Multi-type unmanned aerial vehicle nest site selection method and device based on routing inspection task distribution

Abstract

The invention discloses a multi-type unmanned aerial vehicle nest site selection method and device based on routing inspection task allocation, and the method comprises the steps: giving out a first objective function corresponding to the number of unmanned aerial vehicle nests and a first constraint condition corresponding to the positions of the unmanned aerial vehicle nests based on a routing inspection task; obtaining a candidate address set based on the first objective function and the first constraint condition; constructing a second objective function and a second constraint condition based on the unmanned aerial vehicle type decision and the inspection task according to the inspection cost, the inspection range and the inspection area of each type of unmanned aerial vehicle; and based on a second objective function and a second constraint condition, through a meta-heuristic optimization algorithm, performing optimization in the candidate address set to obtain a target nest position set corresponding to each type of unmanned aerial vehicles. Reasonable arrangement of the nest of each type of unmanned aerial vehicle is realized, each type of unmanned aerial vehicle is combined for use in the electric power inspection process, respective advantages are fully exerted, and efficient inspection of the unmanned aerial vehicle is ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of drone nest site selection, and in particular relates to a method and device for multi-type drone nest site selection based on inspection task allocation. Background Art

[0002] With the continuous expansion of the power grid and the continuous increase of power facilities, the task of power inspection has become more and more arduous and complicated. The traditional inspection method mainly relies on manual work, but this method has problems such as low inspection efficiency, great safety hazards, and high costs. Therefore, the use of drones for power inspection has become a trend.

[0003] Drone inspection has the advantages of flexibility, efficiency and safety, which can significantly improve inspection efficiency and quality and reduce inspection costs. In drone inspection, fixed-wing drones and multi-rotor drones each have their own advantages. Fixed-wing drones have the advantages of high speed and long range, which are suitable for large-scale and long-distance inspections; while multi-rotor drones have the advantages of low flight altitude, maneuverability, simple operation and low cost, which are suitable for refined inspection tasks.

[0004] At present, in actual applications, fixed-wing drones or multi-rotor drones are used alone for power inspections. Although fixed-wing drones are fast and have a long range, they cannot hover and take off and land vertically like multi-rotor drones, so their inspection capabilities in complex terrain and narrow spaces are limited. Although multi-rotor drones are maneuverable and flexible, their endurance, flight speed, and communication distance are relatively limited, making it difficult to meet the needs of large-scale and long-distance inspections.

[0005] How to combine fixed-wing drones and multi-rotor drones in the process of power inspection and give full play to their respective advantages is a problem that needs to be solved at present. However, before using the combined inspection, how to reasonably set up the nests of fixed-wing drones and multi-rotor drones to ensure efficient inspection of drones is a problem that needs to be solved at present. Summary of the invention

[0006] In view of the defects in the above-mentioned prior art, the present invention provides a method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation, including:

[0007] Based on the inspection task, the first objective function corresponding to the number of drone nests and the first constraint condition corresponding to the location of the drone nest are given respectively;

[0008] Based on the first objective function and the first constraint condition, a candidate address set is obtained;

[0009] According to the inspection cost, inspection scope and inspection area of ​​each type of drone, a second objective function and a second constraint condition based on drone type decision and inspection task are constructed;

[0010] Based on the second objective function and the second constraint, a meta-heuristic optimization algorithm is used to optimize the candidate address set and obtain the target nest location set corresponding to each type of UAV.

[0011] It can be seen from the above technical solution that by analyzing the inspection tasks, the number of drone nests is limited and a set of candidate addresses is obtained. Then, the inspection cost, inspection range and inspection area of ​​each type of drone are further limited at the economic level. A meta-heuristic optimization algorithm is used to optimize the candidate address set, and the target nest location set is obtained, which realizes the reasonable setting of each type of drone nest, allowing each type of drone to be used in combination during the power inspection process, giving full play to their respective advantages and ensuring efficient drone inspection.

[0012] In a possible implementation manner, the first objective function is specifically expressed as:

[0013]

[0014] The first constraint is specifically expressed as:

[0015]

[0016] Among them, x j is the result of establishing a nest at the jth pending address, and x j is a Boolean variable, J is the set of pending addresses in the inspection area, X ij is the relationship between the distance between the jth pending address and the ith target inspection device and the inspection range of the drone, and X ij is a Boolean variable, and I is the set of all target inspection devices in the inspection area.

[0017] In a possible implementation, based on the first objective function and the first constraint condition, a candidate address set is obtained, including:

[0018] Based on the first objective function and the first constraint condition, a set of pending addresses is obtained;

[0019] Based on the pending address set, a corresponding relationship is established between each target inspection device in the inspection area and the nearest pending address to obtain a candidate address set.

[0020] In a possible implementation, the inspection cost includes construction cost, flight cost and penalty cost.

[0021] According to the inspection cost, inspection scope and inspection area of ​​each type of drone, the second objective function and second constraint conditions based on drone type decision and inspection task are constructed, including:

[0022] Obtain the construction cost, flight cost and penalty cost of each type of drone, combine the correspondence between each type of drone and the target inspection equipment and the flight distance, analyze the probability of failure of each type of drone nest, and construct the second objective function;

[0023] The inspection range of each type of UAV and the flight distance between each type of UAV and the target inspection equipment are obtained, the corresponding relationship between each type of UAV and the target inspection equipment is analyzed, and the second constraint condition is constructed based on the number of each type of UAV.

[0024] In a possible implementation, the types of drones include fixed-wing drones and multi-rotor drones; the second objective function is specifically expressed as:

[0025]

[0026] Among them, J * is the candidate address set for building the nest, I is the set of all target inspection equipment in the inspection area, c1 is the construction cost of the fixed-wing UAV nest, y j is the result of building a fixed-wing drone nest at the jth candidate address, and y j is a Boolean variable, c0 is the construction cost of the multi-rotor drone nest, z j is the result of building a multi-rotor drone nest at the jth candidate address, and z j is a Boolean variable, T is the total running time of the inspection system including all types of drones, and f i is the inspection service demand frequency of the i-th target inspection equipment, α is the flight cost per unit distance of the UAV, d ij is the flight distance between the i-th target inspection device and the j-th candidate address, Y ij is the correspondence between the fixed-wing UAV in the jth candidate address and the i-th target inspection equipment, and Y ij is a Boolean variable, P fail is the probability of failure of the multi-rotor drone nest, Z ij is the correspondence between the multi-rotor drone in the jth candidate address and the i-th target inspection equipment, and Z ij is a Boolean variable, β is the penalty cost per unit distance of manual inspection when the multi-rotor drone nest fails, is the penalty distance when the i-th target inspection equipment is not covered by the multi-rotor drone nest.

[0027] In a possible implementation, the types of drones include fixed-wing drones and multi-rotor drones; the second constraint condition is specifically expressed as:

[0028]

[0029] Among them, J * is the candidate address set for building the machine nest, I is the set of all target inspection devices in the inspection area, d ij is the flight distance between the i-th target inspection device and the j-th candidate address, Y ij is the correspondence between the fixed-wing UAV in the jth candidate address and the i-th target inspection equipment, and Y ij is a Boolean variable, Z ij is the correspondence between the multi-rotor drone in the jth candidate address and the i-th target inspection equipment, and Z ij is a Boolean variable, y j is the result of building a fixed-wing drone nest at the jth candidate address, and y j is a Boolean variable, z j is the result of building a multi-rotor drone nest at the jth candidate address, and z j is a Boolean variable, R is the maximum service radius of the candidate address, N1 is the maximum number of candidate addresses for fixed-wing UAVs, and N0 is the maximum number of candidate addresses for multi-rotor UAVs.

[0030] In a possible implementation, based on the second objective function and the second constraint condition, a meta-heuristic optimization algorithm is used to search for the best among the candidate address sets to obtain a target nest location set corresponding to each type of drone, specifically including:

[0031] Based on the model operator for changing the type of the drone at the candidate address, the type of the drone at the candidate address in the candidate address set is changed, and optimization iteration is performed to obtain the first optimization set;

[0032] Based on the first optimization set, the relational operator and the hybrid operator are combined to change the drone type of the candidate address and / or adjust the corresponding relationship between the candidate address and the target inspection equipment, and perform optimization iteration to obtain a second optimization set, wherein the relational operator is used to adjust the corresponding relationship between the candidate address and the target inspection equipment, and the hybrid operator includes the model operator and the relational operator;

[0033] Based on the relational operator, the corresponding relationship between the candidate addresses and the target inspection equipment in the second optimization set is adjusted, and optimization iteration is performed to obtain the target machine nest position set.

[0034] In a possible implementation, based on the first optimization set, in combination with the relational operator and the hybrid operator, the drone type of the candidate address is changed and / or the corresponding relationship between the candidate address and the target inspection equipment is adjusted, and optimization iteration is performed to obtain the second optimization set, which specifically includes:

[0035] Compare the random number with the preset optimization threshold to obtain a comparison result;

[0036] Determine an operator corresponding to the comparison result, and optimize the first optimization set based on the operator corresponding to the comparison result;

[0037] If the number of iterations reaches a preset second iteration threshold, a second optimization set is obtained.

[0038] In a possible implementation, based on the relational operator, the corresponding relationship between the candidate addresses and the target inspection devices in the second optimization set is adjusted, and optimization iteration is performed to obtain the target machine nest location set, specifically including:

[0039] Perform the mth iteration on the second optimization set, adjust the correspondence between the candidate addresses and the target inspection equipment, and obtain the iterative result set Q m , where m is a positive integer and m is greater than 1;

[0040] According to Q m The corresponding second objective function value and Q m-1 The difference between the corresponding second objective function values ​​determines the current optimal result set and updates the iteration parameter T corresponding to the iteration result set m,sa ;

[0041] When T m,sa When m is greater than a preset temperature threshold or m is greater than a preset maximum iteration threshold, the current optimal result set is used as the target machine nest position set.

[0042] The present invention also discloses a multi-type UAV nest site selection device based on inspection task allocation, comprising:

[0043] A first condition building module is used to provide a first objective function corresponding to the number of drone nests and a first constraint condition corresponding to the location of the drone nest based on the inspection task;

[0044] A candidate address determination module, configured to obtain a candidate address set based on a first objective function and a first constraint condition;

[0045] The second condition construction module is used to construct a second objective function and a second constraint condition based on the drone type decision and the inspection task according to the inspection cost, inspection scope and inspection area of ​​each type of drone;

[0046] The target location determination module is used to search for the best among the candidate address sets based on the second objective function and the second constraint condition through a meta-heuristic optimization algorithm to obtain a target nest location set corresponding to each type of UAV.

[0047] The method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation provided by the present invention has at least the following beneficial effects:

[0048] By establishing a two-stage multi-type UAV nest site selection method based on inspection task allocation, the target nest location set with the minimum total cost is determined. The target nest location set includes the deployment location of the nest, the type of UAV in the nest, and the distribution of target inspection equipment among the nests. In the first stage, the number of UAV nests is limited by analyzing the inspection tasks and a set of candidate addresses is obtained. In the second stage, the inspection cost, inspection scope and inspection area of ​​each type of UAV are further limited at the economic level. A meta-heuristic optimization algorithm is used to optimize the candidate address set to obtain a set of target nest locations, which realizes the reasonable setting of fixed-wing UAV nests and multi-rotor UAV nests, allowing fixed-wing UAVs and multi-rotor UAVs to be used in combination during the power inspection process, giving full play to their respective advantages and ensuring efficient UAV inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A flow chart of a method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation provided by an embodiment of the present invention;

[0050] Figure 2 A flow chart of a meta-heuristic optimization algorithm provided by an embodiment of the present invention;

[0051] Figure 3 A schematic diagram of the operation of the machine model operator provided in an embodiment of the present invention;

[0052] Figure 4 A schematic diagram of the operation of a relational operator provided in an embodiment of the present invention;

[0053] Figure 5 A schematic diagram of the structure of the inspection area provided by an embodiment of the present invention;

[0054] Figure 6 A schematic diagram of the structure of a candidate address set provided by an embodiment of the present invention;

[0055] Figure 7 A schematic diagram of the structure of a target machine nest position set provided by an embodiment of the present invention;

[0056] Figure 8 A structural diagram of a multi-type UAV nest site selection device based on inspection task allocation provided in an embodiment of the present invention.

[0057] Among them, 201 is a first condition building module; 202 is a candidate address determination module; 203 is a second condition building module; 204 is a target location determination module. DETAILED DESCRIPTION

[0058] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0059] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings, and "multiple" generally includes at least two.

[0060] It should also be noted that the term "includes", "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, so that a commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprising a ..." do not exclude the existence of other identical elements in the commodity or device including the elements.

[0061] There are several target inspection equipment to be inspected distributed in the inspection area, including UHV transmission line towers, transmission lines, power towers, substations and distribution facilities. In order to improve the automation level of power inspection, further improve inspection efficiency and reduce costs, it is necessary to deploy drone nests in the inspection area to realize automatic power inspection by drones. Each drone nest is paired with a drone, which is responsible for the inspection of several target inspection equipment within the coverage of the nest. When performing each inspection task, the drone departs from the nest to the target detection equipment to be inspected after the nest is fully charged. After the inspection task is completed, it returns to the nest to charge and wait for the next task.

[0062] The site selection of the drone nest needs to comprehensively consider multiple factors such as terrain, climate, and distribution of power facilities to ensure safe take-off and landing of drones and efficient inspections. The allocation of inspection tasks needs to be comprehensively considered based on factors such as the performance characteristics of the drone, inspection objectives and their priorities to optimize inspection efficiency and quality.

[0063] Once the location of the nest is selected, it will not be easily changed. Due to the limitations of power supply, signal, network and safety factors, the drone nest can only be deployed near the UHV transmission line. In actual operation, the drone nest may fail due to weather, mechanical failure and other reasons, which will have a serious impact on power inspection. Due to the different inspection task requirements of UHV transmission lines, the quality and maintenance investment of drone nests, drone nests are usually divided into fixed-wing drone nests and multi-rotor drone nests. Fixed-wing drone nests are generally deployed in exclusive areas near UHV transmission lines. They are large in size, high in construction cost and require sufficient operation and maintenance costs. The present invention believes that fixed-wing drone nests always maintain normal operation during the mission cycle, and fixed-wing drones have a long endurance. Multi-rotor drone nests are generally deployed directly on UHV transmission towers. They are small in size and low in construction cost, but there is a certain probability that they will fail and cannot work properly. Once the multi-rotor drone nest fails, the corresponding inspection task needs to be completed manually, thereby generating additional manual inspection costs. In the actual machine nest inspection problem, the cost of building a machine nest accounts for a large proportion of the total cost, so decision makers usually first hope to use the least number of machine nests to complete the full coverage of all target inspection equipment to be inspected in the inspection area. Then, the type of machine nest to be used and the distribution plan of inspection tasks are selected based on factors such as the inspection frequency of the UHV transmission line, the distance between the transmission tower and the machine nest, the reliability of the machine nest, and labor costs.

[0064] The present invention proposes a method for selecting a multi-type UAV nest site based on inspection task allocation. By determining the deployment locations and types of fixed-wing UAV nests and multi-rotor UAV nests and the distribution of inspection demand points among the nests, the nest site selection and the type of UAV corresponding to the nest location are determined. The inspection tasks of all target inspection equipment in the inspection area are achieved on the basis of reducing costs, while improving the efficiency and quality of power inspections, providing a strong guarantee for the safe operation of power facilities.

[0065] like Figure 1 As shown, an embodiment of the present invention provides a method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation, and the specific steps are as follows:

[0066] S101: Based on the inspection task, a first objective function corresponding to the number of drone nests and a first constraint condition corresponding to the location of the drone nests are respectively given.

[0067] Specifically, the first objective function is expressed as:

[0068]

[0069] The first constraint is specifically expressed as:

[0070]

[0071] Among them, x j is the result of establishing a nest at the jth pending address, and x j is a Boolean variable. If x j = 0, it means that the machine nest is not established at the jth pending address. j =1, it means that the machine nest is established at the jth pending address, J is the set of pending addresses in the inspection area, X ij is the relationship between the distance between the jth pending address and the ith target inspection device and the inspection range of the drone, and X ij is a Boolean variable. If X ij = 0, it means that the distance between the jth pending address and the ith target inspection device is not within the inspection range of the drone. ij =1, it means that the distance between the jth pending address and the ith target inspection device is within the inspection range of the drone, and I is the set of all target inspection devices in the inspection area.

[0072] The first objective function is used to minimize the number of drone nest construction points, that is, to build the least number of nests to achieve the inspection tasks of all target inspection equipment in the inspection area. The first formula in the first constraint condition indicates that for any target inspection equipment i∈I, it is within the coverage of at least one drone nest j∈J. The second formula in the first constraint condition indicates that only when a nest is built at the pending address j∈J, the drone can start from the pending address j to inspect the target inspection equipment i∈I.

[0073] S102: Obtain a candidate address set based on the first objective function and the first constraint condition.

[0074] Specifically, based on the first objective function and the first constraint condition, a set of pending addresses is obtained. Based on the set of pending addresses, a corresponding relationship is established between each target inspection device in the inspection area and the nearest pending address to obtain a set of candidate addresses.

[0075] In a specific implementation, the above-mentioned undetermined address set can be obtained by inputting the first objective function and the first constraint condition into the gurobi solver to obtain a candidate address set, or by using other optimization algorithms, which is not limited to this.

[0076] It is understandable that there is no corresponding relationship between the candidate addresses in the candidate address set obtained according to the first constraint condition and the first objective function and the target inspection device. It is only necessary to satisfy that the target inspection device is at least within the coverage range of one candidate address. There may be a situation where multiple candidate addresses cover the same target inspection device. Based on this, the present invention establishes a corresponding relationship between each target inspection device and the nearest pending address in the pending address set by adopting a greedy algorithm to obtain a candidate address set. And the candidate address set is used as the initial solution in the solution process based on the second objective function and the second constraint condition. In a specific example, since the construction cost of a multi-rotor drone nest is lower than that of a fixed-wing drone nest, the drone types of all candidate addresses in the candidate address set are set to multi-rotor drones. In other embodiments, the drone type in the initial solution can be set to other drone types, without limitation.

[0077] S103: According to the inspection cost, inspection scope and inspection area of ​​each type of UAV, a second objective function and a second constraint condition based on UAV type decision and inspection task are constructed.

[0078] Specifically, the inspection cost includes construction cost, flight cost and penalty cost. The construction cost, flight cost and penalty cost of each type of drone are obtained, and the probability of failure of each type of drone nest is analyzed in combination with the corresponding relationship and flight distance between each type of drone and the target inspection equipment, and the second objective function is constructed. The inspection range of each type of drone and the flight distance between each type of drone and the target inspection equipment are obtained, and the corresponding relationship between each type of drone and the target inspection equipment is analyzed. The second constraint condition is constructed in combination with the number of each type of drone.

[0079] Furthermore, the second objective function is specifically expressed as:

[0080]

[0081] Among them, J * is the candidate address set for building the nest, I is the set of all target inspection equipment in the inspection area, c1 is the construction cost of the fixed-wing UAV nest, y j is the result of building a fixed-wing drone nest at the jth candidate address, and y j is a Boolean variable. If y j = 0, it means that the fixed-wing UAV nest will not be built at the jth candidate address. j =1, it means that a fixed-wing UAV nest is built at the jth candidate address, c0 is the construction cost of the multi-rotor UAV nest, z j is the result of building a multi-rotor drone nest at the jth candidate address, and z j is a Boolean variable. If zj = 0, it means that the multi-rotor drone nest will not be built at the jth candidate address. j =1, it means that a multi-rotor drone nest is built at the jth candidate address, T is the total running time of the inspection system including various types of drones, and f i is the inspection service demand frequency of the i-th target inspection equipment, α is the flight cost per unit distance of the UAV, d ij is the flight distance between the i-th target inspection device and the j-th candidate address, Y ij is the correspondence between the fixed-wing UAV in the jth candidate address and the i-th target inspection equipment, and Y ij is a Boolean variable. If Y ij = 0, it means that there is no corresponding relationship between the fixed-wing UAV in the j candidate addresses and the i-th target inspection equipment. ij =1, it means that there is a corresponding relationship between the fixed-wing UAV in the j candidate addresses and the i-th target inspection equipment, P fail is the probability of failure of the multi-rotor drone nest, Z ij is the correspondence between the multi-rotor drone in the jth candidate address and the i-th target inspection equipment, and Z ij is a Boolean variable. If Z ij = 0, it means that there is no corresponding relationship between the multi-rotor drone in the j candidate addresses and the i-th target inspection equipment. ij =1, it means that there is a corresponding relationship between the multi-rotor drone in the j candidate addresses and the i-th target inspection equipment. β is the penalty cost per unit distance of manual inspection when the multi-rotor drone nest fails. is the penalty distance when the i-th target inspection equipment is not covered by the multi-rotor drone nest.

[0082] It can be understood that the second objective function is to minimize the total cost, which includes the construction cost of the two machine nests, the total inspection cost of fixed-wing UAVs, and the total inspection cost of multi-rotor UAVs. Among them, the total inspection cost of multi-rotor UAVs includes the total cost of normal inspection of multi-rotor UAVs and the total cost of manual supplementary inspection of multi-rotor UAVs in the event of machine nest failure.

[0083] Furthermore, the second constraint condition is specifically expressed as:

[0084]

[0085] Among them, J * is the candidate address set for building the machine nest, I is the set of all target inspection devices in the inspection area, d ij is the flight distance between the i-th target inspection device and the j-th candidate address, Y ijis the correspondence between the fixed-wing UAV in the jth candidate address and the i-th target inspection equipment, and Y ij is a Boolean variable, Z ij is the correspondence between the multi-rotor drone in the jth candidate address and the i-th target inspection equipment, and Z ij is a Boolean variable, y j is the result of building a fixed-wing drone nest at the jth candidate address, and y j is a Boolean variable, z j is the result of building a multi-rotor drone nest at the jth candidate address, and z j is a Boolean variable, R is the maximum service radius of the candidate address, N1 is the maximum number of candidate addresses for fixed-wing UAVs, and N0 is the maximum number of candidate addresses for multi-rotor UAVs.

[0086] It can be understood that the second constraint condition limits the number of fixed-wing UAV nests, the number of multi-rotor UAV nests, the correspondence between target inspection equipment and candidate addresses, and the service radius of candidate addresses.

[0087] S104: Based on the second objective function and the second constraint condition, a meta-heuristic optimization algorithm is used to search for the best among the candidate address sets to obtain a set of target nest locations corresponding to each type of UAV.

[0088] Specifically, refer to Figure 2 , based on the model operator of changing the type of drone at the candidate address, change the type of drone at the candidate address in the candidate address set, perform optimization iterations, and if the number of iterations reaches the preset first iteration threshold, obtain the first optimization set. Based on the first optimization set, combine the relational operator and the hybrid operator to change the type of drone at the candidate address and / or adjust the correspondence between the candidate address and the target inspection equipment, perform optimization iterations, and obtain the second optimization set, wherein the relational operator is used to adjust the correspondence between the candidate address and the target inspection equipment, and the hybrid operator includes the model operator and the relational operator. Based on the relational operator, adjust the correspondence between the candidate address and the target inspection equipment in the second optimization set, perform optimization iterations, and obtain the target machine nest position set.

[0089] Further, the random number is compared with a preset optimization threshold to obtain a comparison result. An operator corresponding to the comparison result is determined, and the first optimization set is optimized based on the operator corresponding to the comparison result. If the number of iterations reaches a preset second iteration threshold, a second optimization set is obtained.

[0090] Furthermore, the second optimization set is iterated for the mth time, and the corresponding relationship between the candidate addresses and the target inspection equipment is adjusted to obtain the iterative result set Q m , where m is a positive integer and m is greater than 1. According to Q mThe corresponding second objective function value and Q m-1 The difference between the corresponding second objective function values ​​determines the current optimal result set and updates the iteration parameter T corresponding to the iteration result set m,sa When T m,sa When m is greater than a preset temperature threshold or m is greater than a preset maximum iteration threshold, the current optimal result set is used as the target machine nest position set.

[0091] T m,sa is the iteration parameter at the mth iteration. The initial value of the above iteration parameter is set according to the actual situation. At the same time, the iteration coefficient γ of the iteration parameter is set. After each iteration, the iteration parameter T is updated according to the iteration coefficient. m,sa =T m-1,sa ×γ.Q m The corresponding second objective function value is obtained by iterating the result set Q m Solve the second objective function to obtain .

[0092] In a specific implementation, the present invention proposes an iterative local search (ILS)-simulated annealing (SA) algorithm, which sets a set of neighborhoods, each of which represents a new set of solutions obtained by performing a specific operation on the current solution. By iteratively selecting neighborhoods and combining the acceptance criteria of simulated annealing, the quality of the solution is continuously improved.

[0093] The above fields include model operators, relational operators, and hybrid operators. Model operators are used to change the type of drone corresponding to a candidate address. The specific process is: randomly select a candidate address from the candidate address set. If the drone type of the current candidate address is a fixed-wing drone, it will be replaced with a multi-rotor drone. Conversely, if the drone type of the current candidate address is a multi-rotor drone, it will be replaced with a fixed-wing drone. Figure 3 , the square represents a type of drone at the candidate address, and the triangle represents a type of drone at the candidate address. Adjust the square to a triangle to change the type of drone. The relational operator is used to change the correspondence between the candidate address and the target inspection device. The specific process is: select the target inspection devices that fall within the coverage range of two or more candidate addresses from the target inspection devices in the inspection area, denoted as set I', and the number of elements in the set is represented by |I'|. Randomly generate a natural number Num∈[1,|I'|], and randomly select Num target inspection devices from I', and randomly redistribute the target inspection devices to the candidate addresses that meet the coverage requirements. Reference Figure 4, add the target inspection devices in the intersection of any two dotted circles to the set I', adjust the target inspection devices connected to the triangle to the target inspection devices connected to the square, and change the corresponding relationship between the candidate address and the target inspection device. The hybrid operator is a combination of the model operator and the relational operator. The hybrid operator includes the operation of the model operator once and the operation of the relational operator once.

[0094] In a specific example, since the construction cost of the machine nest accounts for a large proportion of the total cost, the machine model operator is applied to the early stage of iteration, that is, when the number of iterations is less than 10% of the maximum number of iterations (i.e., the first iteration threshold), the machine model operator is used to generate a new solution to obtain the first optimization set. The relational operator and the hybrid operator are applied to the middle stage of iteration, that is, when the number of iterations is less than 50% of the maximum number of iterations (i.e., the second iteration threshold), by using the random number generated by the random function to compare with the preset optimization threshold, the relational operator and the hybrid operator are randomly used to expand the search space and improve the quality of the solution. The relational operator is applied to the late stage of iteration, that is, when the number of iterations is less than the maximum number of iterations, the relational operator is continuously used to optimize the matching relationship between the candidate address and the target inspection equipment. In other embodiments, the use stage and proportion of the machine model operator, the relational operator and the hybrid operator are adjusted according to actual needs, and there is no limitation on this.

[0095] Since the ILS algorithm can only accept solutions that are better than the current optimal solution, it is easy to fall into the local minimum, so the acceptance criterion of simulated annealing is used. The specific process includes: t Better than the current optimal solution * When s * Updated to t Otherwise, the acceptance criterion will accept the worse iterative solution s with a certain probability t , will s * Updated to t For example, when the current optimal solution s * Better than iterative solution t When using random functions and iterative solutions s t Compare the probability of the optimal solution and choose whether to accept the worse iterative solution s according to the comparison result t ,in, ΔQ is Q m The corresponding second objective function value and Q m-1 The difference between the corresponding second objective function values. If the random number obtained by the random function is less than the iterative solution s t is the probability of the optimal solution, then accept the worse iterative solution s t , which will be * Updated to t .

[0096] In a specific example, referring to Figure 5 ,A transmission tower area of ​​approximately 400 square kilometers was selected as the inspection area for algorithm verification and scenario comparison.

[0097] The present invention considers establishing several drone nests in the above inspection area to serve 2062 power equipment that need to be inspected, i.e., target inspection equipment. Each drone nest can serve all the towers to be inspected within a radius of 3km. The goal is to select the minimum number of towers and establish drone nests at their locations so that all the towers to be inspected can be covered by at least one drone nest, thereby achieving the lowest construction cost. Figure 6 ,Through the solution of Gurobi optimizer, 17 optimal nest site selection points, i.e., candidate address sets, are obtained. Among them, the tower to be inspected, i.e., the target inspection equipment, is represented by a dot, and the nest site selection point, i.e., the candidate address, is represented by a cross, with a total of 17 candidate addresses.

[0098] Reference Figure 7 ,Through the present invention, a set of optimal UAV nest deployment and task allocation schemes are obtained. Figure 7 The drone types and service areas of each drone nest are shown. The points and lines corresponding to different numbers represent different nests and the target inspection equipment they cover. Fixed-wing drone nests and their corresponding mission points are marked with squares, and multi-rotor drone nests are marked with triangles. Figure 7 The total cost of this solution is 1,684,373 yuan, including 13 fixed-wing drone nests and 4 multi-rotor drone nests. Figure 7 ,It can be found that the locations of the drone nests are ,distributed relatively evenly, and the task allocation scheme ,effectively covers all the task points.

[0099] The performance comparison between the scheme of the present invention and the greedy algorithm in solving the problem of drone nest site selection and task allocation is shown in Table 1. The main indicators for comparison include total cost, number of nests used, and proportion of fixed-wing nests.

[0100] Table 1 Comparison results between ILS-SA and greedy algorithm

[0101]

[0102] Reference Figure 8 The embodiment of the present invention provides a multi-type UAV nest site selection device based on inspection task allocation, including:

[0103] The first condition building module 201 is used to provide a first objective function corresponding to the number of drone nests and a first constraint condition corresponding to the location of the drone nest based on the inspection task;

[0104] A candidate address determination module 202, configured to obtain a candidate address set based on a first objective function and a first constraint condition;

[0105] The second condition construction module 203 is used to construct a second objective function and a second constraint condition based on the drone type decision and the inspection task according to the inspection cost, inspection range and inspection area of ​​each type of drone;

[0106] The target location determination module 204 is used to search for the best among the candidate address sets based on the second objective function and the second constraint condition through a meta-heuristic optimization algorithm to obtain a target nest location set corresponding to each type of UAV.

[0107] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0108] Although preferred embodiments of the present invention have been described, additional changes and modifications may be made to these embodiments by those skilled in the art once the basic inventive concepts are known. 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. Obviously, those skilled in the art may 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 selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation, characterized in that: include: Based on the inspection task, the first objective function corresponding to the number of drone nests and the first constraint condition corresponding to the location of the drone nest are given respectively; Based on the first objective function and the first constraint condition, a candidate address set is obtained; According to the inspection cost, inspection scope and inspection area of ​​each type of drone, a second objective function and a second constraint condition based on drone type decision and inspection task are constructed; Based on the second objective function and the second constraint, a meta-heuristic optimization algorithm is used to optimize the candidate address set and obtain the target nest location set corresponding to each type of UAV.

2. The method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation according to claim 1, characterized in that: The first objective function is specifically expressed as: The first constraint is specifically expressed as: Among them, x j is the result of establishing a nest at the jth pending address, and x j is a Boolean variable, J is the set of pending addresses in the inspection area, X ij is the relationship between the distance between the jth pending address and the ith target inspection device and the inspection range of the drone, and X ij is a Boolean variable, and I is the set of all target inspection devices in the inspection area.

3. The method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation according to claim 1, characterized in that: Based on the first objective function and the first constraint condition, a candidate address set is obtained, including: Based on the first objective function and the first constraint condition, a set of pending addresses is obtained; Based on the pending address set, a corresponding relationship is established between each target inspection device in the inspection area and the nearest pending address to obtain a candidate address set.

4. The method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation according to claim 1, characterized in that: Inspection costs include construction costs, flight costs, and penalty costs; According to the inspection cost, inspection scope and inspection area of ​​each type of drone, the second objective function and second constraint conditions based on drone type decision and inspection task are constructed, including: Obtain the construction cost, flight cost and penalty cost of each type of drone, combine the correspondence between each type of drone and the target inspection equipment and the flight distance, analyze the probability of failure of each type of drone nest, and construct the second objective function; The inspection range of each type of UAV and the flight distance between each type of UAV and the target inspection equipment are obtained, the corresponding relationship between each type of UAV and the target inspection equipment is analyzed, and the second constraint condition is constructed based on the number of each type of UAV.

5. The method for selecting a site for a multi-type drone nest based on inspection task allocation according to claim 4 is characterized in that: The types of drones include fixed-wing drones and multi-rotor drones; the second objective function is specifically expressed as: Among them, J * is the candidate address set for building the nest, I is the set of all target inspection equipment in the inspection area, c1 is the construction cost of the fixed-wing UAV nest, y j is the result of building a fixed-wing drone nest at the jth candidate address, and y j is a Boolean variable, c0 is the construction cost of the multi-rotor drone nest, z j is the result of building a multi-rotor drone nest at the jth candidate address, and z j is a Boolean variable, T is the total running time of the inspection system including all types of drones, and f i is the inspection service demand frequency of the i-th target inspection equipment, α is the flight cost per unit distance of the UAV, d ij is the flight distance between the i-th target inspection device and the j-th candidate address, Y ij is the correspondence between the fixed-wing UAV in the jth candidate address and the i-th target inspection equipment, and Y ij is a Boolean variable, P fail is the probability of failure of the multi-rotor drone nest, Z ij is the correspondence between the multi-rotor drone in the jth candidate address and the i-th target inspection equipment, and Z ij is a Boolean variable, β is the penalty cost per unit distance of manual inspection when the multi-rotor drone nest fails, is the penalty distance when the i-th target inspection equipment is not covered by the multi-rotor drone nest.

6. The method for selecting a site for a multi-type drone nest based on inspection task allocation according to claim 4, characterized in that: The types of drones include fixed-wing drones and multi-rotor drones; the second constraint is specifically expressed as: Among them, J * is the candidate address set for building the machine nest, I is the set of all target inspection devices in the inspection area, d ij is the flight distance between the i-th target inspection device and the j-th candidate address, Y ij is the correspondence between the fixed-wing UAV in the jth candidate address and the i-th target inspection equipment, and Y ij is a Boolean variable, Z ij is the correspondence between the multi-rotor drone in the jth candidate address and the i-th target inspection equipment, and Z ij is a Boolean variable, y j is the result of building a fixed-wing drone nest at the jth candidate address, and y j is a Boolean variable, z j is the result of building a multi-rotor drone nest at the jth candidate address, and z j is a Boolean variable, R is the maximum service radius of the candidate address, N1 is the maximum number of candidate addresses for fixed-wing UAVs, and N0 is the maximum number of candidate addresses for multi-rotor UAVs.

7. The method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation according to claim 1, characterized in that: Based on the second objective function and the second constraint condition, the meta-heuristic optimization algorithm is used to search for the best solution in the candidate address set to obtain the target nest location set corresponding to each type of drone, including: Based on the model operator for changing the type of the drone at the candidate address, the type of the drone at the candidate address in the candidate address set is changed, and optimization iteration is performed to obtain the first optimization set; Based on the first optimization set, the relational operator and the hybrid operator are combined to change the drone type of the candidate address and / or adjust the corresponding relationship between the candidate address and the target inspection equipment, and perform optimization iteration to obtain a second optimization set, wherein the relational operator is used to adjust the corresponding relationship between the candidate address and the target inspection equipment, and the hybrid operator includes the model operator and the relational operator; Based on the relational operator, the corresponding relationship between the candidate addresses and the target inspection equipment in the second optimization set is adjusted, and optimization iteration is performed to obtain the target machine nest position set.

8. The method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation according to claim 7, characterized in that: Based on the first optimization set, combined with the relational operator and the hybrid operator, the drone type of the candidate address is changed and / or the corresponding relationship between the candidate address and the target inspection equipment is adjusted, and the optimization iteration is performed to obtain the second optimization set, which specifically includes: Compare the random number with the preset optimization threshold to obtain a comparison result; Determine an operator corresponding to the comparison result, and optimize the first optimization set based on the operator corresponding to the comparison result; If the number of iterations reaches a preset second iteration threshold, a second optimization set is obtained.

9. The method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation according to claim 7, characterized in that: Based on the relational operator, the corresponding relationship between the candidate addresses and the target inspection equipment in the second optimization set is adjusted, and optimization iteration is performed to obtain the target machine nest location set, which specifically includes: Perform the mth iteration on the second optimization set, adjust the correspondence between the candidate addresses and the target inspection equipment, and obtain the iterative result set Q m , where m is a positive integer and m is greater than 1; According to Q m The corresponding second objective function value and Q m-1 The difference between the corresponding second objective function values ​​determines the current optimal result set and updates the iteration parameter T corresponding to the iteration result set m,sa ; When T m,sa When m is greater than a preset temperature threshold or m is greater than a preset maximum iteration threshold, the current optimal result set is used as the target machine nest position set.

10. A multi-type drone nest site selection device based on inspection task allocation, characterized in that: The method for selecting a nest site for multiple types of unmanned aerial vehicles based on inspection task allocation as described in any one of claims 1 to 9 comprises: A first condition building module is used to provide a first objective function corresponding to the number of drone nests and a first constraint condition corresponding to the location of the drone nest based on the inspection task; A candidate address determination module, configured to obtain a candidate address set based on a first objective function and a first constraint condition; The second condition construction module is used to construct a second objective function and a second constraint condition based on the drone type decision and the inspection task according to the inspection cost, inspection scope and inspection area of ​​each type of drone; The target location determination module is used to search for the best among the candidate address sets based on the second objective function and the second constraint condition through a meta-heuristic optimization algorithm to obtain a target nest location set corresponding to each type of UAV.

Images