Line planning method, system and medium for connecting centralized vehicle charging stations to distribution networks
Through the improved artificial potential field method and RRT* algorithm, combined with simulated annealing and complete path planning algorithm, the problem of obstacle impact in power line planning is solved, and the power line path optimization is achieved that is more in line with actual construction, ensuring the power line Feasibility and minimize length.
Patent Information
- Application Number
- CN202210213218.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-03-04
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2042-03-04
AI Technical Summary
The existing power line planning method for connecting the electric vehicle charging station special transformer to the distribution network is to minimize the straight line distance from the charging station to the switch cabinet. It is impossible to consider obstacles in the laying of actual power lines, resulting in the optimization plan not in line with the actual situation. Construction cost.
The improved artificial potential field method and RRT* algorithm are used, combined with simulated annealing algorithm and complete path planning algorithm, and the optimization of power line paths by obstacles is considered, and converted into path optimization problems to minimize the power line path length.
The resulting power line laying length is more in line with the actual construction length, providing an optimization foundation for charging stations to access the distribution network, ensuring the feasibility of power line planning and minimizing path length.
Smart Images

Figure CN114610026B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of charging station access to a distribution network, and in particular relates to a line planning method, system and medium for centralized automobile charging station access to a distribution network. Background Art
[0002] The speed of electric vehicle charging affects the speed of electric vehicle popularization. With the development of electric vehicle charging technology, the power of charging piles has been continuously improved, and the capacity of public fast charging stations has also been continuously improved. Taking the 200kW Tesla Supercharger as an example, the total capacity of an eight-pile charging station reaches 1600kW. This large-capacity centralized electric vehicle charging station needs to be connected to the 10kV medium-voltage distribution network through a public switch station via a 10kV dedicated transformer. Therefore, it is necessary to plan the path for the centralized electric vehicle charging station to access the power line of the distribution network.
[0003] The existing power line planning method for connecting the dedicated transformer of the electric vehicle charging station to the distribution network is to connect to the public medium-voltage switchgear with the closest straight-line distance required by the load capacity constraint. The actual laying path of the power line is adjusted by the construction unit during the specific construction.
[0004] The traditional power line planning method aims to minimize the straight-line distance from the charging station to the switch cabinet, but fails to take into account obstacles such as buildings and rivers encountered during the actual laying of power lines. The straight-line distance cannot truly reflect the actual length of the power line that needs to be laid. Therefore, the optimized plan is not the one with the lowest actual construction cost. Summary of the invention
[0005] The main purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art and to provide a line planning method, system and medium for connecting a centralized vehicle charging station to a distribution network.
[0006] In order to achieve the above object, the present invention adopts the following technical solutions:
[0007] One aspect of the present invention provides a method for planning a line for connecting a centralized vehicle charging station to a distribution network, comprising the following steps:
[0008] Set the location of the charging station and the switch station, and obtain the obstacle coordinate set based on the building block GIS information;
[0009] The charging station location is set as the starting point, the switch station location is set as the end point, and the path planning is performed based on the improved artificial potential field method;
[0010] Determine whether the planned path connects the switch station and the charging station: if so, straighten the search point set to obtain the final planned path; if not, set the switch station location as the starting point and the charging station location as the end point, and perform path planning based on the improved artificial potential field method;
[0011] Determine whether the planned path connects the switch station and the charging station: if so, merge the search point sets of the two artificial potential field methods, and straighten the search point sets to obtain the final planned path; if not, use the local optimal solutions of the two artificial potential fields as the starting point and end point, perform path search based on the improved RRT*, merge the search point sets of the three searches, and straighten the search point sets to obtain the final planned path.
[0012] As a preferred technical solution, the improved artificial potential field method is specifically:
[0013] Set the initial target, initial position, and model obstacles;
[0014] Calculate the potential function based on the repulsive field and the gravitational field, and calculate the moving path of the intelligent agent;
[0015] Determine whether the moving distance each time is too short or the resultant force is zero. If so, execute the simulated annealing algorithm to move to the new position and proceed to the next step; otherwise, proceed to the next step;
[0016] Determine whether the initial target has been reached. If so, end the algorithm; if not, move to the next point according to the moving path and jump to the step of calculating the potential function and calculating the moving path of the agent.
[0017] As a preferred technical solution, the potential function of the gravitational field is specifically:
[0018]
[0019] Among them, U req (q) is the potential function, U att (q) is the gravitational potential function determined by the target point, ζ is the gravitational gain, is the set threshold, d(q,q goal ) is the distance between the agent and the target point. When d(q,q goal ) is less than or equal to d, the gravitational force is proportional to the square of the distance; when the distance between the agent and the target point is greater than d, the gravitational force is regarded as a constant.
[0020] As a preferred technical solution, the simulated annealing algorithm adds disturbance to the current optimal solution to generate a new solution, and accepts the new solution with a certain probability, including the following steps:
[0021] Given an initial temperature T and initial state X, and define the annealing rate;
[0022] Add a uniformly distributed random perturbation ΔX that is much smaller than the initial state X to the initial state X, and get X i =X+ΔX, and calculate the potential field strength U(X i ) and the difference ΔU between the potential field strength U(X) in the initial state;
[0023] When ΔU>0, accept X i is the new state; when U>0, with probability Accept X i is the new state, where k is the Boltzmann constant;
[0024] Determine whether to escape from the local minimum at this time: if so, terminate the simulated annealing algorithm and move to the new position X i ; If not, lower the temperature T and jump to the step of adding random perturbation ΔX.
[0025] As a preferred technical solution, the improved RRT* is specifically:
[0026] Set the starting point to the iteration starting point of the growing tree;
[0027] Generate a random point in the agent’s modeling space;
[0028] Find the point closest to the random point on the growth tree as the growth point;
[0029] Determine whether an obstacle is encountered after growing a unit step: If not, add the grown branches and endpoints to the growth tree and jump to the step of generating a random point in the modeling space of the agent; if yes, proceed to the next step;
[0030] Determine whether obstacles are encountered continuously: if not, jump to the step of generating a random point in the agent's modeling space; if yes, proceed to the next step;
[0031] Determine whether to reduce the step size multiple times: if not, reduce the search step size and jump to the step of generating a random point in the modeling space of the agent; if yes, proceed to the next step;
[0032] Reset the search step size, take the current point as the starting point, change the search direction determination method, and continuously search the iterative growth tree N times, jumping to the step of generating a random point in the modeling space of the agent.
[0033] As a preferred technical solution, the method for changing the search direction is specifically as follows:
[0034] First draw the search point distribution radius based on the search step length;
[0035] All points on the distribution radius that are connected to the agent and do not overlap with obstacles are taken as candidate search points, and this fan-shaped area is the search direction.
[0036] As a preferred technical solution, a complete path planning algorithm is used to replace the improved RRT* for path search; the complete path planning algorithm includes the A* algorithm and the Dijkstra algorithm.
[0037] As a preferred technical solution, the straightening of the search point set is specifically as follows:
[0038] Add the starting point to the straightened point set, and start judging from the second point in the search point set;
[0039] Determine whether the straight line formed by the last point in the straightened point set and the current point overlaps with the obstacle: if not, continue to perform obstacle overlap determination on the next point in the search point set; if yes, add the previous point of the current point to the straightened point set;
[0040] Determine whether the current point is the last point in the search point set: if not, continue to perform obstacle overlap judgment on the next point in the search point set; if so, connect the points in the straightened point set to obtain the final planned path.
[0041] Another aspect of the present invention provides a line planning system for accessing a centralized vehicle charging station to a distribution network, which is applied to the above-mentioned line planning method for accessing a centralized vehicle charging station to a distribution network, and includes a preprocessing module, an artificial potential field planning module, an improved RRT* search module, and a straightening module;
[0042] The pre-processing module is used to set the charging station location and the switch station location, and obtain the obstacle coordinate set according to the building block GIS information;
[0043] The artificial potential field planning module is used to plan a path based on the improved artificial potential field method with the charging station location as the starting point and the switch station location as the end point; the path is planned based on the improved artificial potential field method with the switch station location as the starting point and the charging station location as the end point;
[0044] The improved RRT* search module is used to perform path search based on the improved RRT* with the local optimal solutions of the two artificial potential fields as the starting point and the end point;
[0045] The straightening module is used to straighten the search point set to obtain the final planned path.
[0046] Another aspect of the present invention provides a storage medium storing a program, which, when executed by a processor, implements the above-mentioned line planning method for connecting a centralized vehicle charging station to a distribution network.
[0047] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0048] (1) The present invention converts the power line planning problem of connecting the dedicated transformer of the electric vehicle centralized charging station with the public switch cabinet into a path optimization problem, and takes into account the method of optimizing the power line path due to obstacles such as buildings and rivers. The obtained power line laying length is more in line with the actual construction length, providing a basis for optimizing the access of charging stations to the distribution network;
[0049] (2) The present invention improves the RRT algorithm based on the special shape of the building block to adapt to the power line path planning for the charging station to access the distribution network;
[0050] (3) The present invention is based on an improved artificial potential field and RRT joint path optimization method, which ensures that a feasible solution for power line planning is obtained under the goal of minimizing the power line path length. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] Figure 1 is a flow chart of a method for planning a line for connecting a centralized vehicle charging station to a distribution network according to an embodiment of the present invention;
[0052] Figure 2 is a flow chart of an improved artificial potential field method according to an embodiment of the present invention;
[0053] Figure 3 is a schematic diagram of an intelligent agent according to an embodiment of the present invention moving from a high point to a low point in a potential field;
[0054] Figure 4 is a flow chart of an improved RRT algorithm according to an embodiment of the present invention;
[0055] Figure 5 is a schematic diagram of an intelligent agent in an embodiment of the present invention when it is in a corner of a narrow and long area;
[0056] Figure 6 is a schematic diagram of an intelligent agent according to an embodiment of the present invention reducing the step length to jump out of a narrow and long area;
[0057] Figure 7 is a schematic diagram of an improved search direction of an intelligent agent according to an embodiment of the present invention;
[0058] Figure 8 is a schematic diagram of an intelligent agent jumping out of a narrow and long closed area according to an embodiment of the present invention;
[0059] Fig. 9 is a flow chart of straightening search points according to an embodiment of the present invention;
[0060] Fig.10 It is a structural schematic diagram of a line planning system for connecting a centralized vehicle charging station to a distribution network according to an embodiment of the present invention;
[0061] Fig.11 It is a schematic diagram of the structure of the storage medium of an embodiment of the present invention.
[0062] Explanation of the accompanying figures: 1. Intelligent agent; 2. Obstacle; 3. Random points generated by RRT; 4. Search direction of the intelligent agent based on the random points; 5. Search point distribution radius under larger step size; 6. Search point distribution under larger step size; 7. Search point distribution radius under smaller step size; 8. Search point distribution under smaller step size; 9. Search direction range determined by larger step size; 10. Search point distribution radius determined by larger step size; 11. Search direction range determined by smaller step size; 12. Search point distribution radius determined by smaller step size; 13. Search point range generated by RRT; 14. Search point distribution radius. DETAILED DESCRIPTION
[0063] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present application.
[0064] Example
[0065] like Figure 1 As shown, this embodiment provides a method for planning a line for connecting a centralized vehicle charging station to a distribution network, comprising the following steps:
[0066] (1) Set the locations of charging stations and switch stations, and obtain the obstacle coordinate set based on the building block GIS information;
[0067] (2) The charging station location is set as the starting point and the switch station location is set as the end point, and the path planning is performed based on the improved artificial potential field method;
[0068] (3) Determine whether the planned path connects the switch station and the charging station (global optimality). If so, jump to step (9); if not, jump to step (4);
[0069] (4) The switch station location is set as the starting point and the charging station location as the end point, and the path planning is performed based on the improved artificial potential field method;
[0070] (5) Determine whether the planned path connects the switch station and the charging station (global optimality). If so, jump to step (8); if not, jump to step (6);
[0071] (6) The local optimal solutions of the two artificial potential fields are used as the starting point and the end point, and the path search is performed based on the improved RRT*;
[0072] (7) Merge the search point sets of the three searches and jump to step (9);
[0073] (8) Merge the search point sets of the two artificial potential field methods;
[0074] (9) Straighten the search point set to obtain the final planned path.
[0075] like Figure 2 As shown, the improved artificial potential field method is specifically as follows:
[0076] (2.1) Set the initial target, initial position, and model obstacles;
[0077] (2.2) Calculate the potential function based on the repulsive field and the gravitational field, and calculate the moving path of the intelligent agent;
[0078] (2.3) Determine whether the distance of each movement is too short or the resultant force is zero (entering the local optimum). If so, jump to step (2.4); otherwise, jump to step (2.5);
[0079] (2.4) Execute the simulated annealing algorithm to move to the new position;
[0080] (2.5) Determine whether the initial target has been reached. If so (global optimality), jump to step (2.6). If not, move to the next point according to the moving path and jump to step (2.2);
[0081] (2.6)End.
[0082] The traditional artificial potential field method has problems such as unreachable targets and local minima:
[0083] (1) It is easy to fall into a local minimum. When the gravitational force and repulsive force on the agent are equal in magnitude and opposite in direction, the resultant force on the agent is 0, and the agent is trapped in a local minimum. The resultant forces at various locations around the local minimum point all point to the local minimum, causing the agent to oscillate around the local minimum and be unable to leave the area on its own.
[0084] (2) The gravitational part of the artificial potential field method is proportional to the distance between the intelligent agent and the target point. When the intelligent agent is far away from the target point, the gravitational part is too large, which may cause the intelligent agent to collide with obstacles.
[0085] (3) When there are obstacles near the target point, the agent may not be able to reach the target point.
[0086] First of all, we need to improve the problem of excessive gravity in the traditional artificial potential field method. The gravity function of the traditional artificial potential field method is proportional to the distance between the agent and the target point. The size of the gravitational field is determined by the distance between the agent and the target point. When the agent is far away from the target point, the gravity will become very large, and the repulsion will become very small relative to the gravity, which will cause the agent to collide with obstacles. Therefore, it is necessary to improve the gravitational field:
[0087] In step (2.2), the potential function of the gravitational field is specifically:
[0088]
[0089] Among them, U req (q) is the potential function, U att (q) is the gravitational potential function determined by the target point, ζ is the gravitational gain, d is the set threshold, d(q,q goal ) is the distance between the agent and the target point. When d(q,q goal ) is less than or equal to d, the gravitational force is proportional to the square of the distance; when the distance between the agent and the target point is greater than d, the gravitational force is regarded as a constant.
[0090] In order to solve the problem of the artificial potential field method falling into the local optimum, the coefficient term is added and combined with the simulated annealing method to improve it. First, a coefficient term that changes with the distance of the target point is added, so that the algorithm will not be unable to reach the target point when the target point is close to the obstacle because the repulsion is greater than the attraction. Then, the simulated annealing method is used to add disturbances to the current optimal solution to generate a new solution, and the new solution is accepted with a certain probability, so that it has the ability to jump out of the local minimum, solving the problem that the artificial potential field method is prone to falling into the local minimum.
[0091] In step (2.4), the simulated annealing algorithm adds disturbance to the current optimal solution to generate a new solution, and accepts the new solution with a certain probability, including the following steps:
[0092] (2.4.1) Given an initial temperature T and initial state X, and defining the annealing rate;
[0093] (2.4.2) Add a uniformly distributed random perturbation ΔX that is much smaller than the initial state X to the initial state X, and we get X i =X+ΔX, and calculate the potential field strength U(X i ) and the difference ΔU between the potential field strength U(X) in the initial state;
[0094] (2.4.3) When ΔU>0, accept X i is the new state; when U>0, with probability Accept X i is the new state, where k is the Boltzmann constant;
[0095] (2.4.4) Determine whether the local minimum is escaped at this time. If yes, jump to step (2.4.5); if not, reduce the temperature T and jump to step (2.4.1);
[0096] (2.4.5) Terminate the simulated annealing algorithm and move to the new position X i .
[0097] Assume that the agent moves freely from the high point of the potential field to the low point, and its potential field height and moving path are as follows: Figure 3 As shown in the figure, when the agent reaches point A, since point A is a minimum value, the agent will be trapped at point A and cannot climb over the rising interval of the potential field, but the agent's goal is to reach the minimum point B. At this time, a search algorithm is needed to continue searching upwards when the agent is trapped at point A to help the agent cross this rising interval of the potential field, and simulated annealing will accept inferior solutions with probability. When the agent moves to point A, it is possible to search for point C through simulated annealing. Once point C is searched, simulated annealing will directly accept the best solution, that is, once point C is searched, point B can be directly searched, so the agent will choose to move to point B, which is a better solution than point A.
[0098] like Figure 4 As shown, in step (6), the improved RRT* is specifically:
[0099] (6.1) Set the starting point to the iteration starting point of the growth tree;
[0100] (6.2) Generate a random point in the agent’s modeling space;
[0101] (6.3) Find the point on the growth tree that is closest to the random point and use it as the growth point;
[0102] (6.4) Determine whether an obstacle is encountered after growing a unit step. If so, jump to step (6.5); if not, add the grown branches and endpoints to the growth tree and jump to step (6.2);
[0103] (6.5) Determine whether obstacles are encountered continuously. If yes, jump to step (6.6); if no, jump to step (6.2);
[0104] (6.6) Determine whether to reduce the step size multiple times. If so, jump to step (6.7); if not, reduce the search step size and jump to step (6.2);
[0105] (6.7) Reset the search step size, take the current point as the starting point, change the search direction determination method, and continuously search N times to the growth tree, and jump to step (6.2).
[0106] Sampling-based planning algorithms generally do not directly plan with the minimum grid resolution on a binary map. Instead, they use particles of a certain density to randomly scatter on the map to abstract the actual map to assist planning. For example, the PRM algorithm and its variants scatter points on the original map, extract the roadmap, and plan on such a topological map; RRT and its excellent variant RRT-connect randomly scatter a point on the map at each step, iteratively grow a tree, connect the start and end points, and finally plan on the connected graph.
[0107] The basic steps of RRT are:
[0108] 1. The starting point is used as a seed, from which the tree is grown iteratively;
[0109] 2. Generate a random point in the agent’s modeling space;
[0110] 3. Find the point on the tree that is closest to the random point as the growth point;
[0111] 4. Grow in the direction of the random point. If there is no obstacle after growing a unit step, add the grown branches and endpoints to the tree and return to step 2.
[0112] The random points 3 generated by RRT are generally evenly distributed in the entire modeling space, so the growth tree will grow approximately evenly in all directions, which can quickly explore the entire modeling space. However, when there are some concave obstacles 2 in the area, such as Figure 5 As shown in , a smaller step size is required to iterate the growth tree, but a smaller step size will increase the calculation time; at the same time, when the agent 1 is in a narrow area and in the corner of the entire area, the random points 3 generated by RRT are distributed in a large number of infeasible directions, that is, Figure 5 According to the random point search direction 4, agent 1 will not be able to jump out of the narrow area.
[0113] In view of the possible problems in the above power line planning, the basic RRT search method is improved and a method for jumping out of narrow areas is provided.
[0114] When the iteration of the growth tree continuously generates infeasible search points, the local optimal strategy is executed. First, the step size is reduced. When the step size is reduced, the search point generated is closer to the current position, and the probability of touching an obstacle is lower. Figure 6 As shown, compared with the search point distribution radius 5 under the larger step size and the search point distribution 6 under the larger step size, the search point distribution radius 7 under the smaller step size and the search point distribution 8 under the smaller step size are closer to the current position and the probability of touching an obstacle is lower.
[0115] After the step size is reduced, the generated search points do not touch obstacles and are all feasible search points. The growth tree can be further explored. However, since agent 1 is located in the corner of the modeling range, the search direction is mostly in one direction, such as Figure 6 The probability of searching the search point in the lower left direction is small, and reducing the step size still cannot jump out of the narrow area. Therefore, it is necessary to improve the determination of the search direction.
[0116] The iterative process of the RRT growth tree is essentially to determine the growth direction by random points, and the step size determines the search point distribution radius. When in a narrow and long area, the growth tree cannot continue to iterate due to the emergence of a large number of infeasible search directions. In order to allow the growth tree to continue to iterate, the search direction is no longer determined by random points, but the search point distribution radius is first drawn based on the search step size. Then, all points on the distribution radius that are connected to the agent and do not overlap with obstacles are taken as candidate search points, and this fan-shaped area is the search direction. In order to make the search direction reflect the characteristics of the narrow and long area, the search step size at this time should not be too small.
[0117] When the search step size is too small, such as Figure 7 As shown in FIG. 1 , the search direction determined by the search direction range 11 determined by the smaller step size and the search point distribution radius 12 determined by the smaller step size cannot reflect the characteristics of the narrow and long area, and the search point still cannot jump out of the narrow and long closed area. However, the search direction range 9 determined by the larger step size and the search point distribution radius 10 determined by the larger step size can reflect the characteristics of the narrow and long area, and the search point can jump out of the narrow and long closed area. The steps to improve the search direction to jump out of the narrow and long closed area are as follows: Figure 7 , Figure 8 As shown; 13 is the search point range generated by RRT, and 14 is the search point distribution radius.
[0118] By improving the RRT algorithm, its ability to jump out of narrow and long areas is increased, which can ensure that a feasible solution for power line path planning is obtained.
[0119] In this embodiment, in step (6), a complete path planning algorithm (such as A* algorithm, Dijkstra algorithm) can also be used instead of the improved RRT* to plan the path of the power line connecting the dedicated transformer of the charging station and the medium-voltage switchgear of the distribution network. The disadvantage is that the calculation complexity is relatively high.
[0120] like Fig. 9 As shown, in step (9), the straightening of the search point set is specifically as follows:
[0121] (9.1) Add the starting point to the straightened point set and start judging from the second point in the search point set;
[0122] (9.2) Determine whether the straight line formed by the last point in the straightened point set and the current point overlaps with the obstacle. If so, jump to step (9.3); if not, jump to step (9.2) to determine the next point in the search point set;
[0123] (9.3) Add the previous point of the current point to the straightened point set;
[0124] (94) Determine whether the current point is the last point. If so, jump to step (9.5); if not, jump to step (9.2) to determine the next point.
[0125] (9.5) Connect the points in the straightened point set to obtain the final planned path.
[0126] like Fig.10 As shown, in another embodiment of the present application, a line planning system for accessing a centralized vehicle charging station to a distribution network is provided, the system comprising a preprocessing module, an artificial potential field planning module, an improved RRT* search module, and a straightening module;
[0127] The pre-processing module is used to set the charging station location and the switch station location, and obtain the obstacle coordinate set according to the building block GIS information;
[0128] The artificial potential field planning module is used to plan a path based on the improved artificial potential field method with the charging station location as the starting point and the switch station location as the end point; the path is planned based on the improved artificial potential field method with the switch station location as the starting point and the charging station location as the end point;
[0129] The improved RRT* search module is used to perform path search based on the improved RRT* with the local optimal solutions of the two artificial potential fields as the starting point and the end point;
[0130] The straightening module is used to straighten the search point set to obtain the final planned path.
[0131] It should be noted here that the system provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure is divided into different functional modules to complete all or part of the functions described above. The system is a line planning method for the centralized vehicle charging station connected to the distribution network applied to the above embodiment.
[0132] like Fig.11 As shown, in another embodiment of the present application, a storage medium is further provided, storing a program, and when the program is executed by a processor, a line planning method for connecting a centralized vehicle charging station to a distribution network is implemented, specifically:
[0133] (1) Set the locations of charging stations and switch stations, and obtain the obstacle coordinate set based on the building block GIS information;
[0134] (2) The charging station location is set as the starting point and the switch station location is set as the end point, and the path planning is performed based on the improved artificial potential field method;
[0135] (3) Determine whether the planned path connects the switch station and the charging station (global optimality). If so, jump to step (9); if not, jump to step (4);
[0136] (4) The switch station location is set as the starting point and the charging station location as the end point, and the path planning is performed based on the improved artificial potential field method;
[0137] (5) Determine whether the planned path connects the switch station and the charging station (global optimality). If so, jump to step (8); if not, jump to step (6);
[0138] (6) The local optimal solutions of the two artificial potential fields are used as the starting point and the end point, and the path search is performed based on the improved RRT*;
[0139] (7) Merge the search point sets of the three searches and jump to step (9);
[0140] (8) Merge the search point sets of the two artificial potential field methods;
[0141] (9) Straighten the search point set to obtain the final planned path.
[0142] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0143] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.
Claims
1. A line planning method for connecting a centralized vehicle charging station to a distribution network, characterized in that: The steps include: Set the location of the charging station and the switch station, and obtain the obstacle coordinate set based on the building block GIS information; The charging station location is set as the starting point, the switch station location is set as the end point, and the path planning is performed based on the improved artificial potential field method; The improved artificial potential field method is specifically: Set the initial target, initial position, and model obstacles; Calculate the potential function based on the repulsive field and the gravitational field, and calculate the moving path of the intelligent agent; Determine whether the moving distance each time is too short or the resultant force is zero. If so, execute the simulated annealing algorithm to move to the new position and proceed to the next step; otherwise, proceed to the next step; The simulated annealing algorithm adds disturbance to the current optimal solution to generate a new solution, and accepts the new solution with a certain probability, including the following steps: Given an initial temperature T and initial state X, and define the annealing rate; Add a uniformly distributed random perturbation ΔX that is much smaller than the initial state X to the initial state X, and get X i =X+ΔX, and calculate the potential field strength U(X i ) and the difference ΔU between the potential field strength U(X) in the initial state; When ΔU>0, accept X i is the new state; when U>0, with probability Accept X i is the new state, where k is the Boltzmann constant; Determine whether to escape from the local minimum at this time: if so, terminate the simulated annealing algorithm and move to the new position X i ; If not, lower the temperature T and jump to the step of adding random perturbation ΔX; Determine whether the initial target has been reached. If so, end the algorithm; if not, move to the next point according to the moving path and jump to the step of calculating the potential function and calculating the moving path of the agent; Determine whether the planned path connects the switch station and the charging station: if so, straighten the search point set to obtain the final planned path; if not, set the switch station location as the starting point and the charging station location as the end point, and perform path planning based on the improved artificial potential field method; Determine whether the planned path connects the switch station and the charging station: if so, merge the search point sets of the two artificial potential field methods, and straighten the search point sets to obtain the final planned path; if not, use the local optimal solutions of the two artificial potential fields as the starting point and end point, perform path search based on the improved RRT*, merge the search point sets of the three searches, and straighten the search point sets to obtain the final planned path.
2. The line planning method for connecting a centralized vehicle charging station to a distribution network according to claim 1, characterized in that: The potential function of the gravitational field is specifically: Among them, U req (q) is the potential function, U att (q) is the gravitational potential function determined by the target point, ζ is the gravitational gain, d is the set threshold, d(q,q goal ) is the distance between the agent and the target point. When d(q,q goal ) is less than or equal to d, the gravitational force is proportional to the square of the distance; when the distance between the agent and the target point is greater than d, the gravitational force is regarded as a constant.
3. The line planning method for connecting a centralized vehicle charging station to a distribution network according to claim 1, characterized in that: The improved RRT* is specifically: Set the starting point to the iteration starting point of the growing tree; Generate a random point in the agent’s modeling space; Find the point closest to the random point on the growth tree as the growth point; Determine whether an obstacle is encountered after growing a unit step: If not, add the grown branches and endpoints to the growth tree and jump to the step of generating a random point in the modeling space of the agent; If yes, proceed to the next step; Determine whether obstacles are encountered continuously: If not, jump to the step of generating a random point in the modeling space of the agent; If yes, proceed to the next step; Determine whether to reduce the step size multiple times: If not, reduce the search step size and jump to the step of generating a random point in the modeling space of the agent; If yes, proceed to the next step; Reset the search step size, take the current point as the starting point, change the search direction determination method, and continuously search the iterative growth tree N times, jumping to the step of generating a random point in the modeling space of the agent.
4. The line planning method for connecting a centralized vehicle charging station to a distribution network according to claim 3 is characterized in that: The method for determining the change of the search direction is specifically as follows: First draw the search point distribution radius based on the search step length; All points on the distribution radius that do not overlap with obstacles and are connected to the agent are taken as candidate search points, and the fan-shaped area formed by the candidate search points and the agent's connection line is the search direction.
5. The line planning method for connecting a centralized vehicle charging station to a distribution network according to claim 1, characterized in that: A complete path planning algorithm is used to replace the improved RRT* for path search; the complete path planning algorithm includes an A* algorithm and a Dijkstra algorithm.
6. The line planning method for connecting a centralized vehicle charging station to a distribution network according to claim 1, characterized in that: The specific method of straightening the search point set is as follows: Add the starting point to the straightened point set, and start judging from the second point in the search point set; Determine whether the straight line formed by the last point in the straightened point set and the current point overlaps with the obstacle: if not, continue to perform obstacle overlap determination on the next point in the search point set; if yes, add the previous point of the current point to the straightened point set; Determine whether the current point is the last point in the search point set: if not, continue to perform obstacle overlap determination on the next point in the search point set; If so, connect the points in the straightened point set to obtain the final planned path.
7. A line planning system for connecting a centralized vehicle charging station to a distribution network, characterized in that: A line planning method for connecting a centralized vehicle charging station to a distribution network as described in any one of claims 1 to 6, comprising a preprocessing module, an artificial potential field planning module, an improved RRT* search module, and a straightening module; The pre-processing module is used to set the charging station location and the switch station location, and obtain the obstacle coordinate set according to the building block GIS information; The artificial potential field planning module is used to plan a path based on an improved artificial potential field method, taking the charging station location as the starting point and the switch station location as the end point; The switch station location is set as the starting point and the charging station location is set as the end point, and the path planning is performed based on the improved artificial potential field method; The improved RRT* search module is used to perform path search based on the improved RRT* with the local optimal solutions of the two artificial potential fields as the starting point and the end point; The straightening module is used to straighten the search point set to obtain the final planned path.
8. A storage medium storing a program, characterized in that: When the program is executed by a processor, the line planning method for connecting a centralized vehicle charging station to a distribution network as described in any one of claims 1 to 6 is implemented.
