed model can be obtained only by carrying out AUV operation in water at different rotating speeds, different rudder angles and different angles before the beginning of a formal test; the problem of navigation error increase caused by DVL failure or fault can be effectively avoided, and the robustness of the navigation system is enhanced.
Drawings
FIG. 1 is a schematic diagram of a navigation method according to an embodiment of the present invention;
FIG. 2 is a comparison of a DVL velocity generated trajectory and a velocity fitted trajectory generated by the method of the present invention.
Detailed Description
In order to make the above objects, features and advantages of the present invention more clearly understood, the present invention will be further described with reference to the accompanying drawings and examples. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
A speed model assisted underwater intelligent navigation method based on dynamics is shown in a schematic diagram of fig. 1 and comprises the following steps:
the method comprises the steps that firstly, AUV (autonomous underwater vehicle) sails in water, and onboard sensor information is periodically collected, wherein the sensor information comprises the speed information of DVL (dynamic velocity logging), the acceleration, the angular velocity and the angle information of AHRS (attitude heading reference system), the rotating speed information of a propeller and the rudder angle information of a rudder blade;
and step two, by taking the idea of the dynamic model into consideration, determining input variables of the water velocity part in the velocity model according to the influence of variables in the dynamic model on the water velocity, wherein the input variables comprise rudder angles of a rudder, a vertical rudder angle, the rotating speed of a propeller, a roll angle and a pitch angle, considering that the measured velocity of the DVL is the bottom velocity and the velocity of the water flow is constant and the direction is constant in a short time, increasing the influence variables of the water velocity part, including a course angle, a rudder angle of the rudder and a vertical rudder angle, so that the course angle is added as the input variables, and the output variable is the bottom velocity of the DVL, and the specific analysis is as follows:
in the embodiment, considering that the speed and direction of the water flow are not changed in a short time, a rudder angle, a course angle and the like are added as input variables of a speed model, the counter-bottom speed of the DVL is used as an output variable, and specifically, firstly, the influence formulas of the weight, the damping terms and the like of a rudder, a vertical rudder, a propeller and an AUV in the fluid on the speed are listed; because the dynamic model does not consider the influence of factors such as ocean current, the output is the velocity of the water, taking the three-degree-of-freedom dynamic model as an example, the motion equation is as follows:
wherein V is [ u, V, r ═ V]The generalized velocity is the generalized velocity of the AUV under an airborne coordinate system, u and v respectively represent the forward velocity and the right velocity of the AUV, and r represents the angular velocity in the direction around the normal line; tau isRB＝[X,Y,N]Representing external forces and external moments, wherein X, Y represents forward and right forces, respectively, through the origin of the AUV coordinate system, N represents the moment of the heading angle, MRBAnd CRBRespectively, an inertia matrix and a rigid-body Coriolis centripetal force matrix, as shown below
Where m is the weight of AUV in the liquid (sea or lake water), IZFor the moment of inertia of the rigid body generated around the normal vector (z-axis) of the AUV, the force and moment vectors of the above equation of motion are solved in such a way that they are decomposed into additional massesInertia, damping, restoring force and moment, control rudder and screw five aspects specifically as follows:
wherein the formula of the additional mass inertia term is expressed as follows:
wherein,are linear hydrodynamic coefficients representing the force applied to the AUV per unit velocity/acceleration in the direction of interest. The equation for the damping term is expressed as follows:
wherein, Xvr、Xvv、Xrr、X|u|u、Y|v|v、Y|r|r、Yuv、Yur、Nuv、Nur、N|v|v、N|r|rAre coupled hydrodynamic coefficients, which represent the coupled hydrodynamic coefficients at X, Y, N caused by the velocities u, v, and r, respectively, as will be explained below. The formula for the restoring force and moment is as follows:
the restoring force and its moment are mainly generated by gravity and buoyancy, wherein W is gravity, B is buoyancy, theta andpitch and roll angles, X, respectivelyb、Yb、NbAre all fixed deviations.
The equations for the effects of the upper and lower rudders and the left and right vertical rudder are shown below:
wherein, Xδδuu、Yδuu、NδuuAre also coupling hydrodynamic coefficients; delta rt、δrbIs the deflection angle delta e of the upper and lower rudder bladesp、δesPort and starboard yaw angles.
The equation for the force and moment generated by the propeller is as follows:
wherein, X|n|n、X|n|uCoupling water power coefficient is also adopted; n represents the rotating speed of the propeller, w represents the wake fraction, and the hydrodynamic coefficient is solved according to the formula, and then substituted into the variables such as the rudder angle of the rudder, the rudder angle of the vertical rudder, the pitch angle, the roll angle, the rotating speed of the propeller, and the like to obtain the water velocity, namely the rudder angle of the rudder, the rudder angle of the vertical rudder, the pitch angle, the course angle and the rotating speed of the propeller are preliminarily determined to be used as input variables of a velocity model.
Considering the cost problem of a standard AUV, the equipped DVL only has the function of outputting the bottom speed, which is one of the reasons that the output can only select the bottom speed, the bottom speed is obtained by summing the water flow speed and the water speed, the influence variable of the water flow speed part needs to be considered, in the process of using a dynamic model for assisting navigation, the water flow speed is generally assumed to be unchanged in size and direction in a short time, the water flow speed is solved and then is directly decomposed to the absolute speed of the AUV according to a course angle, and the influence of the change of the rudder angle on the AUV is not considered. The invention adds the same assumed water flow speed into the speed model, adds the influence of variables such as the rudder, the vertical rudder, the course angle and the like on the water flow speed component, ensures that the model is more accurate, and continuously adds the rudder angle of the rudder, the vertical rudder, the rudder angle, the course angle and other factors influencing the water flow speed component in each direction of the AUV.
In conclusion, variables such as a rudder, a vertical rudder, a propeller rotating speed, a course angle, a roll angle and a pitch angle are selected as input variables, the counter-bottom speed of the DVL is selected as an output variable, the decomposition of the water flow speed in the model in different directions of the AUV is automatically added to the speed output, and the influence of factors such as a rudder angle and a course angle of a rudder piece on the water flow speed is considered, so that the speed model of the scheme is more complete.
Step three, according to the input variable and the output variable of the speed model obtained in the step two, constructing the speed model and training the speed model, wherein in the specific implementation, a training set can be enriched continuously in the AUV advancing process as long as the DVL does not detect the fault, and the specific construction process of the speed model is as follows:
(1) according to the obtained input variables and output variables, a training data set X { (xi, yi | xi ∈ R is constructedn,yi∈Rm,i＝1,2,…,M)}，xiAnd yiRespectively representing N-dimensional input vectors and m-dimensional output vectors, and constructing an SLFN network with N hidden layer nodes by adopting an ELM algorithm, wherein the mathematical expression of an output function is as follows:
β thereiniIs a parameter, ω, connecting the ith hidden layer node and the output nodeiAnd biThe above formula can be simplified into the following vector form:
Hβ＝Y
wherein
H is the output matrix of the hidden layer node of the neural network, the ith column HiIs the output of the ith hidden layer node to the input. Randomly initializing parameter omega of hidden layer nodeiAnd biIs, wherein i ═ 1,2, …, N; calculating an output matrix H of the hidden layer node; according toSolving for parameters β, whereinMoore-Penrose transpose of H.
(2) For the problems of ELM model with relevant or irrelevant variables in the training dataset, a Multiple Response Sparse Regression (MRSR) is used to combine the Regression matrix HTAnd adding each column of H into the speed model one by one to realize the importance sequencing of the hidden layer nodes.
(3) According to the importance ranking provided by MRSR, a Leave-One-Out (LOO) method is used for pruning off partial unimportant hidden layer nodes, and the LOO error and the neuron number are mainly evaluated by the following formula, so that the number of the reserved hidden layer neurons is finally determined.
Training data sets can be enriched in the AUV launching process, after an optimal speed model is obtained through training, an ideal current speed can be obtained by inputting a corresponding variable value, and for the accuracy of output speed, the average value of multiple (for example, 10) model outputs can be used as the final obtained base speed.
Step four, when the DVL is detected to be in fault or the data is invalid, the output speed of the model trained in the step three is used as the base speed for navigation calculation, and on the premise of not increasing the hardware cost, a sensor substitution/redundancy scheme is provided to improve the navigation precision;
when the AUV navigates on the water surface, navigation positioning is carried out according to data such as GPS, DVL, AHRS, IPS and the like, wherein the GPS is used as a correction value; when the AUV navigates underwater, because the GPS signal is attenuated quickly underwater, the position data can not be measured, and the position of the AUV is calculated mainly according to the numerical values of DVL, AHRS and IPS by means of a Filter algorithm based on Kalman Filter (KF), Particle Filter (PF) and the like or an optimization algorithm based on graph optimization.
After DVL faults are detected through state residual error detection, KF characteristic detection methods and the like, the traditional methods for isolating fault sources or reducing information dependence have certain posterior, and can cause navigation system errors to increase and even system paralysis. As a result, in fig. 2, the solid line indicates the position trajectory estimated by combining with other sensor data when there is no DVL failure data, and the dotted line indicates the position trajectory estimated by combining the velocity obtained by the trained model with other sensor data.
The above description is only a preferred embodiment of the present invention, and not intended to limit the present invention in other forms, and any person skilled in the art may apply the above modifications or changes to the equivalent embodiments with equivalent changes, without departing from the technical spirit of the present invention, and any simple modification, equivalent change and change made to the above embodiments according to the technical spirit of the present invention still belong to the protection scope of the technical spirit of the present invention.
Claims (4)
1. The method for assisting underwater intelligent navigation by using the speed model based on dynamics is characterized by comprising the following steps of:
a, AUV, launching a sailing ship, and periodically collecting information of an onboard sensor, wherein the information of the onboard sensor comprises speed information of DVL, acceleration, angular velocity and angle information of AHRS, rotating speed information of a propeller and rudder angle information of a rudder sheet;
b, determining an input variable and an output variable of a speed model to be established based on the sensor data acquired in the step A:
the input variables comprise rudder angle of a rudder, rudder angle of a vertical rudder, rotating speed of a propeller, roll angle, pitch angle and course angle, and the output variables are the base speed of the DVL;
step C, according to the determined input variable and output variable of the speed model, constructing the speed model and training the speed model, and continuously enriching a training set when fault data are not detected by the DVL;
and D, when the DVL is detected to be in fault or the data is invalid, performing navigation analysis on the base speed based on the speed output by the model trained in the step C, and further realizing model-assisted navigation.
2. The dynamics-based velocity model assisted underwater intelligent navigation method of claim 1, wherein: in the step B, when determining the input variable, the following method is specifically used:
step B1, based on the concept of the dynamic model, preliminarily determining a rudder angle of a rudder, a rudder angle of a vertical rudder, a rotating speed of a propeller, a roll angle and a pitch angle as input variables according to the influence of variables in the dynamic model on the speed of the water-repellent flow:
because the dynamic model does not consider the influence of ocean current factors, the output is the water velocity, a three-degree-of-freedom dynamic model is established, and the motion equation is as follows:
wherein V is [ u, V, r ═ V]The generalized velocity is the generalized velocity of the AUV under an airborne coordinate system, u and v respectively represent the forward velocity and the right velocity of the AUV, and r represents the angular velocity in the direction around the normal line; tau isRB＝[X,Y,N]Representing external forces and external moments, wherein X, Y represents forward and right forces, respectively, through the origin of the AUV coordinate system, N represents the moment of the heading angle, MRBAnd CRBThe inertia matrix and the rigid coriolis centripetal force matrix are respectively expressed as follows:
wherein m is the weight of AUV in water, IZFor the rigid body moment of inertia generated around the AUV normal vector (z axis), the force and moment vectors of the above equation of motion are solved, firstly, the vector of motion is decomposed into the following five influence terms, and a specific expression of each term is obtained respectively:
wherein the formula of the additional mass inertia term is expressed as follows:
wherein,are all hydrodynamic coefficients;
the equation for the damping term is expressed as follows:
wherein, Xvr、Xvv、Xrr、X|u|u、Y|v|v、Y|r|r、Yuv、Yur、Nuv、Nur、N|v|v、N|r|rAre also hydrodynamic coefficients;
the formula for the restoring force and moment is as follows:
the restoring force and its moment are mainly generated by gravity and buoyancy, wherein W is gravity, B is buoyancy, theta andpitch and roll angles, X, respectivelyb、Yb、NbAre all fixed deviations;
the equations for the effects of the upper and lower rudders and the left and right vertical rudder are shown below:
wherein, Xδδuu、Yδuu、NδuuAre all hydrodynamic coefficients; delta rt、δrbIs the deflection angle delta e of the upper and lower rudder bladesp、δesYaw angles for port and starboard;
the equation for the force and moment generated by the propeller is as follows:
wherein, X|n|n、X|n|uThe hydrodynamic coefficient is solved, n represents the rotating speed of the propeller, w represents the wake fraction, and the hydrodynamic coefficient is introduced into a rudder angle of a rudder, a rudder angle of a vertical rudder, a pitch angle, a roll angle and a rotating speed variable of the propeller to obtain the water velocity according to the formula, namely the input variables of the used velocity model are preliminarily determined to comprise the rudder angle of the rudder, the rudder angle of the vertical rudder, the pitch angle, the course angle and the rotating speed of the propeller;
and step B2, considering that the speed measured by the DVL is the bottom speed and the water flow speed and the direction are not changed in a short time, considering the influence variables of the water flow speed part including the course angle, the rudder angle of the direction rudder and the rudder angle of the vertical rudder on the basis of the step B1, and adding the course angle as an input variable.
3. The dynamics-based velocity model assisted underwater intelligent navigation method of claim 1, wherein: in the step C, the speed model is constructed specifically by the following steps:
step C1, SLFN construction by ELM:
constructing a training data set X { (X)i,yi|xi∈Rn,yi∈Rm,i＝1,2,…,M)}，xiAnd yiRespectively representing N-dimensional input vectors and m-dimensional output vectors, and constructing an SLFN network with N hidden layer nodes by adopting an ELM algorithm, wherein the mathematical expression of an output function is as follows:
β thereiniIs a parameter, ω, connecting the ith hidden layer node and the output nodeiAnd biThe node parameters of the hidden layer are simplified into the following vector form by the formula (1):
Hβ＝Y
wherein
H is the output matrix of the hidden layer node of the neural network, the ith column HiIs the output of the ith hidden layer node to the input, randomly initializing the parameter omega of the hidden layer nodeiAnd biWherein i ═ 1,2, …, N; calculating an output matrix H of the hidden layer node; according toSolving for parameters β, whereinMoore-Penrose transpose of H;
step C2, carrying out importance sequencing based on MRSR:
regression matrix H based on multiple response sparse regressionTEach column of H is added to the velocity model one by one,realizing the importance sequencing of the nodes of the hidden layer;
step C3, determining the number of hidden layer neurons based on LOO, and further obtaining an optimized velocity model:
according to the importance ranking provided by MRSR, a leave-one-out method is used to prune off the partially unimportant hidden layer nodes, namely based on a formulaEvaluating LOO error and neuron quantity, and finally determining the number of the reserved hidden layer neurons;
and C4, continuously enriching the training set of the speed model when the DVL does not detect the fault data, and taking the average value of multiple model outputs as the final output speed after the fault data appears.
4. The dynamics-based velocity model assisted underwater intelligent navigation method of claim 1, wherein: in the step D, when the AUV navigates on the water surface, navigation positioning is carried out according to GPS, DVL, AHRS and IPS data, wherein the GPS is used as a correction value; and when the AUV navigates underwater, calculating the position of the AUV based on a Kalman filtering or particle filtering algorithm or an optimization algorithm based on graph optimization according to the numerical values of DVL, AHRS and IPS, and after detecting the DVL data fault, continuing navigation and positioning by using the speed output by the speed model based on dynamics.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910038062.4A CN109634308B (en) | 2019-01-16 | 2019-01-16 | Speed model assisted underwater intelligent navigation method based on dynamics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910038062.4A CN109634308B (en) | 2019-01-16 | 2019-01-16 | Speed model assisted underwater intelligent navigation method based on dynamics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109634308A true CN109634308A (en) | 2019-04-16 |
| CN109634308B CN109634308B (en) | 2020-07-07 |
Family
ID=66060959
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910038062.4A Active CN109634308B (en) | 2019-01-16 | 2019-01-16 | Speed model assisted underwater intelligent navigation method based on dynamics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109634308B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110209197A (en) * | 2019-06-25 | 2019-09-06 | 湖北航天技术研究院总体设计所 | A kind of flight control system design method |
| CN110554359A (en) * | 2019-09-11 | 2019-12-10 | 哈尔滨工程大学 | seabed flight node positioning method integrating long baseline positioning and single beacon positioning |
| CN110906933A (en) * | 2019-11-06 | 2020-03-24 | 中国海洋大学 | AUV (autonomous Underwater vehicle) auxiliary navigation method based on deep neural network |
| CN111026145A (en) * | 2019-12-24 | 2020-04-17 | 上海海事大学 | Ups and downs and gesture control system suitable for upper ocean robot |
| CN111091159A (en) * | 2019-12-27 | 2020-05-01 | 哈尔滨工程大学 | Weak fault feature extraction method for autonomous underwater robot propeller |
| CN111258326A (en) * | 2020-02-27 | 2020-06-09 | 中国科学院自动化研究所 | Bionic robot fish roll stability control method based on reaction wheel |
| CN111811536A (en) * | 2020-05-29 | 2020-10-23 | 中国船舶重工集团公司第七0七研究所 | Inertial navigation position error estimation method based on gravity and base velocity information |
| CN112818819A (en) * | 2021-01-28 | 2021-05-18 | 青岛澎湃海洋探索技术有限公司 | AUV state monitoring method based on dynamic model and complex network theory |
| CN113934159A (en) * | 2021-10-25 | 2022-01-14 | 中国船舶工业综合技术经济研究院 | Unmanned ship reliability test environment model construction method |
| CN115576184A (en) * | 2022-08-27 | 2023-01-06 | 华中科技大学 | Fault online diagnosis and fault-tolerant control method for underwater robot propeller |
| CN117250970A (en) * | 2023-11-13 | 2023-12-19 | 青岛澎湃海洋探索技术有限公司 | Method for realizing AUV fault detection based on model embedding generation countermeasure network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103576693A (en) * | 2013-11-11 | 2014-02-12 | 哈尔滨工程大学 | Underwater robot three-dimensional path tracking control method based on second-order filter |
| CN103942383A (en) * | 2014-04-17 | 2014-07-23 | 哈尔滨工程大学 | Dynamics and kinematics estimation method for deep sea operation type ROV |
| CN104881038A (en) * | 2015-04-22 | 2015-09-02 | 哈尔滨工业大学 | Unmanned underwater vehicle (UUV) track tracking control optimization method under environmental interference |
| CN108663939A (en) * | 2018-06-12 | 2018-10-16 | 哈尔滨工程大学 | Consider the UUV level of constellation face path tracking control method of communication packet loss |
-
2019
- 2019-01-16 CN CN201910038062.4A patent/CN109634308B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103576693A (en) * | 2013-11-11 | 2014-02-12 | 哈尔滨工程大学 | Underwater robot three-dimensional path tracking control method based on second-order filter |
| CN103942383A (en) * | 2014-04-17 | 2014-07-23 | 哈尔滨工程大学 | Dynamics and kinematics estimation method for deep sea operation type ROV |
| CN104881038A (en) * | 2015-04-22 | 2015-09-02 | 哈尔滨工业大学 | Unmanned underwater vehicle (UUV) track tracking control optimization method under environmental interference |
| CN108663939A (en) * | 2018-06-12 | 2018-10-16 | 哈尔滨工程大学 | Consider the UUV level of constellation face path tracking control method of communication packet loss |
Non-Patent Citations (3)
| Title |
|---|
| 彭树萍: "AUV动力学模型辅助的航位推算方法研究", 《万方学位论文库》 * |
| 谢连妮: "基于水下航行器的SINS_DVL组合导航技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 韩波: "基于集成学习的极限学习机的算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110209197A (en) * | 2019-06-25 | 2019-09-06 | 湖北航天技术研究院总体设计所 | A kind of flight control system design method |
| CN110554359B (en) * | 2019-09-11 | 2021-07-09 | 哈尔滨工程大学 | Seabed flight node positioning method integrating long baseline positioning and single beacon positioning |
| CN110554359A (en) * | 2019-09-11 | 2019-12-10 | 哈尔滨工程大学 | seabed flight node positioning method integrating long baseline positioning and single beacon positioning |
| CN110906933A (en) * | 2019-11-06 | 2020-03-24 | 中国海洋大学 | AUV (autonomous Underwater vehicle) auxiliary navigation method based on deep neural network |
| CN111026145A (en) * | 2019-12-24 | 2020-04-17 | 上海海事大学 | Ups and downs and gesture control system suitable for upper ocean robot |
| CN111026145B (en) * | 2019-12-24 | 2023-06-23 | 上海海事大学 | Sinking and floating and attitude control system suitable for upper ocean robot |
| CN111091159A (en) * | 2019-12-27 | 2020-05-01 | 哈尔滨工程大学 | Weak fault feature extraction method for autonomous underwater robot propeller |
| CN111091159B (en) * | 2019-12-27 | 2023-03-17 | 哈尔滨工程大学 | Weak fault feature extraction method for autonomous underwater robot propeller |
| CN111258326A (en) * | 2020-02-27 | 2020-06-09 | 中国科学院自动化研究所 | Bionic robot fish roll stability control method based on reaction wheel |
| CN111811536A (en) * | 2020-05-29 | 2020-10-23 | 中国船舶重工集团公司第七0七研究所 | Inertial navigation position error estimation method based on gravity and base velocity information |
| CN111811536B (en) * | 2020-05-29 | 2022-09-16 | 中国船舶重工集团公司第七0七研究所 | Inertial navigation position error estimation method based on gravity and base velocity information |
| CN112818819A (en) * | 2021-01-28 | 2021-05-18 | 青岛澎湃海洋探索技术有限公司 | AUV state monitoring method based on dynamic model and complex network theory |
| CN113934159A (en) * | 2021-10-25 | 2022-01-14 | 中国船舶工业综合技术经济研究院 | Unmanned ship reliability test environment model construction method |
| CN115576184A (en) * | 2022-08-27 | 2023-01-06 | 华中科技大学 | Fault online diagnosis and fault-tolerant control method for underwater robot propeller |
| CN117250970A (en) * | 2023-11-13 | 2023-12-19 | 青岛澎湃海洋探索技术有限公司 | Method for realizing AUV fault detection based on model embedding generation countermeasure network |
| CN117250970B (en) * | 2023-11-13 | 2024-02-02 | 青岛澎湃海洋探索技术有限公司 | Method for realizing AUV fault detection based on model embedding generation countermeasure network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109634308B (en) | 2020-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109634308B (en) | Speed model assisted underwater intelligent navigation method based on dynamics | |
| Skjetne et al. | A nonlinear ship manoeuvering model: Identification and adaptive control with experiments for a model ship | |
| Roberts et al. | Advances in unmanned marine vehicles | |
| CN111736617B (en) | Track tracking control method for preset performance of benthonic underwater robot based on speed observer | |
| Kim et al. | Estimation of hydrodynamic coefficients for an AUV using nonlinear observers | |
| Skjetne et al. | Modeling, identification, and adaptive maneuvering of Cybership II: A complete design with experiments | |
| Karmozdi et al. | INS-DVL navigation improvement using rotational motion dynamic model of AUV | |
| Yoshimura | Mathematical model for manoeuvring ship motion (MMG Model) | |
| Lv et al. | Underwater navigation methodology based on intelligent velocity model for standard AUV | |
| Randeni P et al. | Parameter identification of a nonlinear model: replicating the motion response of an autonomous underwater vehicle for dynamic environments | |
| CN109579850B (en) | Deepwater intelligent navigation method based on auxiliary inertial navigation to water velocity | |
| Mukhopadhyay et al. | Collaborative autonomous surveys in marine environments affected by oil spills | |
| Racine et al. | CFD-based method for simulation of marine-vehicle maneuvering | |
| Tomera | Dynamic positioning system for a ship on harbour manoeuvring with different observers. Experimental results | |
| Tran et al. | Least squares optimisation algorithm based system identification of an autonomous underwater vehicle | |
| CN107315348B (en) | Position and posture control method of unmanned underwater vehicle based on punishment type wavelet network | |
| Yan et al. | A navigation accuracy compensation algorithm for low-cost unmanned surface vehicles based on models and event triggers | |
| Ko et al. | Fault detection and diagnosis of sensors and actuators for unmanned surface vehicles | |
| Siregar et al. | Design and construction of hybrid autonomous underwater glider for underwater research | |
| CN113934159A (en) | Unmanned ship reliability test environment model construction method | |
| Nelson et al. | Hydrodynamic force decoupling using a distributed sensory system | |
| CN108459614B (en) | UUV real-time collision avoidance planning method based on CW-RNN network | |
| Rasal | Navigation & control of an automated swath surface vessel for bathymetric mapping | |
| Skoglund et al. | Modeling and sensor fusion of a remotely operated underwater vehicle | |
| Fenco et al. | Model Reference Adaptive Fuzzy Controller of a 6-DOF Autonomous Underwater 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 |