CN113420896B - Transformer substation inspection auxiliary method and system based on artificial intelligence and big data analysis - Google Patents
Transformer substation inspection auxiliary method and system based on artificial intelligence and big data analysis Download PDFInfo
- Publication number
- CN113420896B CN113420896B CN202110968904.3A CN202110968904A CN113420896B CN 113420896 B CN113420896 B CN 113420896B CN 202110968904 A CN202110968904 A CN 202110968904A CN 113420896 B CN113420896 B CN 113420896B
- Authority
- CN
- China
- Prior art keywords
- equipment
- inspection
- flow direction
- correlation coefficient
- power flow
- 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
- 238000007689 inspection Methods 0.000 title claims abstract description 207
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 11
- 238000007405 data analysis Methods 0.000 title claims abstract description 9
- 230000002159 abnormal effect Effects 0.000 claims abstract description 71
- 239000013598 vector Substances 0.000 claims abstract description 46
- 239000011159 matrix material Substances 0.000 claims abstract description 38
- 238000013528 artificial neural network Methods 0.000 claims abstract description 17
- 238000010586 diagram Methods 0.000 claims description 19
- 238000010606 normalization Methods 0.000 claims description 11
- 230000004931 aggregating effect Effects 0.000 claims description 9
- 238000010219 correlation analysis Methods 0.000 claims description 5
- 238000012423 maintenance Methods 0.000 abstract description 3
- 230000005856 abnormality Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000013024 troubleshooting Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C1/00—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people
- G07C1/20—Checking timed patrols, e.g. of watchman
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Primary Health Care (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a transformer substation inspection auxiliary method and system based on artificial intelligence and big data analysis. The method comprises the following steps: the power grid current direction of the polling is a target current direction, and the polling attention of the equipment is obtained according to the correlation coefficient of the equipment and the target current direction and the danger degree of the equipment; determining the secondary polling strength of the equipment according to the sum of covariance matrix elements of abnormal state vectors of the equipment which is polled for multiple times; overlapping historical routing inspection tracks of the transformer substation to obtain a routing inspection heat map; obtaining a first associated feature map and a second associated feature map according to the inspection attention, the secondary inspection strength and the position of the substation equipment; and inputting the first associated feature map and the second associated feature map into a neural network to generate an auxiliary routing inspection track. The invention can ensure that the abnormity of the equipment can be found in time during inspection, thereby rapidly detecting the reason of the abnormity of the substation equipment for maintenance and ensuring that the substation can safely and stably operate.
Description
Technical Field
The application relates to the technical field of big data and smart power grids, in particular to a transformer substation inspection auxiliary method and system based on artificial intelligence and analysis.
Background
The power transmitted by the power station needs to be distributed by the substation after being boosted or reduced by the substation. For a large-scale transformer substation, the equipment is numerous, the power system is complex, the power supply scale is large, and various faults and safety problems may exist in the transformer substation equipment. And with the development of large-scale and intelligent power grid configuration, the forms of complex operation environment, high operation technical difficulty and wide working face are gradually presented in the operation and maintenance and overhaul operation of the electric power, but in order to reduce the risk of the transformer substation and timely find and remove faults, the transformer substation equipment needs to be regularly patrolled and examined by human and intelligent robots and data of the transformer substation equipment is recorded.
However, in the prior art, the routing inspection track during routing inspection is low in reasonableness, the whole operation state and the historical fault data of the power grid are ignored, so that no emphasis is placed on routing inspection, and the routing inspection of the abnormal substation equipment is not performed preferentially, so that the working efficiency is reduced, and the safety detection and fault troubleshooting of the power grid are not facilitated.
Disclosure of Invention
Aiming at the problems, the invention provides a transformer substation inspection auxiliary method based on artificial intelligence and big data analysis, and the adopted technical scheme is as follows:
the power grid current direction of the current patrol is a target current direction, and a correlation coefficient between equipment and the target current direction is obtained according to the correlation between the equipment state of the substation and the power grid state in the historical patrol data of the target current direction; obtaining the inspection attention of the equipment according to the correlation coefficient of the equipment and the target tide direction and the danger degree of the equipment;
acquiring a covariance matrix of abnormal state vectors of the repeated inspection equipment from historical inspection data of the target power flow direction; determining the secondary inspection strength of the equipment according to the sum of the elements of the covariance matrix;
overlapping historical routing inspection tracks of the transformer substation to obtain a routing inspection heat map; generating a patrol attention heat degree diagram according to the position of the substation equipment and the patrol attention degree; generating a secondary inspection force heat degree diagram according to the position of the substation equipment and the secondary inspection force; and inputting the inspection heat map, the inspection attention heat map and the secondary inspection force heat map into the neural network to obtain an auxiliary inspection track.
Preferably, the method for inputting the inspection heat map, the inspection attention heat map and the secondary inspection force heat map into the neural network to obtain the auxiliary inspection track comprises the following steps: fusing the inspection heat map and the inspection attention heat map to obtain a first correlation characteristic map; fusing the secondary inspection force heat map and the inspection heat map to obtain a second correlation characteristic map; and inputting the first associated feature map and the second associated feature map into a neural network to generate an auxiliary routing inspection track.
Preferably, the incidence relation between the substation equipment state and the power grid state is obtained according to a correlation coefficient between the equipment and the power flow direction; the obtaining of the correlation coefficient of the equipment and the power flow direction comprises the following steps: aggregating the power grid state vectors in the same tidal current direction in the historical patrol to obtain a power grid state matrix in each tidal current direction; aggregating the abnormal state vectors of multiple patrolling of the same transformer substation equipment in the same tidal current direction to obtain an equipment abnormal state matrix in each tidal current direction; and carrying out correlation analysis on the power grid state matrix and the equipment abnormal state matrix in the same power flow direction to obtain a correlation coefficient between the equipment abnormal state and the power flow direction.
Preferably, the correlation coefficient between the substation equipment and the target power flow direction includes: obtaining a correlation coefficient of the equipment and a target power flow direction and a correlation coefficient ratio of the equipment and other power flow directions; obtaining a normalization coefficient according to the correlation coefficient between the equipment and the target power flow direction and the correlation coefficient between the equipment and each power flow direction; and obtaining a correlation coefficient between the equipment and the target power flow direction according to the correlation coefficient ratio and the normalization coefficient.
Preferably, the state of the grid comprises the voltage phasor and current phasor composition of the grid; the abnormal state vector of the equipment includes an abnormal degree of the sound of the equipment, an abnormal degree of the temperature of the equipment, and an abnormal degree of the amount of oil of the equipment.
The application also provides an auxiliary system is patrolled and examined to transformer substation based on artificial intelligence and big data analysis, and this system mainly includes: the inspection attention acquisition module is used for acquiring the association coefficient between equipment and a target power flow direction according to the association relation between the equipment state of the substation and the power grid state in historical inspection data of the target power flow direction, wherein the power grid power flow direction of the current inspection is the target power flow direction; obtaining the inspection attention of the equipment according to the correlation coefficient of the equipment and the target tide direction and the danger degree of the equipment;
the secondary inspection force acquisition module is used for acquiring a covariance matrix of abnormal state vectors of the multi-time inspection equipment from historical inspection data of the target tide direction; determining the secondary inspection strength of the equipment according to the sum of the elements of the covariance matrix;
the auxiliary inspection track acquisition module is used for superposing the historical inspection tracks of the transformer substation to obtain an inspection heat map; generating a patrol attention heat degree diagram according to the position of the substation equipment and the patrol attention degree; generating a secondary inspection force heat degree diagram according to the position of the substation equipment and the secondary inspection force; and inputting the inspection heat map, the inspection attention heat map and the secondary inspection force heat map into the neural network to obtain an auxiliary inspection track.
The technical scheme of the invention has the following beneficial effects: the method comprises the steps of obtaining the association coefficient of equipment and power grid power flow and the danger degree of the equipment by utilizing historical state data of substation equipment in historical inspection and the power flow direction of a power grid, and obtaining inspection attention and secondary inspection attention of the equipment, so that the equipment with high inspection attention and secondary inspection attention is preferentially inspected and repeatedly inspected when the substation equipment is inspected, the abnormity of the equipment can be timely found during inspection, the abnormal reason of the substation equipment is rapidly checked out for maintenance, and the substation can operate safely and stably.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
To further illustrate the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description will be given to a method and system for efficiently compressing video communication data based on artificial intelligence according to the present invention, with reference to the accompanying drawings and preferred embodiments, and the detailed description thereof. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
Example 1:
the embodiment provides a transformer substation inspection auxiliary method based on artificial intelligence and big data analysis, and a flow chart of the method is shown in fig. 1.
The specific scenario addressed by the present embodiment is as follows: and (5) routing inspection tracks of the substation equipment. The patrol inspection of this embodiment can be that the people is patrolled and examined and also can be that intelligent robot patrols and examines, patrols and examines personnel or patrol and examine the data that intelligent robot can accurate record every equipment in patrolling and examining of equipment, patrols and examines at every turn and all can maintain fault equipment after accomplishing and guarantee its normal operating.
Firstly, the power grid current direction of the current inspection is a target current direction, and a correlation coefficient between equipment and the target current direction is obtained according to the correlation between the equipment state of the substation and the power grid state in historical inspection data of the target current direction; and obtaining the inspection attention of the equipment according to the correlation coefficient of the equipment and the target power flow direction and the danger degree of the equipment.
Collecting historical data of the transformer substation and extracting characteristics: and acquiring the abnormal vector of each device in the corresponding tidal current direction and the state vector of the power grid according to the state data of the devices of the transformer substation during each inspection in history.
Obtain the spatial grid structure of transformer substation, equipment on the spatial grid of transformer substation includes: transformers, disconnectors, buses, circuit breakers, current transformers, and the like. The transformer substation has the functions of transformation, distribution, load transfer, steady state adjustment and the like, the power flow directions in the power grid of the transformer substation at different inspection moments are possibly different, and the power flow directions are used for representing which power transmission lines or power equipment pass through when power is consumed; in different power flow directions, the operating state of each device is different, the load of the power grid is different, the power environment of each device is different, the cause of the fault and the safety accident are different, for example, the noise of some devices is low in a certain power flow, and the noise of the same device is high in other power flows, which indicates that the operating state of the device is different in different power flows.
And acquiring the connection relation among the devices according to the grid structure. Acquiring historical inspection data, acquiring the flow direction of the transformer substation during the nth inspection, wherein each flow direction corresponds to a power dispatching path and is used for representing a path or a power transmission line through which power flows when the power is distributed from the transformer substation to a power utilization area; the method comprises the steps of acquiring equipment on a power dispatching path, and acquiring state data of the equipment, wherein the state data are routing inspection data, such as state data of sound, temperature and the like, and the state data can be acquired by sensors, and the sensors comprise a sound sensor, a temperature sensor and the like. Due to different equipment, the state data acquired by each equipment is different during each inspection, for example, the state data required to be acquired by the transformer comprises state data such as abnormal degree of sound, temperature, oil quantity and the like; the method for acquiring the abnormal degree of the sound is to input a sound sequence into a TCN network to acquire the abnormal degree of the sound, and the abnormal degree of the sound is divided into 10 levels: 0.1, 0.2, … …,1.0, the more abnormal degree of sound indicates more abnormal compared with normal sound, the TCN network belongs to the prior art; the oil quantity of the transformer is obtained by the liquid level meter, and the temperature of the transformer is obtained by the temperature sensor. Similarly, the state data such as the temperature and the abnormal sound degree of the equipment such as the isolating switch, the bus, the mutual inductor and the like can be obtained.
So far, some state data of each device can be obtained according to the routing inspection data, and the state data is regarded as an abnormal state vector of each device; and simultaneously obtaining voltage phasor (phasor refers to the amplitude and phase angle of voltage) and current phasor of all bus nodes of the power grid, wherein the phasors can describe the state of the power grid, and a vector formed by the phasors is called as a state vector of the power grid.
Obtaining the incidence relation between the substation equipment state and the power grid state according to the correlation coefficient between the equipment and the power flow direction; the obtaining of the correlation coefficient of the equipment and the power flow direction comprises the following steps: aggregating the power grid state vectors in the same tidal current direction in the historical patrol to obtain a power grid state matrix in each tidal current direction; aggregating the abnormal state vectors of multiple patrolling of the same transformer substation equipment in the same tidal current direction to obtain an equipment abnormal state matrix in each tidal current direction; and carrying out correlation analysis on the power grid state matrix and the equipment abnormal state matrix in the same power flow direction to obtain a correlation coefficient between the equipment and the power flow direction. Specifically, the power grid power flow direction and routing data during multiple routing inspections in the historical routing inspection process are obtained (one power flow may not be constant because the power flow of the power grid may not be constant)The direction corresponds to a state of the power grid, so that the current power grid current direction and routing inspection data are recorded during each routing inspection, and the current directions of a plurality of power grids can be obtained after multiple routing inspections); the power grid states with the same power flow direction are combined into one type, so that multiple types of power flow directions can be obtained, and one type of power flow direction corresponds to one power grid state. And setting the i-th class tide direction to contain P times of inspection data, aggregating the abnormal state vectors of each device in the P times of inspection into an abnormal state matrix, and aggregating the state vectors of the power grid in the P times of inspection into a power grid state matrix. Performing typical correlation analysis on the abnormal state matrix of the equipment and the power grid state matrix in the same power flow direction to obtain a correlation coefficient between each equipment and the power flow direction
(namely, through the ith type power flow data, one device corresponds to one abnormal state matrix, the abnormal state matrix and the power grid state matrix calculate one correlation coefficient, and the correlation coefficients can be obtained by a plurality of devices). In addition, the correlation coefficient of each state data of the equipment and each state data of the power grid can be analyzed respectively, and the mean value of all the correlation coefficients is the correlation coefficient of the equipment and the power flow direction required by the method. Let the correlation coefficient between the nth device and the ith class tide direction be
. When in use
(Times)
. The correlation coefficient is larger than 0, and the larger the correlation coefficient is, the abnormal state of the equipment and the power grid state have the consistent change rule, for example, when the power grid load is increased, the abnormality of the power station equipment is increased; when the correlation coefficient approaches to 0, it indicates that the abnormal state of the equipment is irrelevant to the power grid state, i.e. the abnormal degree of the equipment is not affected by the specific power grid state.
The correlation coefficient of the substation equipment and the target power flow direction comprises the following steps: obtaining a correlation coefficient of the equipment and a target power flow direction and a correlation coefficient ratio of the equipment and other power flow directions; obtaining a normalization coefficient according to the correlation coefficient between the equipment and the target power flow direction and the correlation coefficient between the equipment and each power flow direction; and obtaining a correlation coefficient between the equipment and the target power flow direction according to the correlation coefficient ratio and the normalization coefficient.
Specifically, calculating a correlation coefficient between the nth equipment and the ith type power flow direction
A represents the common class a tidal current direction. Wherein
The ratio of the correlation between the nth device and the ith flow direction to the correlation between the nth device and the flow directions of other types except the ith flow direction is shown, and the larger the value, the larger the correlation between the nth device and the ith flow direction is larger than that between the nth device and the flow directions of other types except the ith flow direction, and the correlation between the nth device and the flow directions of other types is small.
Is a normalized coefficient.
The relevance coefficient indicates whether the abnormality of the equipment is related to the power flow direction of the power grid. For example, when the power grid supplies power in a certain tidal current direction, abnormal noise occurs in the transformer when the power grid is overloaded or the power grid disturbance is larger, or abnormal heating of the bus is too large; when the power of the power grid is too large or the power grid disturbance is large when the power grid is switched to other power flow directions, the abnormality of the equipment can not be caused, and the abnormality of the equipment is related to one power flow direction.
The danger degree of the equipment is obtained according to the abnormal state vector of the equipment in the history inspection: acquiring any two abnormal state vectors in the abnormal state vectors of the equipment to obtain cosine similarity of the two abnormal state vectors; superposing the abnormal state vector by using a forgetting coefficient to obtain a first superposition result; and obtaining the danger degree of each substation device according to any two abnormal state vectors, the cosine similarity of any two abnormal state vectors and the first superposition result.
Specifically, the degree of danger is statistically obtained from historical patrol data, and a larger value indicates that the equipment is more susceptible to danger. The specific calculation method of the equipment risk degree in the embodiment comprises the following steps: assuming N rounds of inspection in the history record, the abnormal vector of each device obtained after the nth round of inspection is
Then the risk level of the nth device is:
because there are N rounds of inspection, then the nth device corresponds to N anomaly vectors, where
The result of superposition of forgetting coefficients is used for the N abnormal vectors (forgetting coefficient superposition method is well known, specifically, superposition method of forgetting coefficients is used
Wherein
，
，
The larger the indication that a larger anomaly is continuously generated, the greater the ultimate risk level).
The degree of correlation between the x-th and y-th vectors in the N abnormal vectors,
the specific calculation method comprises the following steps: obtaining any two anomaly vectors
、 
The more the similarity is, the more the abnormal states are consistent, and when the similar abnormal states are frequently appeared, the abnormal states mainly reflect the danger degree of the equipment, and the abnormal vector
、 
Has a cosine similarity of
。
A vector is represented which is a function of,
representing a vector
And vector
The result is still a vector.
The modulo length, which represents a vector, is a scalar quantity. Finally, the data is obtained according to the historical patrol data
The larger the number of devices, the higher the risk level of the nth device, and the more attention is paid to the nth device in inspection.
Obtaining the current direction of the power grid when the current inspection is started, and if the current direction of the power grid is the same as the ith type of power flow direction in the history, the inspection attention of the nth equipment at the current inspection is the power
. The greater the inspection attention, the more necessary the inspection is to be performed with priority inspection and emphasis inspection. It should be noted that, in order to improve the representation ability of the inspection attention,
normalization should be performed.
Then, acquiring a covariance matrix of abnormal state vectors of the repeated inspection equipment from historical inspection data of the target power flow direction; and determining the secondary polling strength of the equipment according to the sum of the elements of the covariance matrix.
Obtaining the secondary inspection force comprises: the method comprises the steps of obtaining the ith type of flow direction in a history record, wherein the type of flow direction comprises a plurality of times of routing inspection data, corresponding a plurality of abnormal vectors to the nth device during the plurality of times of routing inspection, calculating a covariance matrix of the abnormal vectors, and obtaining a value of an L1 norm of the matrix, wherein the larger the value is, the stronger the change amplitude of the state data of the nth device is, and the more difficult the state of the nth device is to be determined, and the value is called as the secondary routing inspection strength of the nth device during the current routing inspection. And setting the second polling force to be 0 when the second polling force is smaller than a certain threshold value. The larger the value is, the more necessary the equipment needs to be repeatedly inspected, and the equipment state of the equipment at different times in the inspection process is determined, so that the troubleshooting of the substation is facilitated. And obtaining the positions of all the devices and the corresponding inspection attention force and the secondary inspection force.
Finally, overlapping the historical inspection tracks of the transformer substation to obtain an inspection heat map; generating a patrol attention heat degree diagram according to the position of the substation equipment and the patrol attention degree; generating a secondary inspection force heat degree diagram according to the position of the substation equipment and the secondary inspection force; fusing the inspection heat map and the inspection attention heat map to obtain a first correlation characteristic map; fusing the secondary inspection force heat map and the inspection heat map to obtain a second correlation characteristic map; and inputting the first associated feature map and the second associated feature map into a neural network to generate an auxiliary routing inspection track.
The inspection track that the intelligent robot patrols and examines that obtains patrolling and examining personnel or patrolling and examining based on the dynamics of patrolling and examining of the dynamics of paying attention to patrolling and examining of every equipment and the secondary patrols and examines includes: acquiring an inspection heat map I3 according to the historical inspection track: establishing a panoramic top view of the transformer substation, wherein the construction method of the panoramic top view is conventional and well-known, and mapping a route of historical data inspected each time onto the panoramic top view, namely the pixel value of a passing position is 1 and the pixel value of a non-passing position is 0 during inspection to form a motion track which is called as an inspection track graph; each inspection in the historical data can obtain an inspection trace map; summing all routing inspection track maps obtained during routing inspection, and then carrying out normalization processing on images obtained after summation to obtain a routing inspection heat map I3, wherein positions with large pixel values indicate that the positions often pass through the positions during routing inspection, and positions with smaller pixel values indicate that the positions do not often pass through the positions, so that the positions cannot pass through or are dangerous to avoid being close to the positions often.
And acquiring an inspection attention intensity heat map I1 according to the inspection attention intensity, acquiring a secondary inspection intensity heat map I2 according to the secondary inspection intensity, and inputting I1, I2 and I3 into a neural network to acquire an auxiliary inspection track. The auxiliary inspection track can be a frame thermodynamic diagram, pixels in the diagram represent the inspection track, and can also be a group of coordinate sequences which can represent the inspection track. The training set of the neural network can be obtained through historical patrol data, and each patrol track is used as label data. In addition, in order to improve the generation accuracy of the auxiliary patrol trace, it is necessary to consider that the patrol trace is adjusted. In order to improve the robustness of the network, a large amount of training data can be generated through a computer simulator, and corresponding patrol trace label data can be generated. The loss function of the neural network adopts a cross entropy loss function.
Further, the present embodiment also provides an implementation method of a neural network, which obtains a feature map
And
to increase the convergence speed of the neural network: the method comprises the steps of mapping the position of a certain device on a panoramic top view, generating a Gaussian hotspot with the radius of D by taking the position of the device as a center, forming a thermodynamic diagram of the device, and multiplying the thermodynamic diagram by the patrol attention strength corresponding to the device to obtain the patrol attention strength thermodynamic diagram of the device. And acquiring inspection attention intensity heat maps of all the equipment, summing the inspection attention intensity heat maps, and then carrying out normalization processing to obtain an inspection attention intensity heat map I1 of all the equipment, and similarly obtaining a secondary inspection intensity heat map I2 of all the equipment. Respectively weighting and summing the inspection attention intensity heat map I1 and the inspection heat map I3 into a first associated feature map
，
(ii) a Weighting and summing the second inspection strength heat map I2 and the inspection heat map I3 to obtain a second associated feature map
，
。
Associating the first characteristic graph
And a second associated feature map
The method comprises the steps of inputting a DNN network, wherein the DNN network is of an Encoder-Decoder structure, and outputting a track binary diagram by the network to represent a patrol track. The method for acquiring the training data set of the DNN network comprises the following steps: because the input of the DNN network is two thermodynamic diagrams
And
the feature information of the two graphs is not complex, so that a large number of acquisition data sets can be randomly generated by calculation. The inventors have tagged data sets. Thus, an auxiliary inspection track is obtained, and inspection personnel can inspect according to the inspection track.
Example 2:
the present embodiment provides a system embodiment. The utility model provides an auxiliary system is patrolled and examined to transformer substation based on artificial intelligence and big data analysis, includes: the inspection attention acquisition module is used for acquiring the association coefficient between equipment and a target power flow direction according to the association relation between the equipment state of the substation and the power grid state in historical inspection data of the target power flow direction, wherein the power grid power flow direction of the current inspection is the target power flow direction; obtaining the inspection attention of the equipment according to the correlation coefficient of the equipment and the target tide direction and the danger degree of the equipment;
the secondary inspection force acquisition module is used for acquiring a covariance matrix of abnormal state vectors of the multi-time inspection equipment from historical inspection data of the target tide direction; determining the secondary inspection strength of the equipment according to the sum of the elements of the covariance matrix;
the auxiliary inspection track acquisition module is used for superposing the historical inspection tracks of the transformer substation to obtain an inspection heat map; generating a patrol attention heat degree diagram according to the position of the substation equipment and the patrol attention degree; generating a secondary inspection force heat degree diagram according to the position of the substation equipment and the secondary inspection force; and inputting the inspection heat map, the inspection attention heat map and the secondary inspection force heat map into the neural network to obtain an auxiliary inspection track.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the spirit and principles of the present invention, and any modifications, equivalents, improvements and the like that are made therein are intended to be included within the scope of the present invention.
Claims (10)
1. A transformer substation inspection auxiliary method based on artificial intelligence and big data analysis is characterized by comprising the following steps:
the power grid current direction of the current patrol is a target current direction, and a correlation coefficient between equipment and the target current direction is obtained according to the correlation between the equipment state of the substation and the power grid state in the historical patrol data of the target current direction; obtaining the inspection attention of the equipment according to the correlation coefficient of the equipment and the target tide direction and the danger degree of the equipment;
acquiring a covariance matrix of abnormal state vectors of the repeated inspection equipment from historical inspection data of the target power flow direction; determining the secondary inspection strength of the equipment according to the sum of the elements of the covariance matrix;
overlapping historical routing inspection tracks of the transformer substation to obtain a routing inspection heat map; generating a patrol attention heat degree diagram according to the position of the substation equipment and the patrol attention degree; generating a secondary inspection force heat degree diagram according to the position of the substation equipment and the secondary inspection force; and inputting the inspection heat map, the inspection attention heat map and the secondary inspection force heat map into the neural network to obtain an auxiliary inspection track.
2. The method of claim 1, wherein inputting the inspection heat map, the inspection attention heat map, and the secondary inspection force heat map into a neural network to obtain the auxiliary inspection trajectory comprises: fusing the inspection heat map and the inspection attention heat map to obtain a first correlation characteristic map; fusing the secondary inspection force heat map and the inspection heat map to obtain a second correlation characteristic map; and inputting the first associated feature map and the second associated feature map into a neural network to generate an auxiliary routing inspection track.
3. The method according to claim 1, wherein the association relationship between the substation equipment state and the grid state is obtained according to a correlation coefficient between equipment and a power flow direction; the obtaining of the correlation coefficient of the equipment and the power flow direction comprises the following steps: aggregating the power grid state vectors in the same tidal current direction in the historical patrol to obtain a power grid state matrix in each tidal current direction; aggregating the abnormal state vectors of multiple patrolling of the same transformer substation equipment in the same tidal current direction to obtain an equipment abnormal state matrix in each tidal current direction; and carrying out correlation analysis on the power grid state matrix and the equipment abnormal state matrix in the same power flow direction to obtain a correlation coefficient between the equipment abnormal state and the power flow direction.
4. The method of claim 1, wherein the correlation coefficient of the substation equipment to the target current direction comprises: obtaining a correlation coefficient of the equipment and a target power flow direction and a correlation coefficient ratio of the equipment and other power flow directions; obtaining a normalization coefficient according to the correlation coefficient between the equipment and the target power flow direction and the correlation coefficient between the equipment and each power flow direction; and obtaining a correlation coefficient between the equipment and the target power flow direction according to the correlation coefficient ratio and the normalization coefficient.
5. The method of claim 2, wherein the state of the grid comprises voltage phasors and current phasors of the grid; the abnormal state vector of the equipment includes an abnormal degree of the sound of the equipment, an abnormal degree of the temperature of the equipment, and an abnormal degree of the amount of oil of the equipment.
6. The utility model provides a transformer substation patrols and examines auxiliary system based on artificial intelligence and big data analysis which characterized in that, the system includes: the inspection attention acquisition module is used for acquiring the association coefficient between equipment and a target power flow direction according to the association relation between the equipment state of the substation and the power grid state in historical inspection data of the target power flow direction, wherein the power grid power flow direction of the current inspection is the target power flow direction; obtaining the inspection attention of the equipment according to the correlation coefficient of the equipment and the target tide direction and the danger degree of the equipment;
the secondary inspection force acquisition module is used for acquiring a covariance matrix of abnormal state vectors of the multi-time inspection equipment from historical inspection data of the target tide direction; determining the secondary inspection strength of the equipment according to the sum of the elements of the covariance matrix;
the auxiliary inspection track acquisition module is used for superposing the historical inspection tracks of the transformer substation to obtain an inspection heat map; generating a patrol attention heat degree diagram according to the position of the substation equipment and the patrol attention degree; generating a secondary inspection force heat degree diagram according to the position of the substation equipment and the secondary inspection force; and inputting the inspection heat map, the inspection attention heat map and the secondary inspection force heat map into the neural network to obtain an auxiliary inspection track.
7. The system according to claim 6, wherein the auxiliary inspection track acquisition module is further configured to fuse the inspection heat map and the inspection attention heat map to obtain a first associated feature map; fusing the secondary inspection force heat map and the inspection heat map to obtain a second correlation characteristic map; and inputting the first associated feature map and the second associated feature map into a neural network to generate an auxiliary routing inspection track.
8. The system of claim 6, wherein the inspection attention obtaining module is further configured to aggregate power grid state vectors in the same tidal current direction in the historical inspection to obtain a power grid state matrix in each tidal current direction; aggregating the abnormal state vectors of multiple patrolling of the same transformer substation equipment in the same tidal current direction to obtain an equipment abnormal state matrix in each tidal current direction; carrying out correlation analysis on the power grid state matrix and the equipment abnormal state matrix in the same power flow direction to obtain a correlation coefficient between the equipment abnormal state and the power flow direction; and obtaining the incidence relation between the substation equipment state and the power grid state according to the correlation coefficient between the equipment and the power flow direction.
9. The system of claim 6, wherein the patrol inspection attention acquisition module is further configured to acquire a correlation coefficient of the device with a target current direction and a ratio of correlation coefficients of the device with other current directions; obtaining a normalization coefficient according to the correlation coefficient between the equipment and the target power flow direction and the correlation coefficient between the equipment and each power flow direction; and obtaining a correlation coefficient between the equipment and the target power flow direction according to the correlation coefficient ratio and the normalization coefficient.
10. The system of claim 6, wherein the state of the grid comprises voltage phasors and current phasors of the grid; the abnormal state vector of the equipment includes an abnormal degree of the sound of the equipment, an abnormal degree of the temperature of the equipment, and an abnormal degree of the amount of oil of the equipment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110968904.3A CN113420896B (en) | 2021-08-23 | 2021-08-23 | Transformer substation inspection auxiliary method and system based on artificial intelligence and big data analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110968904.3A CN113420896B (en) | 2021-08-23 | 2021-08-23 | Transformer substation inspection auxiliary method and system based on artificial intelligence and big data analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113420896A CN113420896A (en) | 2021-09-21 |
| CN113420896B true CN113420896B (en) | 2021-11-16 |
Family
ID=77719287
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110968904.3A Active CN113420896B (en) | 2021-08-23 | 2021-08-23 | Transformer substation inspection auxiliary method and system based on artificial intelligence and big data analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113420896B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115221851B (en) * | 2022-06-23 | 2023-08-29 | 智瞻科技股份有限公司 | Analysis processing method and analysis processing system for operation and maintenance inspection form data of electric power station |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106597136A (en) * | 2015-10-14 | 2017-04-26 | 山东鲁能智能技术有限公司 | Intelligent route inspection method for abnormal equipment based on transformer route inspection robot |
| CN110533771A (en) * | 2019-08-21 | 2019-12-03 | 广西电网有限责任公司电力科学研究院 | A kind of intelligent polling method of substation |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8718323B2 (en) * | 2010-10-19 | 2014-05-06 | Raytheon Company | Batch detection association for enhanced target descrimination in dense detection environments |
-
2021
- 2021-08-23 CN CN202110968904.3A patent/CN113420896B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106597136A (en) * | 2015-10-14 | 2017-04-26 | 山东鲁能智能技术有限公司 | Intelligent route inspection method for abnormal equipment based on transformer route inspection robot |
| CN110533771A (en) * | 2019-08-21 | 2019-12-03 | 广西电网有限责任公司电力科学研究院 | A kind of intelligent polling method of substation |
Non-Patent Citations (3)
| Title |
|---|
| Identification of Electrical Equipment Based on Faster LSTM-CNN Network;Xiaoping Xiong 等;《IEEE》;20201030;全文 * |
| 图像识别技术在调度二次设备巡检中的应用;王诤 等;《电气设计》;20181231(第9期);全文 * |
| 基于大数据架构的变电站设备智能化巡检***设计;褚大可 等;《制造业自动化》;20180831;第40卷(第8期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113420896A (en) | 2021-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hu et al. | Real-time transient stability assessment in power system based on improved SVM | |
| CN109086518A (en) | A kind of method of intelligent substation power transmission and transformation primary equipment status assessment | |
| CN113420896B (en) | Transformer substation inspection auxiliary method and system based on artificial intelligence and big data analysis | |
| CN116359652A (en) | State monitoring system for power equipment | |
| Zhu et al. | Integrated data-driven power system transient stability monitoring and enhancement | |
| Xie et al. | Massively digitized power grid: opportunities and challenges of use-inspired AI | |
| Jafarzadeh et al. | A CNN-based post-contingency transient stability prediction using transfer learning | |
| CN117612345A (en) | Power equipment state monitoring and alarming system and method | |
| Zhu et al. | Robust representation learning for power system short-term voltage stability assessment under diverse data loss conditions | |
| CN113536607B (en) | Transformer substation signal transmission system evaluation method and system | |
| CN113344434B (en) | Transformer substation site selection method and system based on artificial intelligence and big data | |
| Haghifam et al. | State estimation in electric distribution networks in presence of distributed generation using the PMUs | |
| Moradzadeh et al. | Image processing-based data integrity attack detection in dynamic line rating forecasting applications | |
| Guan et al. | Grid monitoring and market risk management | |
| CN111127251A (en) | Attack identification method based on LSTM neural network and grid-connected interface device | |
| Myrda | Optimizing Assets [In My View] | |
| Zhao et al. | Application of deep neural networks for fault diagnosis in a hybrid AC/DC power grid | |
| Vinogradov et al. | Structure of a System for Monitoring Operation Modes of Electrical Network and Consumers | |
| Zhao et al. | A Method of Complementing Missing Power Data in Low-Voltage Stations Based on Improved Deep Convolutional Self-Encoding Network | |
| Liu et al. | Power grid fault diagnosis method based on alarm information and PMU fusion | |
| Saifuddin et al. | Apprehending fault crises for an autogenous nanogrid system: Sustainable buildings | |
| CN114825418B (en) | Multifunctional flexible complementary digital power supply management method | |
| CN117852773B (en) | Fault positioning system for power distribution network | |
| Ma et al. | Design of Intelligent Information Monitoring System for Distribution Network and Adjustment of Alarm Threshold. | |
| Hao et al. | Fault Diagnosis Using Deep Convolutional Neural Network with Dynamic Computer-visualised Power Flow |
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 |