A constrained multi-modal multi-objective path optimization method based on dynamic sorting

Through dynamic sorting and co-evolution strategies, the constraints and convergence of paths are balanced, and multiple equivalent optimal paths are generated, which solves the path optimization problems under multi-objectives and constraints in the existing technology, and provides a more suitable solution for practical applications. .

CN117689092BActive Publication Date: 2024-11-26NINGBO UNIV
View PDF 8 Cites 0 Cited by

Patent Information

Application Number
CN202311482987.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-11-09
Publication Date
2024-11-26
Estimated Expiration
2043-11-09

AI Technical Summary

Technical Problem

Existing path optimization methods usually only consider the length of the optimal path and ignore the constraint feature information of the path, making it difficult to provide feasible and convergent multiple equivalent optimal paths under multi-objective and constraint conditions.

Method used

Using a constrained multimodal multi-objective path optimization method based on dynamic sorting, the constrained multimodal multi-objective path optimization method is obtained, and the path population can be collaboratively searched according to the preset co-evolution strategy and weights to balance the constraint quality and convergence quality. , iterate out the optimization target path that meets the constraints.

Benefits of technology

It effectively solves the path optimization problem under multi-objective and constraints, provides multiple equivalent optimal paths, balances the feasibility of the path and population convergence, and is suitable for practical applications.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117689092B_ABST
    Figure CN117689092B_ABST
Patent Text Reader

Abstract

The present invention discloses a constrained multi-modal multi-objective path optimization method based on dynamic sorting, which relates to the field of path optimization, generates a first path population and a second path population including N connected path individuals, determines a first weight and a second weight for balancing the conflict between constraint quality and convergence quality under the current number of iterations, and performs collaborative iteration on the first path population and the second path population according to a preset co-evolution strategy, a first weight and a second weight. The strategy is set according to the objective function and constraint conditions of the model to be solved to iterate out a first path population that better satisfies the optimization goal of the objective function on the basis of satisfying the constraint conditions. When the preset iteration termination condition is reached, it is determined that the N path individuals in the first path population at this time are N equivalent optimal paths. The present application balances the conflict between the feasibility and convergence of the path, provides multiple equivalent optimal paths for decision makers to choose, and is conducive to actual path optimization applications.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of path optimization, and in particular to a constrained multi-modal multi-objective path optimization method based on dynamic sorting. Background Art

[0002] The path optimization problem is closely related to daily life and industrial production, such as path optimization of mobile robots, path optimization of vehicles, and path planning of aircraft. In this problem, usually only one optimal path is considered, but in practical applications, the optimal path may be impassable for some reasons, affecting transportation efficiency. It can be seen that it is necessary to provide multiple equivalent optimal paths for decision makers to choose from. In addition, most of the current methods only consider the length of the optimal path as the primary goal, ignoring some constraint feature information of the path, such as the optimal path needs to pass through a certain location, the length of the congested area of ​​the route is the shortest, and other constraints. There is currently no effective solution to this problem that requires achieving multiple target requirements under multiple constraints. Therefore, how to provide an effective path optimization solution for this situation is a problem that needs to be solved urgently. Summary of the invention

[0003] The problem solved by the present invention is to provide a constrained multi-modal multi-objective path optimization method based on dynamic sorting, which effectively solves the path optimization problem under multiple objectives and constraints, and balances the conflict between solving constraints, that is, the feasibility and convergence of the path, and provides multiple equivalent optimal paths for decision makers to choose, which is more conducive to practical application.

[0004] To solve the above problems, the present invention provides a constrained multi-modal multi-objective path optimization method based on dynamic sorting, comprising:

[0005] Obtain the model to be solved;

[0006] Generate a first path population and a second path population including N connectable path individuals for the model to be solved, where N is an integer greater than 1;

[0007] Determine a first weight and a second weight for balancing the conflict between constraint quality and convergence quality at the current number of iterations;

[0008] Performing collaborative search iteration on the first path population and the second path population according to a preset collaborative evolution strategy, the first weight, and the second weight, wherein the preset collaborative evolution strategy is set according to the objective function and constraint conditions of the model to be solved to iterate out a first path population that further satisfies the optimization goal of the objective function on the basis of satisfying the constraint conditions;

[0009] When a preset iteration termination condition is reached, it is determined that the N path individuals in the first path population are N equivalent optimal paths representing the solution of the model to be solved.

[0010] The beneficial effects of the present invention are as follows: the present application effectively solves the path optimization problem under multiple objectives and constraints, and balances the conflict between solving constraints, namely, the feasibility of the path and the convergence of the population, and provides multiple equivalent optimal paths for decision makers to choose from, which is more conducive to practical application.

[0011] Further, the step of determining a first weight for balancing the conflict between constraint quality and convergence quality at the current number of iterations includes:

[0012] Determine a first weight for balancing the conflict between constraint quality and convergence quality according to the current number of iterations and a first preset relationship;

[0013] The first preset relationship is:

[0014]

[0015] Wherein, ω1 is the first weight, g is the current number of iterations, G Max The preset maximum number of iterations.

[0016] Further, the step of determining a second weight for balancing the conflict between constraint quality and convergence quality at the current number of iterations includes:

[0017] Determine a second weight for balancing the conflict between constraint quality and convergence quality according to the current number of iterations and a second preset relationship;

[0018] The second preset relationship is:

[0019]

[0020] Wherein, ω2 is the second weight.

[0021] In this solution, through the above settings, the first weight can be reduced as the current number of iterations increases, and the second weight can be increased as the current number of iterations increases, which is conducive to balancing the conflict between the feasibility of the path and the convergence of the population.

[0022] Further, performing collaborative search iteration on the first path population and the second path population according to a preset collaborative evolution strategy, the first weight, and the second weight includes:

[0023] S21: using an intelligent optimization algorithm to respectively determine the first offspring of the first path population and the second offspring of the second path population at the current number of iterations;

[0024] S22: determining a first constraint ranking value corresponding to each path individual in a first combined population according to a first dynamic ranking strategy, the first weight, and the second weight, so as to determine a new first path population according to the first constraint ranking value and the objective function, wherein the first combined population includes the first path population, the first offspring, and the second offspring;

[0025] S23: determining a second constraint ranking value corresponding to each path individual in a second combined population according to a second dynamic ranking strategy, the first weight, and the second weight, so as to determine a new second path population according to the second constraint ranking value and the objective function, wherein the second combined population includes the second path population, the first offspring, and the second offspring;

[0026] S24: Determine whether the current number of iterations reaches a preset maximum number of iterations; if so, proceed to the step of determining whether the N path individuals in the first path population are N equivalent optimal paths representing the solution of the model to be solved; if not, proceed to S25;

[0027] S25: Set the current number of iterations = the current number of iterations + 1, and proceed to the step of determining the first weight and the second weight for balancing the conflict between the constraint quality and the convergence quality under the current number of iterations.

[0028] In this scheme, the first path population and the second path population are co-evolved. When the objective function is to minimize the objective function, the first constraint sorting values ​​are sorted in order from small to large and selected to obtain a new first path population, and the second constraint sorting values ​​are sorted in order from small to large and selected to obtain a new second path population.

[0029] Further, determining the first constraint ranking value corresponding to each path individual in the first combined population according to the first dynamic ranking strategy, the first weight and the second weight includes:

[0030] For each path individual in the first combined population, perform the following steps:

[0031] Determine a first local convergence value corresponding to the path individual;

[0032] Determine a first local constraint value corresponding to the path individual;

[0033] Determine a first constraint ranking value corresponding to the path individual according to a third preset relationship;

[0034] The third preset relationship is:

[0035] F1(x)=ω1×F cons (x)+ω2×F conv (x)

[0036] Wherein, F1(x) is the first constraint ranking value, x is the path individual, ω1 is the first weight, and F cons (x) is the first local constraint value, ω2 is the second weight, F conv (x) is the first local convergence value.

[0037] In this solution, the above settings can make the first constraint ranking value focus on the local constraint quality first and then the local convergence quality during the iteration process, which is beneficial to the co-evolution iteration.

[0038] Further, determining a first local convergence value corresponding to the path individual includes:

[0039] Determine a first local convergence value corresponding to the path individual according to a fourth preset relationship;

[0040] The fourth preset relationship is:

[0041]

[0042] Among them, F conv (x) is the first local convergence value, P1 is the first combined population, y is any path individual in the first combined population except x, R conv (x, y) is determined based on a fifth preset relationship;

[0043] The fifth preset relationship is:

[0044]

[0045] Among them, x cv is the constraint violation degree corresponding to x, y cv is the degree of constraint violation corresponding to y.

[0046] Further, determining the first local constraint value corresponding to the path individual includes:

[0047] Determine a first local constraint value corresponding to the path individual according to a sixth preset relationship;

[0048] The sixth preset relationship is:

[0049]

[0050] Among them, F cons (x) is the first local constraint value, R cons (x, y) is determined based on the seventh preset relationship;

[0051] The seventh preset relationship is:

[0052]

[0053] Among them, x cv is the constraint violation degree corresponding to x, y cv is the degree of constraint violation corresponding to y.

[0054] Further, determining the second constraint ranking value corresponding to each path individual in the second combined population according to the second dynamic ranking strategy, the first weight and the second weight includes:

[0055] For each path individual in the second combined population, perform the following steps:

[0056] Determine a second local convergence value corresponding to the path individual;

[0057] Determine a second local constraint value corresponding to the path individual;

[0058] Determine a second constraint ranking value corresponding to the path individual according to an eighth preset relationship;

[0059] The eighth preset relationship is:

[0060] F2(x)=ω2×F cons (x)+ω1×F conv (x)

[0061] Among them, F2(x) is the second constraint ranking value.

[0062] In this solution, the above settings can make the second constraint ranking value focus on the local convergence quality first and then the local constraint quality during the iteration process, which is beneficial to the co-evolution iteration. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 A flowchart of a constrained multi-modal multi-objective path optimization method based on dynamic sorting provided by the present invention. DETAILED DESCRIPTION

[0064] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0065] Please refer to Figure 1 , Figure 1 A flowchart of a constrained multi-modal multi-objective path optimization method based on dynamic sorting provided by the present invention.

[0066] The present invention provides an embodiment, a constrained multi-modal multi-objective path optimization method based on dynamic sorting, comprising:

[0067] S11: Obtain the model to be solved;

[0068] S12: generating a first path population and a second path population including N connected path individuals for the model to be solved, where N is an integer greater than 1;

[0069] S13: Determine a first weight and a second weight for balancing the conflict between constraint quality and convergence quality at the current number of iterations;

[0070] S14: performing collaborative search iteration on the first path population and the second path population according to a preset collaborative evolution strategy, a first weight, and a second weight, wherein the preset collaborative evolution strategy is set according to the objective function and constraint conditions of the model to be solved to iterate out a first path population that better satisfies the optimization goal of the objective function on the basis of satisfying the constraint conditions;

[0071] S15: When a preset iteration termination condition is reached, it is determined that the N path individuals in the first path population are N equivalent optimal paths representing a solution to the model to be solved.

[0072] It should be noted that the present application does not impose any particular limitation on the specific form of the model to be solved. As an example, the constrained multi-modal multi-objective path optimization method based on dynamic sorting can be applied to the path optimization of urban networks. The buildings that cannot be passed through in the urban network are pre-calibrated, and the urban roads and road congestion areas are clearly defined. The road intersection areas are calibrated as intersections. On this basis, assuming that you are currently going from a certain starting point A to a certain end point B, the position that must be passed along the way is called the constraint point C. Then the model to be solved established at this time can be: the first objective function is from the starting point The path from A to the end point B is the shortest, and the second objective function is to have the fewest intersections from the starting point A to the end point B; the first constraint condition is that the number of traffic lights at the intersections from the starting point A to the end point B that can be passed directly without waiting is less than m, and the second constraint condition is that the constraint point C must be passed from the starting point A to the end point B, and the number of constraint points C is at least one; after determining the model to be solved, all intersections in the map corresponding to the urban network are numbered in sequence from left to right and from top to bottom, and then N equivalent optimal paths can be obtained according to the above steps for decision makers to choose.

[0073] It should also be noted that the connectivity in S12 means that the path individual is a valid and actually passable path; the specific method of generating the first path population and the second path population includes but is not limited to random initialization, which is not particularly limited here.

[0074] In addition, the current number of iterations gradually increases with the collaborative search iteration process, and the preset iteration termination condition may be to determine whether the current number of iterations reaches a preset maximum number of iterations.

[0075] In summary, the present application provides a constrained multi-modal and multi-objective path optimization method based on dynamic sorting, which effectively solves the path optimization problem under multiple objectives and constraints, and balances the conflict between solving constraints, namely, the feasibility of the path and the convergence of the population, and provides multiple equivalent optimal paths for decision makers to choose from, which is more conducive to practical application.

[0076] As a preferred embodiment, the step of determining a first weight for balancing the conflict between constraint quality and convergence quality at the current number of iterations includes:

[0077] Determine a first weight for balancing the conflict between constraint quality and convergence quality according to the current number of iterations and a first preset relationship;

[0078] The first preset relationship is:

[0079]

[0080] Among them, ω1 is the first weight, g is the current iteration number, G Max The preset maximum number of iterations.

[0081] Specifically, the current number of iterations increases continuously as the iteration process continues. The specific value of the preset maximum number of iterations is not particularly limited here and can be set according to actual needs.

[0082] As a preferred embodiment, the step of determining a second weight for balancing the conflict between constraint quality and convergence quality at the current number of iterations includes:

[0083] Determine a second weight for balancing the conflict between constraint quality and convergence quality according to the current number of iterations and a second preset relationship;

[0084] The second preset relationship is:

[0085]

[0086] Among them, ω2 is the second weight.

[0087] Specifically, through the above settings, it can be achieved that the first weight decreases as the current number of iterations increases, and the second weight increases as the current number of iterations increases, which is conducive to balancing the conflict between the feasibility of the path and the convergence of the population.

[0088] As a preferred embodiment, the first path population and the second path population are collaboratively searched and iterated according to the preset collaborative evolution strategy, the first weight and the second weight, including:

[0089] S21: using an intelligent optimization algorithm to respectively determine the first offspring of the first path population and the second offspring of the second path population at the current number of iterations;

[0090] S22: determining a first constraint ranking value corresponding to each path individual in the first combined population according to the first dynamic ranking strategy, the first weight, and the second weight, so as to determine a new first path population according to the first constraint ranking value and the objective function, wherein the first combined population includes the first path population, the first offspring, and the second offspring;

[0091] S23: determining a second constraint ranking value corresponding to each path individual in the second combined population according to the second dynamic ranking strategy, the first weight and the second weight, so as to determine a new second path population according to the second constraint ranking value and the objective function, wherein the second combined population includes the second path population, the first offspring and the second offspring;

[0092] S24: Determine whether the current number of iterations reaches the preset maximum number of iterations; if so, proceed to the step of determining whether the N path individuals in the first path population are N equivalent optimal paths representing the solution of the model to be solved; if not, proceed to S25;

[0093] S25: Set the current number of iterations = the current number of iterations + 1, and proceed to the step of determining the first weight and the second weight for balancing the conflict between the constraint quality and the convergence quality under the current number of iterations.

[0094] Specifically, the intelligent optimization algorithm here includes but is not limited to a genetic algorithm, which is not particularly limited here; after determining the first constraint ranking value corresponding to each path individual in the first combined population, if the objective function is to minimize the objective function, the first constraint ranking value is sorted in ascending order, and the first N path individuals after sorting are selected as the new first path population to ensure that higher quality path individuals can be preserved; after determining the second constraint ranking value corresponding to each path individual in the second combined population, if the objective function is to minimize the objective function, the second constraint ranking value is sorted in ascending order, and the first N path individuals after sorting are selected as the new second path population to ensure that higher quality path individuals can be preserved.

[0095] It can be seen that the present application performs a co-evolutionary search through the first path population and the second path population, taking into account the constraints and multi-objective functions. The constraints define the feasibility of the path. The first path population and the second path population both consider convergence and constraints at the same time, balancing the conflict between path feasibility and population convergence. The determination method of the new first path population and the new second path population always ensures that the path individuals with better quality are selected and saved in the corresponding path population.

[0096] As a preferred embodiment, determining the first constraint sorting value corresponding to each path individual in the first combined population according to the first dynamic sorting strategy, the first weight and the second weight includes:

[0097] For each path individual in the first combined population, perform the following steps:

[0098] Determine the first local convergence value corresponding to the path individual;

[0099] Determine the first local constraint value corresponding to the path individual;

[0100] Determine a first constraint ranking value corresponding to the path individual according to a third preset relationship;

[0101] The third preset relationship is:

[0102] F1(x)=ω1×F cons (x)+ω2×F conv (x)

[0103] Among them, F1(x) is the first constraint ranking value, x is the path individual, ω1 is the first weight, and F cons (x) is the first local constraint value, ω2 is the second weight, F conv (x) is the first local convergence value.

[0104] Specifically, the first constraint ranking value is determined by considering the constraint priority, and the dynamic reduction and F cons (x) The corresponding constraint weight (the first weight will decrease as the number of current iterations increases) and the dynamic increase with F conv (x) The corresponding convergence weight (the second weight will increase with the increase of the current number of iterations) is used to balance the constraints and convergence, so that the first constraint ranking value first focuses on the local constraint quality and then the local convergence quality during the iteration process. The balance between convergence and constraints is achieved through the information exchange between the first path population and the second path population. Therefore, in the constrained multi-modal multi-objective path optimization, more feasible and equivalent optimal paths can be searched under the conditions of considering constraints and convergence.

[0105] As a preferred embodiment, determining the first local convergence value corresponding to the path individual includes:

[0106] Determine a first local convergence value corresponding to the path individual according to a fourth preset relationship;

[0107] The fourth preset relationship is:

[0108]

[0109] Among them, F conv (x) is the first local convergence value, P1 is the first combined population, y is any path individual in the first combined population except x, R conv (x, y) is determined based on a fifth preset relationship;

[0110] The fifth preset relationship is:

[0111]

[0112] Among them, x cv is the constraint violation degree corresponding to x, y cv is the degree of constraint violation corresponding to y.

[0113] As a preferred embodiment, determining the first local constraint value corresponding to the path individual includes:

[0114] Determine a first local constraint value corresponding to the path individual according to a sixth preset relationship;

[0115] The sixth preset relationship is:

[0116]

[0117] Among them, F cons (x) is the first local constraint value, R cons (x, y) is determined based on the seventh preset relationship;

[0118] The seventh preset relationship is:

[0119]

[0120] Among them, x cv is the constraint violation degree corresponding to x, y cv is the degree of constraint violation corresponding to y.

[0121] As a preferred embodiment, determining the second constraint sorting value corresponding to each path individual in the second combined population according to the second dynamic sorting strategy, the first weight and the second weight includes:

[0122] For each path individual in the second combined population, perform the following steps:

[0123] Determine the second local convergence value corresponding to the path individual;

[0124] Determine the second local constraint value corresponding to the path individual;

[0125] Determine the second constraint ranking value corresponding to the path individual according to the eighth preset relationship;

[0126] The eighth preset relationship is:

[0127] F2(x)=ω2×F cons (x)+ω1×F conv (x)

[0128] Wherein, F2(x) is the second constraint ranking value.

[0129] Specifically, the method for determining the second local convergence value corresponding to the path individual is the same as the method in the fourth preset relationship and the fifth preset relationship. It should be noted that at this time y∈P2, P2 is the second combined population; the method for determining the second local constraint value corresponding to the path individual is the same as the method in the sixth preset relationship and the seventh preset relationship. It should be noted that at this time y∈P2.

[0130] It can be seen that the second constraint ranking value is determined by taking into account the convergence priority, and by dynamically increasing the value of F cons (x) The corresponding constraint weight (the second weight will decrease as the number of current iterations increases) and the dynamic reduction and F conv (x) The corresponding convergence weight (the first weight will decrease as the current number of iterations increases) is used to balance the constraints and convergence, so that the second constraint ranking value focuses on the local convergence quality first and then the local constraint quality during the iteration process. The balance between convergence and constraints is achieved through the information exchange between the first path population and the second path population. Therefore, in the constrained multi-modal multi-objective path optimization, more feasible and equivalent optimal paths can be searched under the conditions of considering constraints and convergence.

[0131] Although the disclosure is disclosed as above, the protection scope of the disclosure is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the disclosure, and these changes and modifications will fall within the protection scope of the present invention.

[0132] It should also be noted that, in this specification, relational terms such as first, second, etc. are merely used to distinguish one entity or operation from another entity or operation, and do not imply any actual relationship or order between these entities or operations.

Claims

1. A constrained multi-modal multi-objective path optimization method based on dynamic sorting, It is characterized in that include: Obtain the model to be solved and apply it to the path optimization of the urban network. Pre-mark the impassable buildings in the urban network, identify the urban roads and road congestion areas, and mark the road intersection areas as intersections. After determining the model to be solved, number all the intersections in the map corresponding to the urban network from left to right and from top to bottom. For the model to be solved, a first path population and a second path population are generated, which include N connected path individuals from a certain starting point A to a certain end point B, where the positions that must be passed along the way are called constraint points C, where N is an integer greater than 1; Determine a first weight and a second weight for balancing the conflict between the constraint quality and the convergence quality at the current number of iterations; wherein the step of determining the first weight for balancing the conflict between the constraint quality and the convergence quality at the current number of iterations includes: Determine a first weight for balancing the conflict between constraint quality and convergence quality according to the current number of iterations and a first preset relationship; The first preset relationship is: ; in, is the first weight, is the current iteration number, is the preset maximum number of iterations; The step of determining a second weight for balancing the conflict between constraint quality and convergence quality at the current iteration number includes: Determine a second weight for balancing the conflict between constraint quality and convergence quality according to the current number of iterations and a second preset relationship; The second preset relationship is: ; in, is the second weight; Performing collaborative search iteration on the first path population and the second path population according to a preset collaborative evolution strategy, the first weight, and the second weight, wherein the preset collaborative evolution strategy is set according to the objective function and constraint conditions of the model to be solved to iterate out a first path population that further satisfies the optimization goal of the objective function on the basis of satisfying the constraint conditions; The step of performing collaborative search iteration on the first path population and the second path population according to a preset collaborative evolution strategy, the first weight, and the second weight includes: S21: using an intelligent optimization algorithm to respectively determine the first offspring of the first path population and the second offspring of the second path population at the current number of iterations; S22: determining a first constraint ranking value corresponding to each path individual in a first combined population according to a first dynamic ranking strategy, the first weight, and the second weight, so as to determine a new first path population according to the first constraint ranking value and the objective function, wherein the first combined population includes the first path population, the first offspring, and the second offspring; S23: determining a second constraint ranking value corresponding to each path individual in a second combined population according to a second dynamic ranking strategy, the first weight, and the second weight, so as to determine a new second path population according to the second constraint ranking value and the objective function, wherein the second combined population includes the second path population, the first offspring, and the second offspring; S24: Determine whether the current number of iterations reaches a preset maximum number of iterations; if so, proceed to the step of determining whether the N path individuals in the first path population are N equivalent optimal paths representing the solution of the model to be solved; if not, proceed to S25; S25: Set the current number of iterations = the current number of iterations + 1, and proceed to the step of determining a second weight for balancing the conflict between constraint quality and convergence quality under the current number of iterations; The objective function setting includes: the first objective function is to minimize the path from the starting point A to the end point B, and the second objective function is to minimize the number of intersections from the starting point A to the end point B; The setting of the constraint conditions includes: the first constraint condition is that the number of traffic lights at the intersections from the starting point A to the end point B that can be directly passed without waiting is less than m, and the second constraint condition is that the starting point A to the end point B must pass through the constraint point C, and the number of constraint points C is at least one; When the preset iteration termination condition is reached, N equivalent optimal paths are obtained for the decision maker to choose, and it is determined that the N path individuals in the first path population are the N equivalent optimal paths representing the solution of the model to be solved.

2. The constrained multi-modal multi-objective path optimization method based on dynamic sorting as claimed in claim 1, It is characterized in that Determining a first constraint sorting value corresponding to each path individual in the first combined population according to the first dynamic sorting strategy, the first weight, and the second weight includes: For each path individual in the first combined population, perform the following steps: Determine a first local convergence value corresponding to the path individual; Determine a first local constraint value corresponding to the path individual; Determine a first constraint ranking value corresponding to the path individual according to a third preset relationship; The third preset relationship is: ; in, sorting values ​​for the first constraint, is the path individual, is the first weight, is the first local constraint value, is the second weight, is the first local convergence value.

3. The constrained multi-modal multi-objective path optimization method based on dynamic sorting as claimed in claim 2, It is characterized in that Determining a first local convergence value corresponding to the path individual includes: Determine a first local convergence value corresponding to the path individual according to a fourth preset relationship; The fourth preset relationship is: ; in, is the first local convergence value, P 1 is the first combined population, y is any path individual in the first combined population except x, Determined based on the fifth preset relationship; the fifth preset relationship is: ; in, is the degree of constraint violation corresponding to x, is the degree of constraint violation corresponding to y.

4. The constrained multi-modal multi-objective path optimization method based on dynamic sorting as claimed in claim 2, It is characterized in that Determining a first local constraint value corresponding to the path individual includes: Determine a first local constraint value corresponding to the path individual according to a sixth preset relationship; The sixth preset relationship is: ; in, is the first local constraint value, Determined based on the seventh preset relationship; y is any path individual in the first combined population except x; P 1 is the first combined population; The seventh preset relationship is: ; in, is the degree of constraint violation corresponding to x, is the degree of constraint violation corresponding to y.

5. The constrained multi-modal multi-objective path optimization method based on dynamic sorting as claimed in claim 1, It is characterized in that Determining a second constraint sorting value corresponding to each path individual in the second combined population according to the second dynamic sorting strategy, the first weight, and the second weight includes: For each path individual in the second combined population, perform the following steps: Determine a second local convergence value corresponding to the path individual; Determine a second local constraint value corresponding to the path individual; Determine the second constraint ranking value corresponding to the path individual according to an eighth preset relationship; The eighth preset relationship is: ; in, Sort values ​​for the second constraint, is the first local constraint value, is the first local convergence value.

Citation Information

Patent Citations

  • Path planning method based on constraint multi-target particle swarm optimization and related equipment

    CN114971078A

  • Multi-equivalent optimal path planning method considering actual multiple constraints

    CN115146866A

  • Constraint multi-objective optimization method based on two-stage division cooperation

    CN115526101A

  • Vehicle path planning method based on hybrid grey wolf particle swarm optimization

    CN116029639A

  • Path planning method and device, vehicle and storage medium

    CN116952234A