Method and system for power grid multi-unmanned aerial vehicle inspection task distribution

Abstract

The invention discloses a power grid multi-unmanned aerial vehicle inspection task distribution method and system, and belongs to the technical field of power grid operation and maintenance. The method comprises the following steps: determining region characteristics of a target inspection region, dividing the target inspection region into a plurality of regions according to the region characteristics, and establishing an unmanned aerial vehicle inspection task distribution model for a to-be-inspected power grid tower in each of the plurality of regions; determining an area covered by the machine nest in each area, and planning a point distribution position of the machine nest in each area based on the area covered by the machine nest; and establishing a task target of a power grid multi-unmanned aerial vehicle inspection task, based on the unmanned aerial vehicle inspection task distribution model, according to the distribution position and the task target of the aircraft nest, generating a distribution decision of the inspection tasks of the multiple unmanned aerial vehicles in the aircraft nest, and based on the distribution decision, distributing the inspection tasks to the multiple unmanned aerial vehicles in the aircraft nest. According to the invention, the efficiency of unmanned aerial vehicle inspection is improved, and the inspection cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid operation and maintenance, and more specifically, to a method and system for allocating inspection tasks of multiple unmanned aerial vehicles (UAVs) in a power grid. Background Art

[0002] In terms of power grid drone inspection, the current main method is to use a work team to carry a drone to inspect several adjacent base towers and their equipment. Each equipment operation and maintenance management unit also prepares inspection plans based on this strategy. Under the current new situation of rapid growth in the scale of power grids, problems such as structural shortages of operation and maintenance professionals are becoming increasingly prominent. This strategy can no longer fully meet the requirements of improving the quality and efficiency of equipment operation and inspection and ensuring the safe and stable operation of the power grid.

[0003] In recent years, some domestic units have begun to explore the possibility that during a patrol mission, the operation team can carry multiple drones and carry out operations simultaneously according to pre-planned patrol routes to improve on-site work efficiency. Specific practices include: for a single-base tower, multiple drones are used to shoot key parts from different angles to reduce the drone's flight time around the tower and the number of threading flights; for multiple towers within the drone's operating radius, multiple drones are used to conduct patrol operations on different towers at the same time.

[0004] The above inspection strategy requires that the UAV equipment has strong mission planning functions, high-precision navigation and positioning, and program-controlled flight performance. The mission planning function plans the operation routes for each UAV participating in the operation, and formulates the action information that should be associated with all the control waypoints on the route; the RTK high-precision navigation and positioning module guides the UAV to accurately fly along the planned route and accurately reach each control waypoint, and automatically executes each action strictly according to the preset.

[0005] Drone inspection has become one of the main means of power grid equipment inspection. However, in the existing technology, due to the limitation of drone flight time, one inspection takes about 20 to 40 minutes, and the inspection area is limited. Therefore, in the future, power grid equipment inspection will develop in the direction of multiple drones working in coordination. The existing collaborative operation mode generally divides the inspection tasks manually by operators according to the principle of proximity, without quantitatively considering factors such as the direction of the line and the influence of the take-off and landing points. The benefits of drone collaboration are not obvious, and there is still room for improvement in operation efficiency. Summary of the invention

[0006] In view of the above problems, the present invention proposes a method for allocating inspection tasks of multiple UAVs in a power grid, comprising:

[0007] Determine the regional characteristics of the target inspection area, and divide the target inspection area into multiple areas according to the regional characteristics, and establish a drone inspection task allocation model for the power grid towers to be inspected in each of the multiple areas;

[0008] Determine the area covered by the machine nest in each area, and plan the location of the machine nest in each area based on the area covered by the machine nest;

[0009] Establish the task objectives of the multi-UAV inspection tasks of the power grid, generate the allocation decision of the inspection tasks of multiple UAVs in the machine nest based on the UAV inspection task allocation model and the location and task objectives of the machine nest, and allocate inspection tasks to multiple UAVs in the machine nest based on the allocation decision.

[0010] Optionally, determining regional characteristics of a target inspection area, and dividing the target inspection area into a plurality of areas according to the regional characteristics, including:

[0011] Determine the regional characteristics of the target inspection area, and based on the regional characteristics, construct a fuzzy relationship matrix describing the attributes of each grid tower to be inspected within the inspection operation range, and use the transitive closure method to divide the target inspection area into multiple areas.

[0012] Optionally, the target inspection area is divided into multiple areas, including:

[0013] Setting available index vectors of the power grid towers to be inspected, constructing an inspection factor hierarchical analysis matrix based on the available index vectors, and performing normalization processing on the inspection factor hierarchical analysis matrix;

[0014] The similarity between any two elements in the normalized inspection factor hierarchy analysis matrix is ​​calculated using the correlation coefficient method;

[0015] The similarities are dynamically clustered using a closure method to obtain dynamic clustering results, and based on the dynamic clustering results, the target inspection area is divided into multiple areas.

[0016] Optionally, the similarities are dynamically clustered using a closure method, including:

[0017] Calculate the similarity distance between any two elements, and calculate the transitive closure of the similarity distance, sort the elements of the transitive closure, and find the corresponding truncation matrix based on the sorting result;

[0018] The elements in the inspection factor hierarchical analysis matrix are classified according to the cut matrix.

[0019] Optionally, the method further includes: optimizing the location of the machine nest, including:

[0020] Establish a collection of power grid towers to be inspected, a collection of candidate locations, and a collection of machine nests within the scope of operation;

[0021] Establish mapping relationships among the set of grid towers to be inspected, the set of candidate locations and the set of machine nests within the operation range;

[0022] Based on the mapping relationship, a mathematical model is established, and based on the mathematical model, the layout positions of the machine nests are optimized.

[0023] Optionally, the mapping relationship includes: towers inspected by drones, candidate locations that have been used, and already deployed drone nests.

[0024] Optionally, the task goal is to complete all the planned tasks in the shortest time;

[0025] The constraints of the task objectives include:

[0026] Different task weights are assigned to each grid tower to be inspected within the scope of the operation task. In each area, grid towers to be inspected with high operation weights are given priority, and the higher the task weight, the longer the assigned operation time.

[0027] On the other hand, the present invention also proposes a system for allocating inspection tasks of multiple UAVs in a power grid, comprising:

[0028] A division area unit is used to determine the regional characteristics of the target inspection area, and divide the target inspection area into multiple areas according to the regional characteristics, and establish a drone inspection task allocation model for the power grid towers to be inspected in each of the multiple areas;

[0029] The machine nest distribution unit is used to determine the area covered by the machine nest in each area, and plan the distribution position of the machine nest in each area based on the area covered by the machine nest;

[0030] The task allocation unit is used to establish the task objectives of the multi-UAV inspection tasks of the power grid, generate allocation decisions for the inspection tasks of multiple UAVs in the machine nest based on the UAV inspection task allocation model and according to the location and task objectives of the machine nest, and allocate inspection tasks to multiple UAVs in the machine nest based on the allocation decisions.

[0031] Optionally, determining regional characteristics of a target inspection area, and dividing the target inspection area into a plurality of areas according to the regional characteristics, including:

[0032] Determine the regional characteristics of the target inspection area, and based on the regional characteristics, construct a fuzzy relationship matrix describing the attributes of each grid tower to be inspected within the inspection operation range, and use the transitive closure method to divide the target inspection area into multiple areas.

[0033] Optionally, the target inspection area is divided into multiple areas, including:

[0034] Setting available index vectors of the power grid towers to be inspected, constructing an inspection factor hierarchical analysis matrix based on the available index vectors, and performing normalization processing on the inspection factor hierarchical analysis matrix;

[0035] The similarity between any two elements in the normalized inspection factor hierarchy analysis matrix is ​​calculated using the correlation coefficient method;

[0036] The similarities are dynamically clustered using a closure method to obtain dynamic clustering results, and based on the dynamic clustering results, the target inspection area is divided into multiple areas.

[0037] Optionally, the similarities are dynamically clustered using a closure method, including:

[0038] Calculate the similarity distance between any two elements, and calculate the transitive closure of the similarity distance, sort the elements of the transitive closure, and find the corresponding truncation matrix based on the sorting result;

[0039] The elements in the inspection factor hierarchical analysis matrix are classified according to the cut matrix.

[0040] Optionally, the point distribution unit is also used to optimize the point distribution position of the machine nest, including:

[0041] Establish a collection of power grid towers to be inspected, a collection of candidate locations, and a collection of machine nests within the scope of operation;

[0042] Establish mapping relationships among the set of grid towers to be inspected, the set of candidate locations and the set of machine nests within the operation range;

[0043] Based on the mapping relationship, a mathematical model is established, and based on the mathematical model, the layout positions of the machine nests are optimized.

[0044] Optionally, the mapping relationship includes: towers inspected by drones, candidate locations that have been used, and already deployed drone nests.

[0045] Optionally, the task goal is to complete all the planned tasks in the shortest time;

[0046] The constraints of the task objectives include:

[0047] Different task weights are assigned to each grid tower to be inspected within the scope of the operation task. In each area, grid towers to be inspected with high operation weights are given priority, and the higher the task weight, the longer the assigned operation time.

[0048] In yet another aspect, the present invention further provides a computing device, comprising: one or more processors;

[0049] a processor for executing one or more programs;

[0050] When the one or more programs are executed by the one or more processors, the above-described method is implemented.

[0051] In yet another aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed, the method described above is implemented.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] The present invention provides a method for allocating inspection tasks of multiple unmanned aerial vehicles (UAVs) in a power grid, comprising: determining the regional characteristics of a target inspection area, and dividing the target inspection area into multiple areas according to the regional characteristics, and establishing an unmanned aerial vehicle inspection task allocation model for the power grid pole towers to be inspected in each area of ​​the multiple areas; determining the area covered by the machine nest in each area, and planning the location of the machine nest in each area based on the area covered by the machine nest; establishing the task objectives of the inspection tasks of multiple unmanned aerial vehicles in the power grid, and based on the unmanned aerial vehicle inspection task allocation model, generating an allocation decision of the inspection tasks of multiple unmanned aerial vehicles in the machine nest according to the location of the machine nest and the task objectives, and allocating inspection tasks to multiple unmanned aerial vehicles in the machine nest based on the allocation decision. The application of the present invention improves the efficiency of unmanned aerial vehicle inspection and reduces the inspection cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a flow chart of the method of the present invention;

[0055] Figure 2 A flowchart of an embodiment of the method of the present invention;

[0056] Figure 3 A schematic diagram of the distribution of poles and towers and drone airports within the inspection area of ​​an embodiment of the method of the present invention;

[0057] Figure 4 is an algorithm flow chart of an embodiment of the method of the present invention;

[0058] Figure 5 It is a structural diagram of the system of the present invention. DETAILED DESCRIPTION

[0059] Now, exemplary embodiments of the present invention are described with reference to the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely and to fully convey the scope of the present invention to those skilled in the art. The terms used in the exemplary embodiments shown in the accompanying drawings are not intended to limit the present invention. In the accompanying drawings, the same units / elements are marked with the same reference numerals.

[0060] Unless otherwise specified, the terms (including technical terms) used herein have the commonly understood meanings to those skilled in the art. In addition, it is understood that the terms defined in commonly used dictionaries should be understood to have the same meanings as those in the context of the relevant fields, and should not be understood as idealized or overly formal meanings.

[0061] Embodiment 1:

[0062] The present invention proposes a method for allocating inspection tasks of multiple unmanned aerial vehicles in a power grid, such as Figure 1 As shown, including:

[0063] Step 1: determine the regional characteristics of the target inspection area, and divide the target inspection area into multiple areas according to the regional characteristics, and establish a drone inspection task allocation model for the power grid towers to be inspected in each of the multiple areas;

[0064] Step 2: determine the area covered by the machine nest in each area, and plan the location of the machine nest in each area based on the area covered by the machine nest;

[0065] Step 3: Establish the task objectives of the multi-UAV inspection tasks for the power grid. Based on the UAV inspection task allocation model, generate an allocation decision for the inspection tasks of multiple UAVs in the machine nest according to the location and task objectives of the machine nest. Based on the allocation decision, allocate inspection tasks to multiple UAVs in the machine nest.

[0066] The regional characteristics of the target inspection area are determined, and the target inspection area is divided into multiple areas according to the regional characteristics, including:

[0067] Among them, the regional characteristics of the target inspection area are determined, and based on the regional characteristics, a fuzzy relationship matrix describing the attributes of each grid tower to be inspected within the inspection operation range is constructed, and the target inspection area is divided into multiple areas using the transitive closure method.

[0068] The target inspection area is divided into multiple areas, including:

[0069] Setting available index vectors of the power grid towers to be inspected, constructing an inspection factor hierarchical analysis matrix based on the available index vectors, and performing normalization processing on the inspection factor hierarchical analysis matrix;

[0070] The similarity between any two elements in the normalized inspection factor hierarchy analysis matrix is ​​calculated using the correlation coefficient method;

[0071] The similarities are dynamically clustered using a closure method to obtain dynamic clustering results, and based on the dynamic clustering results, the target inspection area is divided into multiple areas.

[0072] The similarities are dynamically clustered using a closure method, including:

[0073] Calculate the similarity distance between any two elements, and calculate the transitive closure of the similarity distance, sort the elements of the transitive closure, and find the corresponding truncation matrix based on the sorting result;

[0074] The elements in the inspection factor hierarchical analysis matrix are classified according to the cut matrix.

[0075] The method further includes: optimizing the location of the machine nest, including:

[0076] Establish a collection of power grid towers to be inspected, a collection of candidate locations, and a collection of machine nests within the scope of operation;

[0077] Establish mapping relationships among the set of grid towers to be inspected, the set of candidate locations and the set of machine nests within the operation range;

[0078] Based on the mapping relationship, a mathematical model is established, and based on the mathematical model, the layout positions of the machine nests are optimized.

[0079] The mapping relationship includes: towers inspected by drones, candidate locations that have been used, and already deployed machine nests.

[0080] Among them, the task goal is to complete all the planned tasks in the shortest time;

[0081] The constraints of the task objectives include:

[0082] Different task weights are assigned to each grid tower to be inspected within the scope of the operation task. In each area, grid towers to be inspected with high operation weights are given priority, and the higher the task weight, the longer the assigned operation time.

[0083] The present invention will be further described below in conjunction with specific embodiments:

[0084] Implementation steps: Figure 2 As shown, including:

[0085] (1) Analysis of UAV inspection conditions in multiple areas:

[0086] According to the geographical distribution of power grid towers, combined with the initial deployment point of drones and the performance status of drones, the scope of the operation mission can be divided into different areas. The main factors affecting the deployment planning of drone airports are: the distribution of transmission line towers under the jurisdiction of the line operation and maintenance unit, geographical environmental factors, distribution of airworthy areas and communication network conditions; the second is the number of drone airports and drones currently available for deployment, drone functions and performance indicators and other parameters; the third is the distribution of locations within the jurisdiction that have the ability to provide logistics support such as venues, electricity, maintenance services, etc. for the deployment of drone airports.

[0087] When formulating a regional drone multi-machine collaborative inspection strategy, two situations need to be considered. The first situation is that a certain number of drone airports have been deployed within the scope of the operation task. The inspection strategy is mainly about how to reasonably allocate the towers to be inspected within the scope of the operation task to each drone airport to collaboratively complete the inspection operation; the second situation is to reasonably select and deploy a certain number of drone airports within the scope of the operation task to maximize the benefits of drone inspections after deployment. After completing the site selection of drone airports, the subsequent collaborative inspection strategy can be formulated according to the first situation.

[0088] For transmission lines of different importance, DLT741 "Overhead Line Operation Regulations" mainly distinguishes in the inspection cycle, but the drone inspection operation procedures are basically the same, requiring the inspection operation plan to be completed according to the relevant standards of line operation and maintenance, and the inspection of fixed points on the tower to be completed according to the standard route. Therefore, for the drone inspection of lines in the region, the calculation indicators are designed based on the minimum time required to complete all the planned operations. Taking into account special conditions such as emergency inspections, it may be required to give priority to the inspection of certain important or critically defective line towers within a limited time, while taking into account the inspection of other line towers under the condition of energy. At this time, different task weights are assigned to each tower within the scope of the operation task, and the inspection area division gives priority to ensuring that towers with high weights are allocated sufficient operation time, and then the remaining time is allocated to as many other towers as possible.

[0089] (2) Task Modeling:

[0090] The classification object of the multi-inspection area division problem is the tower. Since the no-fly zone, important buildings and facility coordinates are known in advance, cluster analysis is used to establish the uncertainty description of the tower sample attributes.

[0091] Assume that the total number of towers to be classified is T base, then the sample set composed of T base towers is:

[0092] T={T1,T2,T3,…,T t}

[0093] The attributes of each tower related to drone inspection include latitude and longitude coordinates, task weight, influence of geographical environment conditions, influence of communication and navigation conditions, and influence of surrounding important facilities and buildings. All attributes together constitute an indicator vector that characterizes the characteristics of the tower.

[0094] The availability index vector of tower drone inspection is set. For each tower, the distance to all available drone airports within the task range is calculated using the traversal method according to its longitude and latitude coordinates. At the same time, the performance of the drone assigned to the airport is queried, and the effective flight time is calculated in turn to obtain the index vectors related to the availability of drone inspection of the tower.

[0095] Tower T i The indicator vector related to each drone airport within the inspection operation mission is.

[0096]

[0097] Construct the hierarchical analysis matrix of inspection elements, set the sample set T = {T1, T2, T3, ..., T t Each sample T in i The characteristic index vector for:

[0098]

[0099] in,

[0100] T is the number of towers to be inspected within the area, T = {T1, T2, T3, ..., T t};

[0101] n is the number of drone nests;

[0102] For the tower T i Airport with drones N j distance;

[0103] K 任i is the weight coefficient of the task influencing factors;

[0104] K 地i is the weight coefficient of the influencing factors of geographical environment conditions;

[0105] K 通i It is the weight coefficient of the factors affecting communication and navigation conditions, including at least three factors: navigation positioning error, navigation signal shielding, and network signal shielding.

[0106] The original matrix U of power grid inspection elements is constructed as:

[0107]

[0108] Each element in the original characteristic index vector matrix U of the power grid inspection element In order to make the elements with different dimensions comparable and clustered, it is necessary to standardize each element to eliminate the dimension effect. i The index vector of the jth column Standardization processing, according to the following formula, obtains the normalized matrix of power grid inspection elements

[0109]

[0110]

[0111] After standardization, the dimensions of each element in the characteristic index vector matrix are eliminated, and the standardized characteristic index vector matrix U′ is obtained:

[0112]

[0113] In the above formula, t′=t, n′=n+3. t and n are the number of towers to be inspected and the number of available drone airports within the scope of the operation task, respectively.

[0114] For different elements in any column of U′ and The similarity The calculation was performed using the correlation coefficient method.

[0115]

[0116] Dynamic clustering is performed using the improved closure method. Calculate the transitive closure of the similarity distance R. Calculate R in turn 2 , R 4 , ..., when it first appears hour, The transitive closure corresponding to the similarity matrix. Arrange the elements of the transitive closure from large to small and find all the corresponding truncation matrices. Classify according to the truncation matrix, traverse the subscripts of all elements of the truncation matrix, and put the same subscripts in one category when any group of subscripts appear in a concentrated manner. In this way, each related element is placed in the same list. The elements that have been put into the list have clear corresponding list values. The value will not be traversed again during the next traversal. Repeat the process until all elements are traversed and the classification result is obtained. The inspection area division result is obtained based on the dynamic clustering result.

[0117] (3) Multi-area and multi-machine inspection task deployment strategy:

[0118] Assuming that the poles and towers to be inspected within the scope of the on-site operation task have been divided into different inspection areas according to their characteristic attributes, and are respectively associated with available drone airports, the deployment of multiple drone inspection tasks for multiple inspection areas mainly involves the following two links. One link is to optimize the deployment of the pole inspection workload undertaken by each drone airport to achieve the highest efficiency in completing all the operation tasks; the other link is that when a drone airport is equipped with multiple drones, each drone needs to be further scheduled to complete the tasks of all the poles and towers to be inspected in the associated inspection area with the highest efficiency.

[0119] Therefore, in order to dispatch m drones taking off and landing from a drone airport to complete the inspection of all n towers in the inspection area, the essence is to minimize the inspection cost. According to the actual line inspection operation, this cost is mainly reflected in achieving the highest inspection efficiency, that is, the shortest task operation time.

[0120] Considering that the position to be avoided in a patrol area is relatively small compared to the entire area, the impact of different avoidance strategies on the flight mileage of the drone is relatively small, so this method uses a rectangular grid to perform relevant processing. For the pole tower, with its longitude and latitude coordinates as the center, a certain length is extended in the positive and negative directions of the x-axis and y-axis of the rectangular coordinate system. The length can be set to 30 meters ± 5 meters to form an avoidance zone. (If it is set to 30m, the 60m × 60m rectangular area formed is used as the avoidance zone). For other obstacles that cannot be crossed, the four vertices of the smallest rectangle that can cover the obstacle are used to mark the actual position and size of the obstacle in the rectangular coordinate system, and then each side of the rectangle is extended by 30 meters in the positive and negative directions of the x-axis and y-axis, and the new rectangular area is used as the avoidance zone. Establish a rectangular coordinate system for a certain inspection area such as Figure 3 shown.

[0121] In each path calculation, the drone is first set to travel back and forth between the drone airport or different towers along the shortest path. When the drone goes to the tower, it is considered that the drone mission deployment is completed when it reaches the edge of the avoidance zone of the tower, and then the operation is completed according to the autonomous inspection route of the tower. When the drone leaves the tower, the edge of the avoidance zone of the tower is also used as the starting point of the drone mission deployment. When the shortest path does not pass through other marked avoidance zones, no additional processing is required during this path calculation process.

[0122] Assume that there are n poles and towers (numbered from 1 to n) waiting to be inspected in a certain inspection area, and the drone airport (numbered 0) can dispatch m drones (numbered from 1 to m) at the same time. The operation duration of each drone is limited by the single flight time. Since the total amount of inspection work is uncertain, during a task dispatch process, the number of drones required to operate may be less than m, that is, some drones are idle; it is also possible that some drones may need to perform multiple operations to complete the inspection of the poles and towers in the inspection area, and the number of operations performed by the multiple drones involved is different. At this time, the total number of operations of all drones participating in the inspection task is m. ′ Greater than or equal to m.

[0123] The control goal of the multi-machine inspection task allocation strategy is to minimize the total inspection time T, that is:

[0124]

[0125] in:

[0126] t ijk is the time it takes for drone k to travel from point i to point j, which can be calculated by the distance between the two points and the flight speed of the drone. If points i and j pass through an avoidance zone, the shortest distance after bypassing the avoidance zone needs to be corrected in the above manner. When i or j ≠ 0, it means that the point is a pole tower; when i or j = 0, it means that the point is a drone airport.

[0127] t′ jk is the time it takes for UAV k to inspect tower j. When j = 0, it means that UAV k is at the airport. At this time, t ′ k =0.

[0128] T k is the flight time of the UAV k, which is determined by the performance of the UAV. is the flight time of drone k at the current position i. Obviously, when i = 0, it means that drone k is at the airport. The meanings of other symbols are the same as above.

[0129] The condition for merging the m′ virtual drone task allocation to make the m real drone inspection operation time St as evenly distributed as possible is max{St i -St j}≤St j . Among them, i, j∈{1, 2,…, m} and i≠j.

[0130] According to the above model, a multi-objective genetic algorithm is used to solve the operation task allocation of m′ virtual drones. After obtaining the optimal solution, the virtual drones are merged according to the number of real drones, so that the inspection operation time of each real drone is distributed as evenly as possible, that is, the multi-machine task allocation process in the inspection area is completed.

[0131] (4) Drone Airport Optimization Strategy

[0132] The main factors of drone airport deployment planning are similar to the aforementioned multi-inspection area division strategy within the scope of the operation task, but the key to airport deployment planning is to maximize the benefits of drone inspections of all transmission lines under its jurisdiction by utilizing the drone airports that can be deployed. The key to multi-inspection area division within the scope of the operation task is to efficiently complete the specific line inspection tasks of a certain inspection operation plan by utilizing the already deployed drone airports.

[0133] When there are enough drone airports available for deployment, reasonable planning can meet the drone inspection needs of all transmission lines under the jurisdiction of the operation and maintenance unit. Under the actual constraints of the site selection of the operation and maintenance unit, it is necessary to gradually increase the number of deployed drone airports in batches. Therefore, when establishing the drone airport optimization strategy model, it cannot be simply based on the idea of ​​"how many drone airports are needed to cover all transmission lines", but should follow the principle of "how many existing drone airports are there, which locations should be selected in turn among the deployment candidate points, and the coverage of the transmission lines range from large to small" for benefit calculation.

[0134] The control objectives of achieving the optimal layout strategy of drone airports can be further described as:

[0135] (1) Make sure that all airports are covered by as many poles as possible. In special cases, the operation and maintenance unit can set which poles must be covered;

[0136] (2) Make the sum of the distances between each airport and all the towers covered by it as small as possible. This control goal ensures that when a certain tower can be covered by the inspection mileage of multiple drone airports, one of the best-located airports can be selected to be responsible for the inspection of the tower, thereby maximizing the inspection efficiency. It should be pointed out that the airport with the best location for the tower is not the one with the smallest distance between the two, but is judged by comparing the sum of the distances of the tower sets associated with different airports.

[0137] Suppose that the operation and maintenance unit is optimizing the deployment of drone airports for the first time within the scope of its operations. It is known that there are Ma drone airports that need to be deployed. For convenience, the newly deployed drone airports and the deployed airports can be numbered together; there are Nt poles and towers within the jurisdiction that can carry out drone inspection operations, and some of the poles and towers are restricted to drone inspections; through research and field visits in advance, it is determined that there are a total of p candidate locations for deploying airports; the location and scope of other special areas such as no-fly zones are known. A data matrix A (Nt×p) can be established, and the element a in the matrix ij =(T i , D ij ,ω 0i ,ω pij ).

[0138] Among them, the row number of matrix A represents the number of all towers that can be used for inspection by drones, that is, The column numbers represent all candidate locations where drone airports can be deployed, i.e. T i D represents the equivalent operation mileage converted from the time required for the inspection of tower i by drone; ij Represents the distance between tower i and candidate location j. Obviously, this distance needs to be corrected according to the above steps.

[0139]

[0140] ω pij Indicates whether the tower is covered by an existing deployed drone airport. Its value is 0 or the drone airport number k. The value is k, indicating that the tower i has been deployed at the airport numbered k at location j to carry out drone inspection operations; for the first deployment airport, it is obvious that all ω in A(Nt×p) pij Both are 0.

[0141] For the convenience of calculation, the inspection mileage coverage of all Ma drone airports and the number of drones deployed are also established as a data matrix B (1×Ma), and the element b in the matrix is k =(Tm k , Na k ), Among them, Tm k Na represents the inspection mileage coverage of the drone airport numbered k; k Indicates the number of drones assigned to the drone airport numbered k.

[0142] Define Nt×p matrix A=[a ij ] and 1×Ma matrix B=[b kThe operation of ] is recorded as A■B, and its elements are the results of the operation of the elements of matrix A and the elements of B in sequence according to the following rules. The resulting matrix is ​​Nt×p. The element c in the matrix ij ={ω p |1≤ω p ≤Ma,ω p ∈Z}. c ij When it is not an empty set, it means that there is an airport represented by the element ωp at location j that can be deployed, and after deployment, drone inspection operations can be carried out on tower i; c ij When it is an empty set, it means that there is no suitable airport available for deployment at location j to carry out drone inspection operations on tower i.

[0143] The operation rules of the elements in matrix A and the elements in B are as follows: first, if ω pij If the value is non-zero, put it into set c ij Secondly, judge (T i +2D ij ) is not greater than Tm k , the Tm that satisfies the condition k The corresponding column number is added to the set c ij Therefore, after completing the A■B operation, the elements c in the obtained matrix C are ij That is, the collection of drone airports available for use on any base tower within the scope of the assigned operational mission and their corresponding deployment locations.

[0144] In order to meet the control goal of the airport layout optimization strategy of making all airports covered by as many towers as possible, the elements in the matrix C can be sorted to obtain the relevant parameters required for modeling. From the sorting, it can be seen that: within the scope of the task, there is a tower set T = {t i} can carry out drone inspection operations, and there is a set of candidate locations P = {p j} and the drone airport set M = {m k}; T i With P j and M k A mapping relationship is established between them, which includes factors such as the towers that must be inspected by drones, candidate locations that have been used, and drone airports that have been deployed. Based on this, a mathematical model can be established as follows.

[0145]

[0146] For any candidate location, only one drone airport can be deployed or none can be deployed.

[0147] So we have:

[0148]

[0149] For any candidate location, if it is not selected as a drone airport, no inspection operation can be carried out on any tower, so:

[0150]

[0151] All airports in the drone airport set M must be deployed, so:

[0152]

[0153] For any drone airport, only one candidate location can be selected for deployment, so there are:

[0154]

[0155] According to the above model, when the number of candidate locations or newly added drone airports to be deployed is small, the exhaustive method can be easily used for solution. In particular, due to the current limitations of drone functions and performance, the number of transmission line towers included in the operation coverage of an airport is very limited, generally only dozens of towers, which further saves the calculation time of the exhaustive method. In this project, the greedy algorithm is combined with the exhaustive algorithm for solution. The main steps are:

[0156] (1) Select each column in the matrix C in turn. According to the definition, this column is one of the candidate locations in the set P. Select each row element of this column in turn. Let the selected element be c ij , find element a in matrix A according to the mapping relationship ij Omega 0i and ω pij value, according to which the element c is determined ij Whether the value meets the requirements. ij , its {ω p} is added to the column to form a new airport set, denoted as {w′ p}. It should be noted that if ω 0i =1 but element c ij If it is an empty set, it means that there is no candidate location that can meet the restriction condition that tower i must carry out drone inspection; if there are multiple rows of elements in this column 0i =1 or ω pij ≠0, but the corresponding {ω p} If there are no common intersection elements, it means that it is impossible to meet all the preset constraints by deploying a drone airport at this candidate location. In both cases, the operation and maintenance unit should adjust the constraints or introduce new candidate locations and recalculate matrices A, B, and C;

[0157] (2) In all cases where the pre-set conditions that must be met are not involved and Select an element w′ from each column of p The candidate location represented by this column is the drone airport to be deployed. After completing the selection of all columns, check whether the constraints required by the model are met. If the constraints are not met, the w′ that is repeated between different columns is p For the convenience of subsequent explanation, the columns involving the preset limiting conditions that must be met are added to the collection Add the other columns to the collection Obviously, after the calculation is completed, |R1|+|R2|=|M|;

[0158] (3) Calculate f = ∑ i∈T y i And compare it with the last calculated value, and keep the larger value. For the first calculation, the last calculated value is 0;

[0159] (4) Delete any column in set R2 and repeat steps (2) and (3) until |R1|+|R2|=|M|;

[0160] (5) Check whether there are other selectable w′ in each column of set R2 p If there is no, the calculation ends; if there is, replace the current w′ one by one p , calculate f respectively, retain the maximum value, and end the calculation. The main process of the algorithm is as follows Figure 4 As shown;

[0161] For the set T, the elements can be divided into two categories: one is that only one airport can carry out inspections; the other is that multiple airports can carry out inspections. To facilitate the distinction, the elements are marked as and The subscript i represents the tower number; the superscript n∈N + Indicates the number of airports that can be used for drone inspection of the tower. Each tower in the system has only one airport available, leaving no room for adjustment in subsequent optimization.

[0162] Based on the above analysis, the mathematical model for establishing secondary optimization of airport layout to optimize the overall inspection efficiency is as follows:

[0163]

[0164] is the number of drones assigned to the airport numbered j in the set M; T i and D ij The meaning is the same as above.

[0165]

[0166] F is to make the overall path shortest; Sj In order to make the time taken by each airport to complete the inspection workload as close as possible, the time-related parameters are not explicitly expressed in the above formula because the operating speed of drone inspection is mainly determined by relevant technical regulations and is a certain value.

[0167] Affected by factors such as the dense distribution of poles and towers or the different inspection coverage radius of the drones assigned to the airport, it is not always possible to meet this constraint condition while meeting the first control goal of the drone airport optimization strategy. Often, adjustments can only be made between airports with common poles and towers within the operation coverage area. Therefore, S1 and S2 are mainly set by the operation and maintenance unit based on local actual conditions. In most cases, due to the long time span of the planned inspection cycle, S1 and S2 are set for each airport. j There is no strong need to optimize the value, so this constraint can be ignored in the model.

[0168] To further reduce the amount of calculation, the invariant can be separated out so that it only needs to be calculated once in the algorithm, thus:

[0169] F=f c +f v

[0170]

[0171] S j =S jc +S jv

[0172]

[0173] For the case of temporarily adding an airport to a certain inspection area to improve inspection efficiency, a similar method in this section can be used after the above calculations are completed. The difference is that the adjustment range is limited to all towers in the inspection area, and these towers can be used for the newly added temporary airport. Obviously, for the special case where there is only one additional candidate location in the area, a simple exhaustive method can be used; when there are multiple additional candidate locations but only one airport can be added, an exhaustive method or a greedy take-away heuristic algorithm can be used; when there are multiple additional candidate locations and multiple airports that can be added, a greedy algorithm is used for calculation.

[0174] The present invention provides a basis for the division of power grid inspection areas in a modeling manner, and reasonably solves the problem of location selection of drone nests. The provided drone task allocation method is also applicable to the nest position after it is fixed, and it is automatically sent to the nest. The nest allocates the inspection tasks in the area to different drones, further improving the efficiency of drone inspection and reducing inspection costs.

[0175] In the present invention:

[0176] Based on drone inspection of lines in a certain area, the calculation indicators are designed with the minimum time required to complete all planned tasks as the basic condition.

[0177] Constraints: Different task weights are assigned to each tower within the scope of the task. The inspection area division gives priority to ensuring that towers with high weights are allocated sufficient working time, and then the remaining time is allocated to as many other towers as possible.

[0178] A fuzzy relationship matrix describing the attributes of each tower to be inspected within the inspection operation range is constructed, and the transitive closure method is used to complete the division of different inspection areas.

[0179] (1) Set the tower available index vector. Tower T i The indicator vector related to each drone airport within the inspection operation mission is.

[0180]

[0181] (2) Construct a hierarchical analysis matrix of inspection elements and normalize the matrix.

[0182] Suppose the sample set T is composed of T-based towers in the region = {T1, T2, T3, ..., T t Each sample T in i The characteristic index vector for:

[0183]

[0184] in:

[0185] T is the number of towers to be inspected within the area, T = {T1, T2, T3, ..., T t};

[0186] n is the number of drone nests;

[0187] For the tower T i Airport with drones N j distance;

[0188] K 任i is the weight coefficient of the task influencing factors;

[0189] K 地i is the weight coefficient of the influencing factors of geographical environment conditions;

[0190] K 通i It is the weight coefficient of the factors affecting communication and navigation conditions, including at least three factors: navigation positioning error, navigation signal shielding, and network signal shielding.

[0191] (3) U ′Different elements in any column and The similarity The calculation was performed using the correlation coefficient method.

[0192]

[0193] (4) Dynamic clustering using improved closure method.

[0194] Compute the transitive closure of the similarity distance R. Compute R in turn 2 , R 4 , ..., when it first appears hour, For the transitive closure corresponding to the similar matrix, arrange the elements of the transitive closure from large to small and find all the corresponding truncation matrices.

[0195] Classify according to the truncation matrix, traverse the subscripts of all elements of the truncation matrix, and put the same subscripts in one category when any group of subscripts appear in a group. In this way, each related element is placed in the same list. The elements that have been put into the list have clear corresponding list values. The value will not be traversed again during the next traversal. Repeat this process until all elements are traversed and the classification result is obtained.

[0196] (5) The inspection area division results are obtained based on the dynamic clustering results.

[0197] To optimize the location of the airport, there are tower sets T = {t i} can carry out drone inspection operations, and there is a set of candidate locations P = {p j} and the drone airport set M = {m k}; T i With P j and M k A mapping relationship is established between them, including the towers that must be inspected by drones, the candidate locations that have been used, and the drone airports that have been deployed, and a mathematical model is established as follows:

[0198]

[0199] Embodiment 2:

[0200] The present invention also proposes a system 200 for allocating inspection tasks of multiple UAVs in a power grid, such as Figure 5 As shown, including:

[0201] The area division unit 201 is used to determine the regional characteristics of the target inspection area, and divide the target inspection area into multiple areas according to the regional characteristics, and establish a drone inspection task allocation model for the power grid towers to be inspected in each of the multiple areas;

[0202] The machine nest point distribution unit 202 is used to determine the area covered by the machine nest in each area, and plan the location of the machine nest in each area based on the area covered by the machine nest;

[0203] The task allocation unit 203 is used to establish the task objectives of the multi-UAV inspection tasks of the power grid, generate allocation decisions for the inspection tasks of multiple UAVs in the machine nest based on the UAV inspection task allocation model and according to the location and task objectives of the machine nest, and allocate inspection tasks to multiple UAVs in the machine nest based on the allocation decisions.

[0204] The regional characteristics of the target inspection area are determined, and the target inspection area is divided into multiple areas according to the regional characteristics, including:

[0205] Determine the regional characteristics of the target inspection area, and based on the regional characteristics, construct a fuzzy relationship matrix describing the attributes of each grid tower to be inspected within the inspection operation range, and use the transitive closure method to divide the target inspection area into multiple areas.

[0206] The target inspection area is divided into multiple areas, including:

[0207] Setting available index vectors of the power grid towers to be inspected, constructing an inspection factor hierarchical analysis matrix based on the available index vectors, and performing normalization processing on the inspection factor hierarchical analysis matrix;

[0208] The similarity between any two elements in the normalized inspection factor hierarchy analysis matrix is ​​calculated using the correlation coefficient method;

[0209] The similarities are dynamically clustered using a closure method to obtain dynamic clustering results, and based on the dynamic clustering results, the target inspection area is divided into multiple areas.

[0210] The similarities are dynamically clustered using a closure method, including:

[0211] Calculate the similarity distance between any two elements, and calculate the transitive closure of the similarity distance, sort the elements of the transitive closure, and find the corresponding truncation matrix based on the sorting result;

[0212] The elements in the inspection factor hierarchical analysis matrix are classified according to the cut matrix.

[0213] The point distribution unit 203 is also used to optimize the point distribution position of the machine nest, including:

[0214] Establish a collection of power grid towers to be inspected, a collection of candidate locations, and a collection of machine nests within the scope of operation;

[0215] Establish mapping relationships among the set of grid towers to be inspected, the set of candidate locations and the set of machine nests within the operation range;

[0216] Based on the mapping relationship, a mathematical model is established, and based on the mathematical model, the layout positions of the machine nests are optimized.

[0217] The mapping relationship includes: towers inspected by drones, candidate locations that have been used, and already deployed machine nests.

[0218] Among them, the task goal is to complete all the planned tasks in the shortest time;

[0219] The constraints of the task objectives include:

[0220] Different task weights are assigned to each grid tower to be inspected within the scope of the operation task. In each area, grid towers to be inspected with high operation weights are given priority, and the higher the task weight, the longer the assigned operation time.

[0221] Embodiment 3:

[0222] Based on the same inventive concept, the present invention also provides a computer device, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding functions, so as to implement the steps of the method in the above embodiment.

[0223] Embodiment 4:

[0224] Based on the same inventive concept, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It is understandable that the computer-readable storage medium here can include both a built-in storage medium in a computer device and an extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the steps of the method in the above embodiment.

[0225] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0226] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0227] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0228] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0229] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0230] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for allocating inspection tasks of multiple unmanned aerial vehicles in a power grid, characterized in that: include: Determine the regional characteristics of the target inspection area, and divide the target inspection area into multiple areas according to the regional characteristics, and establish a drone inspection task allocation model for the power grid towers to be inspected in each of the multiple areas; Determine the area covered by the machine nest in each area, and plan the location of the machine nest in each area based on the area covered by the machine nest; Establish the task objectives of the multi-UAV inspection tasks of the power grid, generate the allocation decision of the inspection tasks of multiple UAVs in the machine nest based on the UAV inspection task allocation model and the location and task objectives of the machine nest, and allocate inspection tasks to multiple UAVs in the machine nest based on the allocation decision.

2. The method according to claim 1, characterized in that: The determining of the regional characteristics of the target inspection area and dividing the target inspection area into a plurality of areas according to the regional characteristics include: Determine the regional characteristics of the target inspection area, and based on the regional characteristics, construct a fuzzy relationship matrix describing the attributes of each grid tower to be inspected within the inspection operation range, and use the transitive closure method to divide the target inspection area into multiple areas.

3. The method according to claim 2, characterized in that The target inspection area is divided into multiple areas, including: Setting available index vectors of the power grid towers to be inspected, constructing an inspection factor hierarchical analysis matrix based on the available index vectors, and performing normalization processing on the inspection factor hierarchical analysis matrix; The similarity between any two elements in the normalized inspection factor hierarchy analysis matrix is ​​calculated using the correlation coefficient method; The similarities are dynamically clustered using a closure method to obtain dynamic clustering results, and based on the dynamic clustering results, the target inspection area is divided into multiple areas.

4. The method according to claim 3, characterized in that The method of dynamically clustering the similarities using the closure method includes: Calculate the similarity distance between any two elements, and calculate the transitive closure of the similarity distance, sort the elements of the transitive closure, and find the corresponding truncation matrix based on the sorting result; The elements in the inspection factor hierarchical analysis matrix are classified according to the cut matrix.

5. The method according to claim 1, characterized in that The method further includes: optimizing the location of the machine nest, including: Establish a collection of power grid towers to be inspected, a collection of candidate locations, and a collection of machine nests within the scope of operation; Establish mapping relationships among the set of grid towers to be inspected, the set of candidate locations and the set of machine nests within the operation range; Based on the mapping relationship, a mathematical model is established, and based on the mathematical model, the layout positions of the machine nests are optimized.

6. The method according to claim 5, characterized in that The mapping relationship includes: towers inspected by drones, candidate locations that have been used, and deployed machine nests.

7. The method according to claim 1, characterized in that The task goal is to complete all the planned tasks in the shortest time possible; The constraints of the task objectives include: Different task weights are assigned to each grid tower to be inspected within the scope of the operation task. In each area, grid towers to be inspected with high operation weights are given priority, and the higher the task weight, the longer the assigned operation time.

8. A system for allocating inspection tasks of multiple unmanned aerial vehicles in a power grid, characterized in that: include: A division area unit is used to determine the regional characteristics of the target inspection area, and divide the target inspection area into multiple areas according to the regional characteristics, and establish a drone inspection task allocation model for the power grid towers to be inspected in each of the multiple areas; The machine nest distribution unit is used to determine the area covered by the machine nest in each area, and plan the distribution position of the machine nest in each area based on the area covered by the machine nest; The task allocation unit is used to establish the task objectives of the multi-UAV inspection tasks of the power grid, generate allocation decisions for the inspection tasks of multiple UAVs in the machine nest based on the UAV inspection task allocation model and according to the location and task objectives of the machine nest, and allocate inspection tasks to multiple UAVs in the machine nest based on the allocation decisions.

9. The system according to claim 8, characterized in that The determining of the regional characteristics of the target inspection area and dividing the target inspection area into a plurality of areas according to the regional characteristics include: Determine the regional characteristics of the target inspection area, and based on the regional characteristics, construct a fuzzy relationship matrix describing the attributes of each grid tower to be inspected within the inspection operation range, and use the transitive closure method to divide the target inspection area into multiple areas.

10. The system according to claim 9, characterized in that The target inspection area is divided into multiple areas, including: Setting available index vectors of the power grid towers to be inspected, constructing an inspection factor hierarchical analysis matrix based on the available index vectors, and performing normalization processing on the inspection factor hierarchical analysis matrix; The similarity between any two elements in the normalized inspection factor hierarchy analysis matrix is ​​calculated using the correlation coefficient method; The similarities are dynamically clustered using a closure method to obtain dynamic clustering results, and based on the dynamic clustering results, the target inspection area is divided into multiple areas.

11. The system according to claim 10, characterized in that The method of dynamically clustering the similarities using the closure method includes: Calculate the similarity distance between any two elements, and calculate the transitive closure of the similarity distance, sort the elements of the transitive closure, and find the corresponding truncation matrix based on the sorting result; The elements in the inspection factor hierarchical analysis matrix are classified according to the cut matrix.

12. The system according to claim 8, characterized in that The point distribution unit is also used to optimize the point distribution position of the machine nest, including: Establish a collection of power grid towers to be inspected, a collection of candidate locations, and a collection of machine nests within the scope of operation; Establish mapping relationships among the set of grid towers to be inspected, the set of candidate locations and the set of machine nests within the operation range; Based on the mapping relationship, a mathematical model is established, and based on the mathematical model, the layout positions of the machine nests are optimized.

13. The system according to claim 12, characterized in that The mapping relationship includes: towers inspected by drones, candidate locations that have been used, and deployed machine nests.

14. The system according to claim 8, characterized in that The task goal is to complete all the planned tasks in the shortest time possible; The constraints of the task objectives include: Different task weights are assigned to each grid tower to be inspected within the scope of the operation task. In each area, grid towers to be inspected with high operation weights are given priority, and the higher the task weight, the longer the assigned operation time.

15. A computer device, characterized in that: include: one or more processors; a processor for executing one or more programs; When the one or more programs are executed by the one or more processors, the method according to any one of claims 1 to 7 is implemented.

16. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed, the method according to any one of claims 1 to 7 is implemented.

Images