A method for generating an optimal path for vehicle travel in a field environment

Abstract

The application discloses a kind of generation method of vehicle advancing optimal path under field environment, select field environment area, extract field obstacle information, establish field digital map containing soil characteristics and geomorphologic feature information;Based on vehicle ground mechanics theory, establish wheel and soil coupling mechanics model under field environment, deduce the motion equation of vehicle and soil coupling under field environment;Based on field environment digital map, generate multiple groups of paths randomly, embed wheel and soil coupling mechanics model under field environment, calculate initial automatic obstacle avoidance path;Establish path optimization algorithm, obtain optimal path in field environment digital map under field environment;The application proposes the method considering wheel and soil coupling effect, so as to calculate the mechanical properties and motion characteristics of vehicle in field environment, improve the authenticity of vehicle advancing optimal path generated under field environment, and the optimal path of vehicle in field environment can be quickly obtained, various obstacles can be successfully avoided.

Description

Technical Field

[0001] This invention belongs to the field of path planning technology, specifically relating to a method for generating the optimal path for vehicle travel in a field environment. Background Technology

[0002] The optimal route for vehicles in the field plays an important role in field travel, search and rescue, and transport of supplies. It also has broad prospects in working environments such as extraterrestrial landing, exploration, and data collection.

[0003] While research on vehicle trafficability and route planning under urban structured roads is relatively mature, the field environment is more complex, with various potholes, mud, and unknown threats that hinder vehicle movement. The complex environment causes difficulties and uncertainties for vehicle movement. Therefore, it is necessary to quickly analyze the field environment, determine efficient and safe field travel routes, and reduce the uncertainty of vehicle movement in the field.

[0004] Current methods for path planning in complex field environments rarely consider the coupled mechanical properties between wheels and soil. Therefore, determining the optimal path based on the distance required for passage does not take into account the actual field environment or threats. This will lead to problems such as low accessibility and poor accuracy of the predicted optimal path when applied in real complex field environments, making it difficult to adapt to field conditions. Summary of the Invention

[0005] This invention provides a method for generating optimal vehicle travel paths in field environments, embedding a wheel-soil coupled mechanical model into a path optimization algorithm. It accurately calculates safe and efficient optimal paths for complex field environments.

[0006] To address the aforementioned technical problems, this invention provides a method for generating the optimal vehicle travel path in a wilderness environment, comprising the following steps:

[0007] Step 100: Select a field environment area, extract field obstacle information, and create a field digital map containing soil properties and landform features;

[0008] Step 200: Based on the theory of vehicle ground mechanics, establish a mechanical model of wheel-soil coupling in the field environment, and derive the motion equations of vehicle-soil coupling in the field environment.

[0009] Step 300: Based on the aforementioned digital map of the wilderness environment, randomly generate multiple sets of paths, embed the wheel-soil coupling mechanical model in the wilderness environment, and calculate the initial automatic obstacle avoidance path;

[0010] Step 400: Establish a path optimization algorithm to obtain the optimal path in the digital map of the field environment.

[0011] As a preferred embodiment of the present invention, in step 100, based on the geographic information data of the current field environment area, the geographic information data is converted into a field environment digital map containing elevation, obstacle distribution and soil property information of different coordinate points using Matlab software.

[0012] Depending on the extent of the obstacle, the obstacle is simplified into circular and polygonal geometries.

[0013] As a preferred embodiment of the present invention, in step 200, the mechanical model of the coupling between the wheel and the soil is established based on the Lagrange equation;

[0014] Based on the coupled mechanical model of wheels and soil, the variation laws of subsidence, slip ratio, hook traction force and travel resistance parameters are obtained, and the motion equation of vehicle-soil coupling in the field environment is established.

[0015] In a preferred embodiment of the present invention, in step 300, the method for embedding the wheel-soil coupled mechanical model is as follows:

[0016] On the digital map of the field environment, the starting point and the ending point are fixed, and multiple sets of coordinate points within the coordinate range of the starting point and the ending point are randomly generated. All coordinate points are integrated to form a generated path, wherein the coordinate points represent the x-coordinate and y-coordinate of the points on the path, respectively.

[0017] Using the motion equations of vehicle-soil coupling in a field environment, the time required for a vehicle to travel on the generated path is calculated.

[0018] Calculate the distance from each coordinate point on the generated path to the starting point coordinates, sort each generated path in order of distance, and then form an initial path population from the N generated paths.

[0019] In a preferred embodiment of the present invention, in step 300, after generating the initial path population, iteratively determining whether the generated path individuals within the initial path population intersect with the obstacles, until a path capable of automatic obstacle avoidance is generated. Specifically, this is implemented as follows:

[0020] Based on the coordinate information of obstacles and soil, the obstacle area and soft soil area are divided into zones in the digital map of the field environment; for the obstacle area, its accessibility is set to 0.

[0021] When generating a path, it is determined whether the path intersects with an obstacle area. If they intersect, the path is removed and a new path is generated.

[0022] As a preferred embodiment of the present invention, the method for determining whether the generated paths in the initial path population intersect with the obstacles is as follows:

[0023] For each generated path, take 50 points from each path segment and determine whether each point falls in the middle of a circular obstacle. At the same time, use the vector cross product method to determine whether each point falls in the middle of a rectangular obstacle.

[0024] As a preferred embodiment of the present invention, in step 400, the implementation method of establishing a path optimization algorithm to obtain the optimal path in the digital map of the field environment is as follows:

[0025] Based on the dynamic model, the motion equation of the vehicle in the field environment is obtained, and the travel time of the vehicle on each generated path in the initial path population is calculated.

[0026] A fitness function is established based on the travel time to find the minimum travel time. Therefore, the fitness function is as follows:

[0027]

[0028] Where F(x) i Let t be the fitness of the i-th generated path. i The time required for vehicles to pass through.

[0029] Based on the roulette wheel selection method, N generated paths with high fitness are selected, and crossover and mutation operations are performed to form new N generated paths.

[0030] As a preferred embodiment of the present invention, the selection method based on the roulette wheel selection method is as follows:

[0031] The probability of each generated path in the initial path population being selected is calculated. Specifically, the selection probability of the i-th generated path is calculated based on the fitness function as follows:

[0032]

[0033] Where p(x) i Let F(x) be the probability that the i-th generated path is selected. i ) represents the fitness of the i-th generation path, and N represents the population size;

[0034] Calculate the cumulative probability of each generated path, then the cumulative probability of the i-th generated path is:

[0035]

[0036] Where, q i The cumulative probability of the i-th generated path is the sum of the probabilities of all generated paths preceding the i-th generated path.

[0037] A random number r is generated in the interval [0, 1]. The random number r is then compared with the cumulative probability q of the i-th generated path in the initial path population. i Compare them sequentially; if r > q i If the i-th generated path is skipped, it is compared with the next generated path until r ≤ q. j Then the j-th generated path will be selected.

[0038] Crossover and mutation operations are performed on the N generated paths to form N new child paths.

[0039] As a preferred embodiment of the present invention, the method for performing crossover and mutation operations on the N generated paths to form N new child paths in the roulette wheel selection method is as follows:

[0040] The crossover operation uses a two-point crossover method, with a crossover probability P set. c Generate a random number r in the range [0,1]. c For each of the N generation paths that serve as the parent, determine if r c <P c Then the i-th generated path is intersected with the (i+1)-th generated path;

[0041] The specific crossover operation is as follows: For the i-th generated path selected for crossover, two coordinate points are randomly selected, and the coordinates in the i-th generated path that are within the range of the two coordinate points are swapped with the (i+1)-th generated path;

[0042] The mutation operation uses a single-point mutation method, with a mutation probability P set. m Generate a random number r in the range [0, 1]. m For each of the N generation paths that serve as the parent, determine if r m <P m If the i-th generated path is mutated, then the i-th generated path will be mutated; otherwise, no mutation will occur.

[0043] The specific mutation operation is as follows: for the i-th generation path of the selected mutation, two mutation points are randomly generated, and then the coordinates of the two mutation points are reversed to obtain the offspring chromosome.

[0044] As a preferred embodiment of the present invention, the fitness of the generated N offspring chromosomes is calculated, and selection, crossover, and mutation operations are repeated on the N offspring chromosomes until the maximum number of generations is reached to obtain the optimal path.

[0045] Compared with the prior art, the present invention has the following advantages:

[0046] This invention proposes a method for calculating the mechanical properties of vehicles in a field environment. For complex real-world environments, it proposes a method that considers the coupling effect between the wheels and the soil, thereby calculating the mechanical properties and motion characteristics of vehicles in a field environment and improving the realism of the optimal vehicle travel path generated in a field environment.

[0047] This invention also proposes a method for calculating the optimal path of a vehicle in a field environment. By integrating the mechanical characteristics of the vehicle traveling in the field, the optimal path of the vehicle can be obtained quickly, enabling it to successfully avoid various obstacles. Attached Figure Description

[0048] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0049] Figure 1 This is a flowchart of finding the optimal travel path in the wild environment provided by an embodiment of the present invention;

[0050] Figure 2 This is a schematic diagram of a circular obstacle in a field environment provided by an embodiment of the present invention;

[0051] Figure 3 This is a diagram showing the optimal path result in a complex field environment provided by an embodiment of the present invention. Detailed Implementation

[0052] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] Path planning in the field is an important method to ensure the efficient, safe and rapid passage of equipment in complex field environments. However, current engineering calculations for path planning do not take into account the characteristics of the field soil and the mechanical behavior of vehicles. In order to solve the above problems, this invention enables vehicles to accurately avoid obstacles and travel efficiently in the field. The invention will be further described below with reference to the embodiments and accompanying drawings.

[0054] like Figure 1 As shown, this embodiment takes a vehicle traveling in a wilderness environment with multiple obstacles as a specific application background, and provides a method for generating the optimal travel path for a vehicle in a wilderness environment, including the following steps:

[0055] Step 100: Select a field environment area, extract field obstacle information, and create a field digital map containing soil properties and landform features.

[0056] In step 100, based on the geographic information data of the current field environment area, the geographic information data is transformed into a digital map of the field environment containing elevation, obstacle distribution and soil characteristics information of different coordinate points using Matlab software; according to the range of the obstacle, the obstacle is simplified into circular and polygonal geometric shapes.

[0057] Preliminary information processing is performed based on on-site exploration or satellite radar information to analyze and merge obstacle areas, thereby determining the location and geometric information of obstacles. In this example, the map size is 50×50, the vehicle center is located at the starting point (0,0), and the ending point is located at (50,50).

[0058] Circular obstacles with radii of 5, 5 and 3 are placed at (4, 15), (20, 20) and (35, 15) respectively. Two rectangular obstacles are also placed with vertex coordinates of {(12, 9), (12, 11), (18, 11), (18, 9)} and {(35, 30), (35, 35), (40, 35), (40, 30)} respectively.

[0059] Therefore, in this embodiment, as Figure 2 and Figure 3 As shown, four circular obstacles and three circular obstacles and two rectangular obstacles were set up in the field environment, respectively. There is no limit to the number of obstacles in actual operation. The fitness calculated in the embodiment is related to vehicle performance and soil characteristics, and then the vehicle passage time is calculated.

[0060] Step 200: Based on the theory of vehicle ground mechanics, establish a mechanical model of wheel-soil coupling in the field environment, and derive the motion equations of vehicle-soil coupling in the field environment.

[0061] Specifically, a wheel-soil coupled mechanical model is established based on the Lagrange equation. Based on the wheel-soil coupled mechanical model, the variation law of subsidence, slip ratio, hook traction force and travel resistance parameters is obtained, and the motion equation of vehicle-soil coupling in the field environment is established.

[0062] Step 300: Based on the digital map of the field environment, randomly generate multiple sets of paths, embed the wheel and soil coupling mechanical model in the field environment, and calculate the initial automatic obstacle avoidance path.

[0063] The method for embedding the wheel-soil coupled mechanical model is as follows:

[0064] (1) Fix the starting point and ending point on the digital map of the field environment, randomly generate multiple sets of coordinate points within the coordinate range of the starting point and ending point, integrate all coordinate points to form a generated path, where the coordinate points represent the horizontal and vertical coordinates of the points on the path respectively.

[0065] (2) Calculate the time required for a vehicle to travel on the generated path by using the motion equations of the vehicle and soil coupling in the field environment.

[0066] (3) Calculate the distance of each coordinate point on the generated path from the starting point coordinate, sort each generated path in order of distance, and then form an initial path population from the N generated paths.

[0067] As one embodiment of this implementation, the initial path population size is set to 1000, the number of generations is 10000, and the initial chromosome number is 4. 1000 sets of paths are randomly generated for both the horizontal and vertical coordinates. Each generated path consists of 4 sets of point coordinates, i.e., the horizontal coordinate generated path genx. i =[genxx i,j genxx i,j+1 genxx i,j+2 genxx i,j+3 ], vertical coordinate generation path geny i =[genyy i,j genyy i,j+1 genyy i,j+2 genyy i,j+3 ] represents the path.

[0068] Set the starting coordinates (x) s y s ) and target coordinates (x) t y t If added to the population, the horizontal coordinate of each generation path becomes genx. i =[x s genxx i,j genxx i,j+1 genxx i,j+2 genxx i,j+3 x t The vertical coordinate of each generated path becomes geny. i =[y s genyy i,j genyy i,j+1 genyy i,j+2 genyy i,j+3 y t ].

[0069] After generating the initial path population, it is necessary to iteratively evaluate the generated path individuals within the initial path population until a path capable of automatic obstacle avoidance is generated. The specific implementation method is as follows:

[0070] Based on the coordinate information of obstacles and soil, obstacle areas and soft soil areas are divided into zones in the digital map of the field environment; for obstacle areas, their accessibility is set to 0.

[0071] When generating a path, it is determined whether the path intersects with an obstacle area. If they intersect, the path is discarded and a new path is generated. Specifically, the method for determining whether generated paths in the initial path population intersect with obstacles is as follows:

[0072] For each generated path, take 50 points from each path segment and determine whether each point falls in the middle of a circular obstacle. At the same time, use the vector cross product method to determine whether each point falls in the middle of a rectangular obstacle.

[0073] In conjunction with the embodiments of this implementation, it is determined whether each point falls in the middle of the circular obstacle, that is:

[0074] (x-xobs(t)) 2 +(y-yobs(t)) 2 -robs(t) 2 ≤0

[0075] Where x and y are the x and y coordinates of the selected point, respectively, xobs(t) is the x coordinate of the center of the t-th obstacle, yobs(t) is the y coordinate of the center of the t-th obstacle, and robs(t) is the radius of the t-th obstacle.

[0076] Simultaneously, to determine whether each point falls within the center of the rectangular obstacle, the cross product of vectors is used for calculation, namely:

[0077] (P2-P1)×(P-P1)*(P4-P3)×(P-P3)≥0

[0078] and,

[0079] (P3-P2)×(P-P2)*(P1-P4)×(P-P4)≥0

[0080] Where P1 to P4 are the four vertices of the rectangle, P is the point taken by each path segment in the individual, the symbol × is the cross product of vectors, and the symbol * is the multiplication sign between numbers.

[0081] If it is determined that the individual intersects with an obstacle, repeat step 300 to regenerate different paths, and continue to determine whether the generated paths in the initial path population intersect with obstacles, until every generated path in the generated population does not intersect with any obstacle.

[0082] Step 400: Establish a path optimization algorithm to obtain the optimal path in the digital map of the field environment.

[0083] The overall process of establishing a path optimization algorithm is as follows:

[0084] (1) Calculate the travel distance of each generated path, calculate the vehicle travel time t(i) based on vehicle dynamics, and then obtain the fitness F(x). i );

[0085] (2) Use roulette wheel selection to filter out the generated paths with low fitness and form a new path population;

[0086] (3) Retain the best generated path in the new path population, judge each generated path, and form a new path population through crossover and mutation operations;

[0087] (4) Repeat steps (1), (2) and (3) above for the new path population, and perform iterative loops to obtain the final optimal path.

[0088] The specific implementation method for establishing a path optimization algorithm to obtain the optimal path in a digital map of the field environment is as follows:

[0089] Based on the dynamic model, the motion equation of the vehicle in the field environment is obtained, and the travel time of the vehicle on each generated path in the initial path population is calculated.

[0090] A fitness function is established based on the travel time to find the minimum travel time. Therefore, the fitness function is as follows:

[0091]

[0092] Where F(x) i Let t be the fitness of the i-th generated path. i The time required for vehicles to pass through.

[0093] Based on the roulette wheel selection method, N generation paths with high fitness are selected, and crossover and mutation operations are performed to form new N generation paths.

[0094] The selection method based on roulette wheel selection is as follows:

[0095] Calculate the probability of each generated path in the initial path population being selected. Specifically, the selection probability of the i-th generated path is calculated based on the fitness function:

[0096]

[0097] Where p(x) i Let F(x) be the probability that the i-th generated path is selected.i ) represents the fitness of the i-th generation path, and N represents the population size;

[0098] Calculate the cumulative probability of each generated path, then the cumulative probability of the i-th generated path is:

[0099]

[0100] Where, q i It represents the cumulative probability of the i-th generated path, which is the sum of the probabilities of all generated paths before the i-th generated path.

[0101] A random number r is generated in the interval [0, 1]. The random number r is then compared with the cumulative probability q of the i-th generated path in the initial path population. i The comparisons are performed sequentially, and the N generation paths with the highest fitness are selected based on the comparison results. Specifically, if r > q i If the i-th generated path is skipped, it is compared with the next generated path until r ≤ q. j If so, then the j-th generated path will be selected.

[0102] Perform crossover and mutation operations on the N generated paths to form N new child paths.

[0103] The implementation of the roulette wheel selection method, which generates N paths, performs crossover and mutation operations to form N new child paths, is as follows:

[0104] The crossover operation uses a two-point crossover method, with a crossover probability P set. c The crossover probability typically ranges from [0.6, 0.9], and then a random number r in the range [0, 1] is generated. c For each of the N generation paths that serve as the parent, determine if r c <P c If the i-th generated path is intersected with the (i+1)-th generated path, then the i-th generated path will be intersected.

[0105] The specific crossover operation is as follows: For the i-th generated path selected for crossover, two coordinate points are randomly selected, and the coordinates in the i-th generated path that are within the range of the two coordinate points are swapped with the (i+1)-th generated path;

[0106] The mutation operation uses a single-point mutation method, with a mutation probability P set. m The mutation probability typically ranges from [0.0001, 0.1], generating a random number r within the range [0, 1]. m For each of the N generation paths that serve as the parent, determine if r m <P mIf the i-th generated path is mutated, then the i-th generated path will be mutated; otherwise, no mutation will occur.

[0107] The specific mutation operation is as follows: for the i-th generation path of the selected mutation, two mutation points are randomly generated, and then the coordinates of the two mutation points are reversed to obtain the offspring chromosome.

[0108] Calculate the fitness of the N offspring chromosomes generated, and repeat selection, crossover, and mutation operations on the N offspring chromosomes until the maximum number of generations is reached to obtain the optimal path.

[0109] The final calculation results are shown below. Figure 3 The results show that the field path planning algorithm proposed in this invention can achieve efficient obstacle avoidance and has good solution capability.

[0110] Based on the above, this embodiment proposes a method for calculating the mechanical properties of vehicles in a field environment. For complex real-world field environments, a method considering the coupling effect between the wheels and the soil is proposed to calculate the mechanical properties and motion characteristics of vehicles in field environments. Based on the method for calculating the mechanical properties of vehicles in field environments, a method for calculating the optimal path of vehicles in field environments is also proposed. By integrating the mechanical properties of vehicles traveling in the field, the optimal path of vehicles traveling in the field can be obtained quickly, enabling them to successfully avoid various obstacles.

[0111] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A method for generating an optimal path for a vehicle to travel in a wild environment, the method comprising the following steps: Step 100: selecting a wild environment area, extracting wild obstacle information, and establishing a digital map of the wild environment containing information about soil properties and landform characteristics; Step 200: based on the theory of vehicle ground mechanics, establishing a wheel-soil coupling mechanics model in the wild environment, and deriving a motion equation of the vehicle-soil coupling in the wild environment; Step 300: based on the digital map of the wild environment, randomly generating a plurality of groups of paths, embedding the wheel-soil coupling mechanics model in the wild environment, and calculating an initial automatic obstacle avoidance path; In the step 300, the embedding method of the wheel-soil coupling mechanics model is as follows: Fixing the starting point and the ending point positions on the digital map of the wild environment, randomly generating a plurality of groups of coordinate points within the coordinate range of the starting point and the ending point, and integrating all the coordinate points to form a generated path, wherein the coordinate points represent the horizontal and vertical coordinates of the points on the path respectively; Using the motion equation of the vehicle-soil coupling in the wild environment, calculating the time required for the vehicle to travel on the generated path; Calculating the length of each coordinate point on the generated path from the starting point coordinate, sequentially sorting each generated path according to the distance, and then forming an initial path population of N generated paths; After generating the initial path population, it is repeatedly judged whether the generated path individuals in the initial path population intersect with the obstacles, until a path capable of automatic obstacle avoidance is generated, and the specific implementation method is as follows: Based on the coordinate information of the obstacles and the soil, the obstacle area and the soft soil area are partitioned in the digital map of the wild environment; for the obstacle area, the passability is set to 0; When generating a path, it is judged whether the path intersects with the obstacle area, and if it does, the path is excluded and a new path is generated; The specific implementation method for judging whether the generated path in the initial path population intersects with the obstacles is as follows: Taking 50 points in each segment of the generated path, it is judged whether each point falls in the middle of the circular obstacle, and at the same time, the calculation method of vector cross product is used to judge whether each point falls in the middle of the rectangular obstacle; Judging whether each point falls in the middle of the circular obstacle, that is: (x - xobs(t)) 2 +(y - yobs(t)) 2 -robs(t) 2 ≤ 0 Wherein, x and y are the horizontal and vertical coordinates of the taken point, xobs(t) is the horizontal coordinate of the center of the tth obstacle, yobs(t) is the vertical coordinate of the center of the tth obstacle, and robs(t) is the radius of the tth obstacle; At the same time, it is judged whether each point falls in the middle of the rectangular obstacle, and the calculation method of vector cross product is used, that is: (P2-P1)×(P-P1)*(P4-P3)×(P-P3)≥0 And, (P3-P2)×(P-P2)*(P1-P4)×(P-P4)≥0 Wherein, P1-P4 are the four vertices of the rectangle, P is the taken point in each segment of the individual, symbol × is the vector cross product, and symbol * is the multiplication sign between numbers; If it is judged that the individual intersects with the obstacles, the work of step 300 is repeated, different paths are re-generated, and it is continued to judge whether the generated paths in the initial path population intersect with the obstacles, until each generated path in the population does not intersect with the obstacles. Step 400, establishing a path optimization algorithm to obtain the optimal path in the digital map of the field environment.

2. The method of claim 1, wherein the method further comprises: In the step 100, based on the geographic information data of the current field environment area, the geographic information data is converted into the digital map of the field environment containing the information of the elevation, obstacle distribution and soil characteristics of different coordinate points by using the Matlab software. According to the range where the obstacle is located, the obstacle is simplified into a circular and polygonal geometric shape.

3. The method of claim 1, wherein the method further comprises: In the step 200, a wheel-soil coupling mechanical model is established based on the Lagrange equation. Based on the wheel-soil coupling mechanical model, the variation law of the sinkage, slip rate, hook traction and travel resistance parameters is obtained, and a motion equation of the vehicle-soil coupling in the field environment is established.

4. The method of claim 1, wherein the method further comprises: In the step 400, the implementation of establishing a path optimization algorithm to obtain the optimal path in the digital map of the field environment is as follows: Based on the mechanical model, the motion equation of the vehicle in the field environment is obtained, and the travel time of the vehicle on each generated path in the initial path population is calculated. Based on the travel time, an adaptive function is established to find the minimum value of the travel time, and the adaptive function is as follows: where F(x i ) is the fitness of the ith generated path, t i is the time required for the vehicle to pass through. Based on the roulette wheel selection method, N generated paths with large adaptability are selected for crossover operation and mutation operation to form new N generated paths.

5. The method of claim 4, wherein the roulette wheel selection method is as follows: The probability of each generated path in the initial path population being selected is calculated, and the selection probability of the i-th generated path is calculated according to the adaptive function as follows: The cumulative probability of each generated path is calculated, and the cumulative probability of the i-th generated path is as follows: where p(x i ) is the probability of the i-th generated path being selected, F(x i ) is the fitness of the i-th generated path, and N is the population size. The N generated paths are subjected to crossover operation and mutation operation to form new N offspring paths. wherein q i is the cumulative probability of the i-th said generation path, i.e. the sum of the probabilities of all generation paths preceding the i-th said generation path; A random number r is generated in the interval [0, 1]. The random number r is then compared with the cumulative probability q of the i-th generated path in the initial path population. i Compare them sequentially; if r > q i If the i-th generated path is skipped, it is compared with the next generated path until r ≤ q. j Then the j-th generated path will be selected.

6. The method of claim 5, wherein the implementation of the roulette wheel selection method for the crossover operation and mutation operation of the N generated paths to form new N offspring paths is as follows: The specific crossover operation is as follows: randomly selecting two coordinate points for the i-th generated path selected for crossover, and exchanging the coordinates in the i-th generated path within the range of the two coordinate points with the i+1-th generated path. The specific mutation operation is as follows: randomly generating two mutation points for the i-th generated path selected for mutation, and then reversing the coordinate points of the two mutation points to obtain the offspring chromosome. The crossover operation adopts a two-point crossover manner, and a crossover probability P is set c A random number r in a range of [0, 1] is generated c Each of the N generated paths serving as a parent is judged, if r c <P c The i-th generated path is crossed with the i+1-th generated path.

7. The method of claim 6, wherein the method further comprises: The mutation operation adopts a single-point mutation manner, and a mutation probability P is set m A random number r in a range of [0, 1] is generated m Each generated path in the N generated paths as the parent is judged, if r m <P m The i-th generated path is subjected to the mutation operation, otherwise no mutation occurs; ​ ​ The fitness of the N generated offspring chromosomes is calculated, and the selection, crossover and mutation operations are repeated on the N offspring chromosomes until a maximum evolution generation is reached, and an optimal path is obtained.

Images