Multi-path planning method and device

Abstract

The invention discloses a multi-path planning method and device, and relates to the technical field of automatic driving. A specific embodiment of the method comprises the following steps: acquiring a starting point and an ending point of a to-be-planned path; initializing a particle swarm based on the starting point and the ending point of the to-be-planned path; each particle in the particle swarm corresponds to a candidate path; based on a simulated annealing particle swarm algorithm and the target function of the round, determining a planned path of the round from the candidate paths; wherein the target function of the current round is obtained according to a sequential ecological niche algorithm and the target function of the previous round. According to the embodiment, the simulated annealing particle swarm optimization and the sequential niche algorithm are combined to realize multi-path planning, the population can be effectively prevented from falling into a local optimal solution, the target function is updated through sequence optimization, the search of all or local optimal paths is ensured, and the efficiency and accuracy of path planning are improved.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and in particular to a method and apparatus for multi-path planning. Background Technology

[0002] Path planning is a key technology for autonomous vehicles. When autonomous vehicles perform tasks, the complexity of the actual environment changes over time, and a single planned path may be ineffective. Alternatively, autonomous vehicles may be required to reach the target area from different directions to perform tasks. Therefore, multi-path planning is an effective method to solve the above problems.

[0003] In related technologies, particle swarm optimization (PSO) is used for path planning to obtain multiple paths. However, the global and local search capabilities of particles in PSO cannot be well balanced, and it is easy to get trapped in local optima too early, making it difficult to obtain the optimal path for the entire region. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method and apparatus for multi-path planning, which can achieve multi-path planning by combining simulated annealing particle swarm optimization algorithm with sequential niche algorithm, effectively prevent the population from getting trapped in local optima, and ensure the search of all or local optimal paths by updating the objective function through sequence optimization, thereby improving the efficiency and accuracy of path planning.

[0005] To achieve the above objectives, according to one aspect of the present invention, a multi-path planning method is provided, comprising:

[0006] Obtain the start and end points of the path to be planned;

[0007] The particle swarm is initialized based on the start and end points of the path to be planned; each particle in the particle swarm corresponds to a candidate path.

[0008] Based on the simulated annealing particle swarm optimization algorithm and the objective function of this round, the planned path for this round is determined from each candidate path;

[0009] The objective function for this round is obtained based on the sequential niche algorithm and the objective function of the previous round.

[0010] Optionally, the objective function for this round is obtained through the following method:

[0011] The niche radius of the candidate path is calculated based on the number of path points in the candidate path;

[0012] Calculate the path distance between the planned path from the previous round and the candidate path;

[0013] Based on the sequential niche algorithm, the depreciation function is constructed according to the niche radius and the path distance;

[0014] The objective function for this round is obtained based on the reduction function and the objective function of the previous round;

[0015] The devaluation function is used to reduce the value of the objective function when the path distance is less than the niche radius.

[0016] Optionally, the niche radius of the candidate path is calculated based on the number of path points in the candidate path, including:

[0017] Obtain the area and variable dimension of the planning region corresponding to the path to be planned;

[0018] The niche radius is calculated based on the area of ​​the planned region, the number of path points, and the dimension of the variables.

[0019] Optionally, calculating the path distance between the planned path from the previous round and the candidate path includes:

[0020] According to preset rules, the planned path and the candidate path of the previous round are segmented to obtain each first segmentation point corresponding to the planned path of the previous round and each second segmentation point corresponding to the candidate path.

[0021] Calculate the segmentation distance between each first segmentation point and its corresponding second segmentation point, and use the sum of all segmentation distances as the path distance.

[0022] Optionally, a devaluation function is constructed based on the niche radius and the path distance, including:

[0023] When the path distance is greater than the niche radius, the devaluation function is a preset value;

[0024] If the path distance is less than or equal to the niche radius, calculate the difference between the niche radius and the path distance, and construct the depreciation function based on the ratio of the difference to the niche radius.

[0025] Optionally, based on the simulated annealing particle swarm optimization algorithm and the objective function of this round, the planned path for this round is determined from each candidate path, including:

[0026] For each particle in the particle swarm, obtain the position of each particle; calculate the objective function value corresponding to each particle based on the objective function of this round and the position of each particle;

[0027] The optimal position of the particle swarm and the optimal position of each particle are updated according to the objective function value corresponding to each particle; the optimal position of the particle swarm is the position of the particle with the optimal objective function value, and the optimal position of each particle is the position corresponding to the optimal objective function value of each particle; the position of each particle is updated according to the optimal position of the particle swarm and the optimal position of each particle.

[0028] Determine whether the iteration has converged. If it has converged, use the candidate path corresponding to the particle at the optimal position of the particle swarm as the planned path for this round.

[0029] Optionally, updating the position of each particle based on the optimal position of the particle swarm and the optimal position of each individual particle includes:

[0030] Obtain the velocity of each particle;

[0031] The velocity learning value is calculated based on the optimal position of the particle swarm and the optimal position of each particle.

[0032] The velocity of each particle is updated based on the learned velocity value;

[0033] Update the position of each particle based on its updated velocity.

[0034] According to another aspect of the present invention, a multi-path planning apparatus is provided, comprising:

[0035] The module retrieves the start and end points of the path to be planned.

[0036] The initialization module initializes a particle swarm based on the start and end points of the path to be planned; each particle in the particle swarm corresponds to a candidate path.

[0037] The determination module, based on the simulated annealing particle swarm optimization algorithm and the objective function of this round, determines the planned path for this round from each candidate path;

[0038] The objective function for this round is obtained based on the sequential niche algorithm and the objective function of the previous round.

[0039] According to another aspect of the present invention, an electronic device is provided, comprising:

[0040] One or more processors;

[0041] Storage device for storing one or more programs.

[0042] When the one or more programs are executed by the one or more processors, the one or more processors implement the multi-path planning method provided by the present invention.

[0043] According to another aspect of the present invention, a computer-readable medium is provided having a computer program stored thereon, which, when executed by a processor, implements the multi-path planning method provided by the present invention.

[0044] One embodiment of the above invention has the following advantages or beneficial effects: The multi-path planning method of this invention first obtains the starting point and ending point of the path to be planned, then initializes a particle swarm based on the starting point and ending point, with each particle in the particle swarm corresponding to a candidate path; then, based on the simulated annealing particle swarm algorithm and the objective function of the current round, the planned path for the current round is determined from each candidate path, wherein the objective function of the current round is obtained based on the sequential niche algorithm and the objective function of the previous round. This method combines the simulated annealing particle swarm algorithm and the sequential niche algorithm for multi-path planning. Based on the high convergence and ease of escaping local optima of the simulated annealing particle swarm algorithm, it can prevent the particle swarm from being lost prematurely and trapped in local optima, effectively solving the global path planning problem of autonomous vehicles; by adaptively adjusting the objective function of the space near the known planned path through the sequential niche algorithm, it avoids the swarm from being trapped in the same space, finds different local optimal paths through sequential optimization, and ensures that the simulated annealing particle swarm algorithm sequentially searches for global or local optimal paths by updating the objective function, thus improving the efficiency and accuracy of multi-path planning.

[0045] The further effects of the aforementioned unconventional alternative methods will be explained below in conjunction with specific implementation methods. Attached Figure Description

[0046] The accompanying drawings are provided to better understand the invention and are not intended to unduly limit the scope of the invention. Wherein:

[0047] Figure 1 This is a schematic diagram of the main flow of a multi-path planning method according to an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of the main flow of another multi-path planning method according to an embodiment of the present invention;

[0049] Figure 3 This is a schematic diagram of the main flow of another multi-path planning method according to an embodiment of the present invention;

[0050] Figure 4 This is a flowchart illustrating a multi-path planning method according to an embodiment of the present invention;

[0051] Figure 5 This is a schematic diagram of the main modules of a multipath planning device according to an embodiment of the present invention;

[0052] Figure 6 This is an exemplary system architecture diagram in which embodiments of the present invention can be applied;

[0053] Figure 7 This is a schematic diagram of the structure of a computer system suitable for implementing terminal devices or servers of the present invention. Detailed Implementation

[0054] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of the present invention, including various details to aid understanding. These details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0055] In the field of autonomous driving, path planning algorithms are a key technology for autonomous vehicles. The core technology is the path planning algorithm. Currently, commonly used path planning methods include ant colony optimization, genetic algorithms, and artificial potential field methods. These methods each have their advantages, but also some limitations, such as high computational complexity, poor convergence, and a tendency to get trapped in local optima, which restrict path planning to some extent.

[0056] Particle Swarm Optimization (PSO) is a swarm intelligence optimization algorithm. This algorithm preserves the global search strategy of the swarm. Through a velocity-displacement model and its own memory characteristics, it achieves dynamic tracking and real-time adjustment of the search. Its advantages include simple algorithm design, ease of implementation, few parameters requiring adjustment, fast convergence speed, and strong global search capability. However, when using PSO for multi-path planning, the global and local search capabilities of particles cannot be well balanced; it is prone to prematurely getting trapped in local optima, making it difficult to obtain the global optimum; and during iteration, the swarm may get stuck in the same niche, leading to a decrease in algorithm efficiency.

[0057] Therefore, this invention provides a multi-path planning method that combines simulated annealing particle swarm optimization with sequential niche optimization. This method can prevent the population from getting trapped in local optima and avoid the population from getting stuck in the same space. It finds multiple different local optima through sequence optimization.

[0058] Figure 1 This is a schematic diagram of the main flow of a multi-path planning method according to an embodiment of the present invention, as shown below. Figure 1 As shown, this multi-path planning method includes the following steps:

[0059] Step S101: Obtain the starting point and ending point of the path to be planned;

[0060] Step S102: Initialize the particle swarm based on the start and end points of the path to be planned; each particle in the particle swarm corresponds to a candidate path;

[0061] Step S103: Based on the simulated annealing particle swarm algorithm and the objective function of this round, determine the planned path for this round from each candidate path;

[0062] The objective function for this round is obtained based on the sequential niche algorithm and the objective function of the previous round.

[0063] In this embodiment of the invention, the multi-path planning method can solve the multi-path planning problem of autonomous vehicles, i.e., driverless cars, thereby planning multiple paths for autonomous vehicles from the starting point to the destination. When planning paths, it is necessary to obtain the starting point and ending point of the paths to be planned, i.e., initialize environmental parameters. Environmental parameters include the planning area boundary, the starting point, the ending point, the threat area (i.e., the restricted area), and the number of paths to be planned. The planning area boundary includes the boundary geographic coordinates, the starting point is the geographic coordinates of the initial position, and the ending point is the geographic coordinates of the final position. The dimension of the geographic coordinates is the variable dimension, which can be preset, such as being two-dimensional or three-dimensional.

[0064] In this embodiment of the invention, after obtaining the start and end points of the path to be planned, a particle swarm is initialized based on these points, with each particle in the swarm representing a candidate path. The population size of the particle swarm, i.e., the number of particles in the swarm, can be preset. Randomly initializing the particle swarm based on its population size allows the position and velocity of each particle to be obtained. Specifically, a Latin hypercube can be used to initialize the particle swarm to generate approximately uniformly distributed particles, ensuring both diversity and uniform distribution of the initialized particles.

[0065] In this embodiment of the invention, after initializing the particle swarm, the planned path for this round is determined from each candidate path based on the Simulated Annealing Particle Swarm Optimization (SAPSO) algorithm and the objective function of this round, specifically including:

[0066] For each particle in the particle swarm, obtain the position of each particle; based on the objective function of this round and the position of each particle, calculate the objective function value corresponding to each particle;

[0067] Update the optimal position of the particle swarm and the optimal position of each individual particle based on the objective function value corresponding to each particle; the optimal position of the particle swarm is the position of the particle with the optimal objective function value, and the optimal position of each individual particle is the position corresponding to the optimal objective function value of each particle; update the position of each particle based on the optimal position of the particle swarm and the optimal position of each individual particle.

[0068] Determine if the iteration has converged. If it has converged, use the candidate path corresponding to the particle at the optimal position of the particle swarm as the planned path for this round.

[0069] Specifically, for each iteration, the position and velocity of each particle in the particle swarm are first obtained. If it is the first iteration, the position and velocity of each particle are the same as those of each particle after the initialization of the particle swarm. Then, the objective function for this round is obtained. The position of each particle is input into the objective function for this round, and the objective function value of each particle is calculated. The objective function value of each particle can be the fitness value of each particle. The objective function can be a function constructed with the position of the particle as the independent variable. The objective function can be a function that indicates the path cost or a function that indicates the path length based on the position of the particle.

[0070] After obtaining the objective function value of each particle in the particle swarm, the objective function values ​​of each particle are compared to obtain the optimal objective function value of the particle swarm. The optimal objective function value of the particle swarm is then compared with the optimal objective function value already stored in the particle swarm. If the optimal objective function value of the particle swarm is better than the optimal objective function value already stored in the particle swarm, the optimal objective function value of the particle swarm is updated, and the position of the particle with the updated optimal objective function value is taken as the optimal position of the particle swarm.

[0071] The objective function value of each particle is compared with the optimal objective function value stored for that particle. If the objective function value of the particle is better than the optimal objective function value stored for that particle, the optimal objective function value of the particle is updated, and the position of the particle corresponding to the updated optimal objective function value is taken as the optimal position of the particle.

[0072] Whether it is the optimal objective function value is determined by the objective function. If the objective function is a function that indicates the path cost, then the optimal objective function value is the minimum objective function value.

[0073] Then, the position and velocity of each particle are updated based on the optimal position of the particle swarm and the optimal position of each individual particle.

[0074] The iteration convergence is then determined by the distance between the optimal position of the particle swarm and the updated position of the corresponding particle. If this distance is less than or equal to a preset clustering threshold, the iteration converges, and the candidate path corresponding to the particle at the optimal position is used as the planning path for this round. If the number of planned paths is less than the preset number of planned paths, the process continues to determine the planning path for the next round. If the distance is greater than the preset distance threshold, the iteration fails to converge, and the iteration continues. That is, the objective function value of each particle is calculated based on its updated position, and then the optimal position of the particle swarm and the optimal position of each particle are updated based on the objective function value of each particle to determine whether the iteration converges.

[0075] In this embodiment of the invention, the maximum number of iterations of the simulated annealing particle swarm optimization algorithm can also be preset. For example, when the maximum number of iterations is reached, the iteration is stopped and the iteration is determined to be converged.

[0076] In this embodiment of the invention, updating the position of each particle based on the optimal position of the particle swarm and the optimal position of each particle includes: obtaining the velocity of each particle; calculating a velocity learning value based on the optimal position of the particle swarm and the optimal position of each particle; updating the velocity of each particle based on the velocity learning value; and updating the position of each particle based on the updated velocity of each particle.

[0077] In this embodiment of the invention, when updating the position of each particle based on the optimal position of the particle swarm and the optimal position of each individual particle, the velocity of each particle can be updated first, and then the position of each particle can be updated based on the updated velocity. Specifically, the updated position of each particle can be obtained by summing its position and its updated velocity. Obtaining the updated velocity of each particle can be achieved by acquiring the velocity of each particle, calculating a velocity learning value based on the optimal position of the particle swarm and the optimal position of each individual particle, such as calculating a first velocity learning value based on the optimal position of each particle, calculating a second velocity learning value based on the optimal position of the particle swarm, and summing the first and second velocity learning values ​​as the velocity learning value. The sum of the velocity and the velocity learning value of each particle can be calculated, and the product of the sum and the inertia weight is used as the updated velocity of each particle.

[0078] In another implementation, obtaining the updated velocity of each particle can also be achieved by calculating the product of the velocity and the inertia weight of each particle, and then summing the product with the velocity learning value as the updated velocity of each particle.

[0079] The first velocity learning value is calculated based on the optimal position of each particle. This can be achieved by calculating the first difference between the optimal position of each particle and the position of each particle. This first difference can be used as the first velocity learning value, or the product of the first difference and a first preset coefficient can be used as the first velocity learning value. The first preset coefficient can be a first learning factor, a first random number, or the product of the first learning factor and the first random number. Similarly, the second velocity learning value is calculated based on the optimal position of the particle swarm. This can be achieved by calculating the second difference between the optimal position of the particle swarm and the position of each particle. This second difference can be used as the second velocity learning value, or the product of the second difference and a second preset coefficient can be used as the second velocity learning value. The second preset coefficient can be a second learning factor, a second random number, or the product of the second learning factor and the second random number.

[0080] In this embodiment of the invention, the velocity of each particle can be updated according to equation (1), and the position of each particle can be updated according to equation (2).

[0081] Equation (1)

[0082] (2)

[0083] In equations (1) and (2), i represents the i-th particle, j is the current round, d is the maximum round, i.e., the number of paths to be planned; k is the current iteration number. Let be the optimal position of the i-th particle in the j-th round and k-th iteration. Let be the optimal position of the particle swarm in the j-th round and k-th iteration. and These are the first learning factor and the second learning factor, respectively. and These are the first and second random numbers, respectively, distributed within the interval [0,1]. It is the velocity of the particle; It represents the position of the particle. For inertial weights, ,in .

[0084] In embodiments of the present invention, such as Figure 2 As shown, the objective function for this round was obtained using the following method:

[0085] Step S201: Calculate the niche radius of the candidate path based on the number of path points in the candidate path;

[0086] Step S202: Calculate the path distance between the planned path and the candidate path from the previous round;

[0087] Step S203: Based on the sequential niche algorithm, construct a devaluation function according to the niche radius and path distance;

[0088] Step S204: Obtain the objective function for this round based on the reduction function and the objective function of the previous round;

[0089] The devaluation function is used to reduce the value of the objective function when the path distance is less than the niche radius.

[0090] In this embodiment of the invention, the objective function of this round is obtained based on the sequential niche algorithm (i.e., the sequential niche algorithm, SNT algorithm) and the objective function of the previous round. That is, the objective function of this round is obtained by modifying the objective function of the previous round based on the sequential niche algorithm. Specifically, the niche radius, i.e., the niche, of the candidate path is calculated based on the number of path points in the candidate path. The niche radius is positively correlated with the number of path points. Then, the path distance between the planned path and the candidate path in the previous round is calculated. Based on the sequential niche algorithm, a reduction function is constructed based on the niche radius and the path distance. The objective function of this round is obtained based on the reduction function and the objective function of the previous round. The reduction function is used to reduce the value of the objective function. Specifically, the objective function of this round is the product of the objective function of the previous round and the reduction function.

[0091] In this embodiment of the invention, calculating the niche radius of a candidate path based on the number of path points in the candidate path includes:

[0092] Obtain the area and variable dimensions of the planning region corresponding to the path to be planned;

[0093] The niche radius is calculated based on the area of ​​the planned region, the number of path points, and the dimension of the variables.

[0094] In this embodiment of the invention, the niche radius is used to define the size of the search area within the domain of the previous round of planning paths. A small niche radius limits its spatial range of influence, meaning the "filling" effect is limited to areas very close to the already found optimal path, i.e., the planning path from the previous round. This may lead to finding another "optimal" path relatively close to the current optimal path, potentially causing spurious minima in the objective function within these local regions. A larger niche radius affects other true minima in the search space. A larger niche radius means the "filling" effect affects a wider area, potentially masking the existence of true minima. In other words, an excessively large niche radius may cause many areas in the search space to be "filled," thus obscuring the truly optimal paths.

[0095] In this embodiment of the invention, the ecological niche radius is negatively correlated with the variable dimension and correlated with the arithmetic square root of the area of ​​the planned region. For example, the ecological board radius can be obtained according to equation (3):

[0096] Equation (3)

[0097] In equation (3), r is the ecological niche radius, p is the number of path points, S is the area of ​​the planned area, z is the preset coefficient, and D is the variable dimension.

[0098] In embodiments of the present invention, such as Figure 3 As shown, the path distance between the planned path and the candidate path in the previous round is calculated, including:

[0099] Step S301: According to the preset rules, the planned path and candidate path of the previous round are divided into segments to obtain the first segmentation points corresponding to the planned path of the previous round and the second segmentation points corresponding to the candidate path.

[0100] Step S302: Calculate the segmentation distance between each first segmentation point and the corresponding second segmentation point, and use the sum of all segmentation distances as the path distance.

[0101] In this embodiment of the invention, when calculating the path distance between the planned path and the candidate path in the previous round, the planned path and the candidate path in the previous round are divided according to preset rules. This can be done by dividing the planned path and the candidate path in the previous round according to preset directions (such as the X-axis, Y-axis, or Z-axis), obtaining first dividing points on the planned path and second dividing points on the candidate path. At the same dividing position, there are corresponding first dividing points and second dividing points. During the division, the number of dividing points can be set, and the path can be divided equally according to the number of dividing points; alternatively, path points on the planned path or candidate path in the previous round can be used as dividing points. After obtaining each first dividing point and each second dividing point, the dividing distance between the first dividing point and the corresponding second dividing point is calculated. This dividing distance can be a Euclidean distance, and then the result of each dividing distance is used as the path distance.

[0102] In this embodiment of the invention, a devaluation function is constructed based on the niche radius and path distance, including:

[0103] When the path distance is greater than the niche radius, the devaluation function is the preset value;

[0104] When the path distance is less than or equal to the niche radius, calculate the difference between the niche radius and the path distance, and construct a deduction function based on the ratio of the difference to the niche radius.

[0105] In this embodiment of the invention, when constructing the devaluation function based on the niche radius and path distance, since the devaluation function is used to correct the value of the objective function within the niche radius, that is, when the path distance is less than or equal to the niche radius, the devaluation function is used to correct the objective function of the previous round. Therefore, it is necessary to determine the magnitude between the path distance and the niche radius to determine the devaluation function. Specifically, when the path distance is greater than the niche radius, the devaluation function is a preset value, which can be 1, that is, no correction is made to the objective function. When the path distance is less than or equal to the niche radius, the difference between the niche radius and the path distance is calculated, and the ratio of the difference to the niche radius is further calculated. The devaluation function is constructed based on the ratio and the devaluation coefficient. The devaluation coefficient can be preset. When setting the devaluation coefficient, it is best to set a sufficiently large devaluation coefficient so that the value of the objective function after updating the objective function increases significantly within the niche, i.e., within the niche radius, so that the particle does not converge to the niche in the next iteration.

[0106] In one embodiment of the present invention, the dereasing function is shown in equation (4), and the objective function is modified according to equation (5).

[0107] Equation (4)

[0108] Equation (5)

[0109] In equations (4) and (5), The corrected objective function for searching the (n+1)th minimum value is the objective function for the (n+1)th round of path planning. The original objective function for searching the nth minimum value is the objective function for the nth round of path planning. For dereasing function, Let m be the nth minimum value, i.e., the planning path in the nth round, where m is the reduction value and r is the niche radius. For candidate path x and The distance between them.

[0110] Figure 4This is a flowchart illustrating a multi-path planning method according to an embodiment of the present invention. First, initialize the environmental parameters, including the planning area boundary, start point, end point, restricted areas, and the number of paths to be planned, K. Calculate the niche radius. Initialize the annealing simulation particle swarm optimization algorithm, such as setting the population size and maximum number of iterations. During the i-th round of path planning, initialize the particle swarm using a Latin hypercube to obtain the position and velocity of each particle. Calculate the fitness of each particle. Compare the fitness of each particle to determine the optimal position (Gbest) of the particle swarm and the optimal position (Pbest) of each particle. Update the position and velocity of each particle based on the optimal positions of the particle swarm and each particle. Perform a convergence check to determine if the iteration has converged. If it has converged, obtain the planned path for this round, i.e., the optimal path Si, and save it. Let i = i + 1, and check if i equals K. If yes, end the process and obtain multiple planned paths. If not, update the objective function according to the derating function and proceed to the next round. If the iteration has not converged, calculate the fitness of each particle based on the updated position of each particle and continue iterating until it converges.

[0111] The multi-path planning method of this invention first obtains the start and end points of the path to be planned, then initializes a particle swarm based on the start and end points, with each particle in the swarm corresponding to a candidate path; then, based on the simulated annealing particle swarm algorithm and the objective function of the current round, the planned path for the current round is determined from each candidate path, wherein the objective function of the current round is obtained based on the sequential niche algorithm and the objective function of the previous round. This method combines the simulated annealing particle swarm algorithm and the sequential niche algorithm for multi-path planning. Leveraging the high convergence and ease of escaping local optima of the simulated annealing particle swarm algorithm, it can prevent the particle swarm from being lost prematurely and trapped in local optima, effectively solving the global path planning problem for autonomous vehicles. The sequential niche algorithm adaptively adjusts the objective function of the space near the known planned path, avoiding the swarm from getting trapped in the same space. Sequential optimization finds different local optimal paths, and updating the objective function ensures that the simulated annealing particle swarm algorithm sequentially searches for global or local optimal paths, improving the efficiency and accuracy of multi-path planning. This invention also provides an effective method for calculating the niche radius of multi-path planning, which can make more appropriate corrections to the objective function in order to find the globally optimal path more accurately.

[0112] According to another aspect of the embodiments of the present invention, such as Figure 5 As shown, a multi-path planning device 500 is provided, comprising:

[0113] Get module 501 to obtain the start and end points of the path to be planned;

[0114] Initialization module 502 initializes the particle swarm based on the start and end points of the path to be planned; each particle in the particle swarm corresponds to a candidate path;

[0115] The determination module 503, based on the simulated annealing particle swarm algorithm and the objective function of this round, determines the planned path for this round from each candidate path;

[0116] The objective function for this round is obtained based on the sequential niche algorithm and the objective function of the previous round.

[0117] In this embodiment of the invention, the determining module 503 is further configured to: calculate the niche radius of the candidate path based on the number of path points in the candidate path; calculate the path distance between the planned path in the previous round and the candidate path; construct a depreciation function based on the niche radius and the path distance using the sequential niche algorithm; and obtain the objective function for the current round based on the depreciation function and the objective function of the previous round; wherein the depreciation function is used to reduce the value of the objective function when the path distance is less than the niche radius.

[0118] In this embodiment of the invention, the determining module 503 is further configured to: obtain the area of ​​the planning region and the variable dimension corresponding to the path to be planned; and calculate the niche radius based on the area of ​​the planning region, the number of path points and the variable dimension.

[0119] In this embodiment of the invention, the determining module 503 is further configured to: segment the planned path and candidate path of the previous round according to preset rules to obtain each first segmentation point corresponding to the planned path of the previous round and each second segmentation point corresponding to the candidate path; calculate the segmentation distance between each first segmentation point and the corresponding second segmentation point, and use the sum of each segmentation distance as the path distance.

[0120] In this embodiment of the invention, the determining module 503 is further configured to: when the path distance is greater than the niche radius, set the devaluation function to a preset value; when the path distance is less than or equal to the niche radius, calculate the difference between the niche radius and the path distance, and construct the devaluation function based on the ratio of the difference to the niche radius.

[0121] In this embodiment of the invention, the determining module 503 is further configured to: obtain the position of each particle in the particle swarm; calculate the objective function value corresponding to each particle based on the objective function of the current round and the position of each particle; update the optimal position of the particle swarm and the optimal position of each particle based on the objective function value corresponding to each particle; the optimal position of the particle swarm is the position of the particle with the optimal objective function value, and the optimal position of each particle is the position corresponding to the optimal objective function value of each particle; update the position of each particle based on the optimal position of the particle swarm and the optimal position of each particle; determine whether the iteration converges, and if the iteration converges, use the candidate path corresponding to the particle with the optimal position of the particle swarm as the planned path for the current round.

[0122] In this embodiment of the invention, the determining module 503 is further configured to: obtain the velocity of each particle; calculate a velocity learning value based on the optimal position of the particle swarm and the optimal position of each particle; update the velocity of each particle based on the velocity learning value; and update the position of each particle based on the updated velocity of each particle.

[0123] According to another aspect of the present invention, an electronic device is provided, comprising: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the multi-path planning method provided by the present invention.

[0124] According to another aspect of the present invention, a computer-readable medium is provided having a computer program stored thereon, which, when executed by a processor, implements the multi-path planning method provided by the present invention.

[0125] Figure 6 An exemplary system architecture 600 is shown that can be applied to the multipath planning method or multipath planning apparatus of the present invention.

[0126] like Figure 6 As shown, system architecture 600 may include terminal devices 601, 602, and 603, a network 604, and a server 605. Network 604 serves as the medium for providing communication links between terminal devices 601, 602, and 603 and server 605. Network 604 may include various connection types, such as wired or wireless communication links or fiber optic cables, etc.

[0127] Users can use terminal devices 601, 602, and 603 to interact with server 605 via network 604 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 601, 602, and 603, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0128] Terminal devices 601, 602, and 603 can be various electronic devices with displays and web browsing capabilities, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0129] Server 605 can be a server that provides various services, such as a backend management server that supports shopping websites browsed by users using terminal devices 601, 602, and 603 (for example only). The backend management server can analyze and process data such as received product information query requests, and feed back the processing results (such as target push information, product information - for example only) to the terminal devices.

[0130] It should be noted that the multi-path planning method provided in the embodiments of the present invention is generally executed by server 605, and correspondingly, the multi-path planning device is generally set in server 605.

[0131] It should be understood that Figure 6 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0132] The following is for reference. Figure 7 It shows a schematic diagram of the structure of a computer system 700 suitable for implementing a terminal device of the present invention. Figure 7 The terminal device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0133] like Figure 7 As shown, the computer system 700 includes a central processing unit (CPU) 701, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 702 or programs loaded from storage section 708 into random access memory (RAM) 703. The RAM 703 also stores various programs and data required for the operation of the system 700. The CPU 701, ROM 702, and RAM 703 are interconnected via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0134] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, mouse, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.

[0135] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 709, and / or installed from removable medium 711. When the computer program is executed by central processing unit (CPU) 701, it performs the functions defined above in the system of this invention.

[0136] It should be noted that the computer-readable medium shown in this invention can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0137] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0138] The modules described in the embodiments of the present invention can be implemented in software or hardware. The described modules can also be housed in a processor; for example, a processor can be described as including an acquisition module, an initialization module, and a determination module. The names of these modules do not necessarily limit the module itself; for example, the acquisition module can also be described as "a module for acquiring the start and end points of the path to be planned."

[0139] In another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiments; or it may exist independently and not assembled into the device. The computer-readable medium carries one or more programs that, when executed by the device, cause the device to include: obtaining the start and end points of a path to be planned; initializing a particle swarm based on the start and end points of the path to be planned; each particle in the particle swarm corresponding to a candidate path; determining the planned path for the current round from the candidate paths based on the simulated annealing particle swarm algorithm and the objective function of the current round; wherein the objective function of the current round is obtained based on the sequential niche algorithm and the objective function of the previous round.

[0140] According to the technical solution of the present invention, the multi-path planning method of the present invention first obtains the starting point and ending point of the path to be planned, and then initializes a particle swarm based on the starting point and ending point, with each particle in the particle swarm corresponding to a candidate path; then, based on the simulated annealing particle swarm algorithm and the objective function of the current round, the planned path for the current round is determined from each candidate path, wherein the objective function of the current round is obtained based on the sequential niche algorithm and the objective function of the previous round. This method combines the simulated annealing particle swarm algorithm and the sequential niche algorithm for multi-path planning. Based on the high convergence and ease of escaping local optima of the simulated annealing particle swarm algorithm, it can prevent the particle swarm from being lost prematurely and trapped in local optima, effectively solving the global path planning problem of autonomous vehicles; by adaptively adjusting the objective function of the space near the known planned path through the sequential niche algorithm, it avoids the swarm from getting trapped in the same space, finds different local optimal paths through sequential optimization, and updates the objective function, ensuring that the simulated annealing particle swarm algorithm sequentially searches for global or local optimal paths, improving the efficiency and accuracy of multi-path planning. This invention also provides an effective method for calculating the niche radius of multi-path planning, which can make more appropriate corrections to the objective function in order to find the globally optimal path more accurately.

[0141] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A multi-path planning method, characterized in that, include: Obtain the start and end points of the path to be planned; Initialize the particle swarm based on the start and end points of the path to be planned; Each particle in the particle swarm corresponds to a candidate path. Based on the simulated annealing particle swarm optimization algorithm and the objective function of this round, the planned path for this round is determined from each candidate path; The objective function for this round is obtained based on the sequential niche algorithm and the objective function of the previous round.

2. The method according to claim 1, characterized in that, The objective function for this round was obtained through the following method: The niche radius of the candidate path is calculated based on the number of path points in the candidate path; Calculate the path distance between the planned path from the previous round and the candidate path; Based on the sequential niche algorithm, the depreciation function is constructed according to the niche radius and the path distance; The objective function for this round is obtained based on the reduction function and the objective function of the previous round; The devaluation function is used to reduce the value of the objective function when the path distance is less than the niche radius.

3. The method according to claim 2, characterized in that, The niche radius of the candidate path is calculated based on the number of path points in the candidate path, including: Obtain the area and variable dimension of the planning region corresponding to the path to be planned; The niche radius is calculated based on the area of ​​the planned region, the number of path points, and the dimension of the variables.

4. The method according to claim 2, characterized in that, Calculating the path distance between the planned path from the previous round and the candidate path includes: According to preset rules, the planned path and the candidate path of the previous round are segmented to obtain each first segmentation point corresponding to the planned path of the previous round and each second segmentation point corresponding to the candidate path. Calculate the segmentation distance between each first segmentation point and its corresponding second segmentation point, and use the sum of all segmentation distances as the path distance.

5. The method according to claim 2, characterized in that, A devaluation function is constructed based on the niche radius and the path distance, including: When the path distance is greater than the niche radius, the devaluation function is a preset value; If the path distance is less than or equal to the niche radius, calculate the difference between the niche radius and the path distance, and construct the depreciation function based on the ratio of the difference to the niche radius.

6. The method according to claim 1, characterized in that, Based on the simulated annealing particle swarm optimization algorithm and the objective function of this round, the planned path for this round is determined from each candidate path, including: For each particle in the particle swarm, obtain the position of each particle; calculate the objective function value corresponding to each particle based on the objective function of this round and the position of each particle; The optimal position of the particle swarm and the optimal position of each particle are updated according to the objective function value corresponding to each particle; the optimal position of the particle swarm is the position of the particle with the optimal objective function value, and the optimal position of each particle is the position corresponding to the optimal objective function value of each particle; the position of each particle is updated according to the optimal position of the particle swarm and the optimal position of each particle. Determine whether the iteration has converged. If it has converged, use the candidate path corresponding to the particle at the optimal position of the particle swarm as the planned path for this round.

7. The method according to claim 6, characterized in that, The position of each particle is updated based on the optimal position of the particle swarm and the optimal position of each individual particle, including: Obtain the velocity of each particle; The velocity learning value is calculated based on the optimal position of the particle swarm and the optimal position of each particle. The velocity of each particle is updated based on the learned velocity value; Update the position of each particle based on its updated velocity.

8. A multi-path planning device, characterized in that, include: The module retrieves the start and end points of the path to be planned. The initialization module initializes the particle swarm based on the start and end points of the path to be planned; Each particle in the particle swarm corresponds to a candidate path. The determination module, based on the simulated annealing particle swarm optimization algorithm and the objective function of this round, determines the planned path for this round from each candidate path; The objective function for this round is obtained based on the sequential niche algorithm and the objective function of the previous round.

9. An electronic device, characterized in that, include: One or more processors; Storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-7.

10. A computer-readable medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.

Images