Electric vehicle charging scheduling method based on genetic algorithm and multi-objective optimization

Abstract

This invention discloses an electric vehicle charging scheduling method based on genetic algorithms and multi-objective optimization to solve the comprehensive optimization problem of charging pile selection and power allocation during long-distance travel. The invention constructs a combined weighted TOPSIS model, which integrates objective entropy weights with user preferences based on multiple linear regression and Fisher classification, obtaining combined weights through the minimum entropy principle to calculate the charging pile suitability score. An improved two-stage genetic algorithm, TPGA, is employed to decompose the optimization process into a charging pile combination selection stage and a dynamic charging capacity allocation stage, achieving efficient path planning and resource allocation under constraints of power continuity, battery capacity, and market balance. Experiments show that compared to existing solutions, this invention achieves significant improvements in core indicators such as convergence speed, additional driving distance, waiting time, and total cost, comprehensively improving user satisfaction and charging decision-making efficiency.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation and vehicle-to-everything (IoV) technology, specifically relating to a method and system for selecting intelligent charging services for long-distance electric vehicle (EV) travel. Background Technology

[0002] The widespread adoption of electric vehicles has driven the development of charging service systems. However, in long-distance, cross-regional travel scenarios for electric vehicles, the range of the power battery limits the need for vehicles to be charged multiple times along the route. The rationality of charging scheduling directly affects the user's travel experience.

[0003] Existing charging recommendations mainly fall into two categories: One approach is to recommend charging stations based on distance or price. These solutions typically only consider factors such as distance and cost, making them suitable for short-distance or single-charge decisions. However, they ignore the differences in users' subjective preferences and fail to meet personalized charging needs, leading to reduced user satisfaction.

[0004] Second, multi-objective intelligent charging scheduling methods, while taking into account multiple objectives such as travel time and cost, often use traditional evolutionary algorithms (such as NSGA-II) for integrated solutions. When dealing with the two closely related sub-problems of "charging station sequence selection" and "charging amount allocation at each station", they suffer from problems such as huge search space, low convergence efficiency, and easy getting trapped in local optima.

[0005] Meanwhile, existing technologies lack dynamism and precision in quantifying user subjective preferences, failing to effectively integrate objective indicators with subjective preferences; furthermore, they do not adequately consider real-time status information of charging piles in the vehicle-to-everything (V2X) environment (such as queue length and price changes), the weight allocation of multiple indicators lacks scientific rigor, and there is still considerable room for improvement in the convergence speed of the algorithm and the stability of the optimization results.

[0006] Therefore, there is an urgent need to design a long-distance charging planning method that can integrate dynamic user preferences, efficiently and collaboratively optimize charging station sequences and charging volume, and comprehensively improve the performance of core user experience indicators, thereby addressing the aforementioned shortcomings in existing technologies. Summary of the Invention

[0007] The purpose of this invention is to provide an electric vehicle charging scheduling method based on genetic algorithms and multi-objective optimization. By integrating multi-criteria decision-making and evolutionary algorithm techniques, a personalized charging strategy optimization system covering single-point selection to path planning is constructed, which is suitable for efficient resource allocation in long-distance cross-regional electric vehicle travel scenarios. This method addresses the challenges of the lack of personalized adaptation of single-point charging schemes and the failure to achieve coordinated optimization of charging pile combination and charging volume allocation in multi-point charging strategies during long-distance cross-regional electric vehicle travel. Through a collaborative method of the charging pile suitability calculation model (CWT) based on combined weight TOPSIS and the TPGA genetic algorithm, a closed-loop decision-making process is achieved, which includes user preference quantification, charging pile sequence selection, and dynamic power allocation.

[0008] To achieve the above-mentioned objectives, the present invention provides the following technical solution: an electric vehicle charging scheduling method based on genetic algorithms and multi-objective optimization, comprising the following steps: S1) Travel information acquisition: Acquire long-distance travel information for electric vehicles, including the origin, destination, initial battery level, and user preference information. S2) Calculate key indicators of charging piles: Based on long-distance travel information, determine the baseline route and candidate charging pile set, and calculate the key indicators of each candidate charging pile, including additional driving distance, expected waiting time and total charging cost. S3) Construct a combined weighted TOPSIS model: integrate objective weights and subjective weights to calculate the applicability score of each candidate charging pile. The objective weights are calculated based on the entropy weight method, and the subjective weights are quantified based on user preferences. S4) Generate charging strategy: Calculate the total charging demand and the number of charging times required. Based on the applicability score of each candidate charging pile, use an improved genetic algorithm to generate a two-stage charging strategy, including the charging pile combination scheme in the first stage and the charging amount allocation scheme of each charging pile in the second stage. S5) Output charging strategy: The generated charging pile combination sequence and corresponding charging amount allocation scheme are returned to the user as the final charging strategy.

[0009] Furthermore, the aforementioned user preference information includes vague language descriptions of additional driving distance, expected waiting time, and total charging cost, corresponding to five levels: not important, not very important, neutral, relatively important, and very important, which are used to quantify subjective weights.

[0010] Furthermore, the objective weights calculated using the entropy weight method in step S3) above include: The objective weights of key indicators are calculated using the entropy weight method. The formula for calculating information entropy is: , , in Let be the information entropy of indicator j, and n be the number of candidate charging piles. Let j be the standardized proportion of index j in sample i. These are the preprocessed, positively standardized values. This is the preprocessed value of the k-th candidate charging pile on the j-th indicator; Using information entropy to derive objective weights The calculation formula is: Where m is the number of key indicators. The objective weight of the k-th indicator. Let be the information entropy of the k-th indicator.

[0011] Furthermore, the subjective weight quantification based on user preferences in step S3) above includes: calculating the group subjective weight using a multiple linear regression model based on the fuzzy preference descriptions provided by users. Fisher classification model is used to calculate individual subjective weights. ; ; in , This represents the regression coefficient of group type X with respect to index j. Indicates the subjective weight of the group. Indicates individual subjective weight, These represent the classification probabilities of an individual user belonging to groups A, B, and C, respectively.

[0012] Furthermore, the integration of objective and subjective weights in step S3 above includes: Using the principle of minimum entropy to fuse objective weights and subjective weight The combined weights used for TOPSIS evaluation are obtained, and the calculation formula is as follows: ; in To minimize information entropy, Entropy weight is the objective weight. For individual subjective weight, The weighted integral is a comprehensive consideration of both objective and subjective factors.

[0013] Furthermore, the calculation of the applicability score in step S3) above includes: The Euclidean distance between each candidate charging pile and the virtual optimal charging pile was calculated using a combined weighted TOPSIS model. Euclidean distance to the worst virtual charging station The applicability score is obtained using the following formula: ; in For applicability score, To determine the distance to the virtual optimal charging station. To determine the distance to the virtual worst-case charging station, For normalized matrix elements, and These are virtual positive and negative ideal solutions, respectively.

[0014] Furthermore, the first-stage charging pile combination scheme in step S4 above specifically includes: S1.1) Encode the charging pile sequence into a first-stage chromosome and sort it in ascending order according to the relative order of the charging piles relative to the baseline path; S1.2) Design a fitness function with the optimization objectives of minimizing total additional driving distance, minimizing total waiting time, and maximizing total applicability, and combine feasibility factors to ensure the continuity of power supply between charging piles; The fitness function is: ;in The fitness score of chromosome i in the first stage. As weight, Total additional driving distance, Total waiting time For total applicability; S1.3) Iterative optimization is performed using elite retention and roulette wheel selection operators, two-point crossover operators, and adaptive mutation operators until convergence is obtained to obtain the optimal charging pile combination scheme. The mutation probability is adaptively adjusted according to the chromosome fitness, and the calculation formula is as follows: ; in For the i-th chromosome, For adaptive mutation probability, The average fitness of the population. These represent the maximum and minimum fitness of the population.

[0015] Furthermore, the charging allocation scheme in the second stage of step S4) above specifically includes: S2.1) Encode the charging allocation ratio as a second-stage chromosome; S2.2) The design uses a fitness function with the optimization objectives of minimizing charging cost and ensuring a positive correlation between charging amount and driving distance, and combines feasibility factors to ensure that the total charging demand is met and that the power is not depleted between segments; The fitness function is: , in The fitness score of chromosome i in the second stage. This indicates the results are sorted in descending order. This indicates the result of sorting in ascending order. The total service cost for the charging strategy represented by chromosome i. For charging capacity and driving range Correlation coefficient; S2.3) Iterative optimization is performed through selection, crossover and mutation operators, and rank normalization is used to match the relationship between the amount of charge and the distance until convergence is obtained to obtain the optimal charge allocation scheme.

[0016] Further, step S4 above includes: The formula for calculating total charging demand is: ,in The expected total charge at the end point. This represents the initial remaining battery power. Baseline energy consumption for the reference path. Energy consumption for detour refueling Total charging demand; The formula for calculating the required number of charging cycles is: ,in For unit charging capacity, The required number of charging cycles.

[0017] The beneficial effects of this invention are as follows: 1. In terms of personalization, precise adaptation to user subjective differences is achieved through fuzzy preferences and combined weights. Regarding performance optimization, TPGA's segmented execution and applicability-driven fitness design effectively improve convergence speed and result stability. In terms of efficiency improvement, the segmentation method and constraint optimization significantly reduce computational burden. Regarding model innovation, the introduction of multi-index fusion and reachability analysis reduces information asymmetry.

[0018] 2. This invention constructs a charging scheme generation method that supports highly personalized, real-time multi-point optimization, and possesses high stability through deep coupling of multi-criteria decision-making and evolutionary algorithms. Its innovation lies in weight fusion, segmented optimization, and applicability-driven approaches, achieving a significant improvement in user satisfaction without substantially increasing computational overhead.

[0019] 3. This invention provides an efficient and personalized charging decision-making solution for scenarios such as long-distance electric vehicle travel, and is expected to promote the evolution of the electric vehicle ecosystem towards a more intelligent and user-friendly direction. Attached Figure Description

[0020] Figure 1 This is a diagram illustrating the overall working architecture of an embodiment of the present invention.

[0021] Figure 2 The following is a detailed description of the application scenarios of the embodiments of the present invention.

[0022] Figure 3 This is a block diagram of the applicability score model system according to an embodiment of the present invention. Detailed Implementation

[0023] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings in the embodiments of the present invention.

[0024] This invention discloses an electric vehicle charging scheduling method based on genetic algorithms and multi-objective optimization, comprising the following steps: S1) Travel Information Acquisition: Acquire long-distance travel information for electric vehicles, including the origin, destination, initial battery level, and user preference information.

[0025] As a preferred embodiment of the present invention, the user preference information includes a vague language description of the additional driving distance, expected waiting time and total charging cost, corresponding to five levels: not important, not very important, average, relatively important and very important, which are used to quantify subjective weight.

[0026] S2) Calculate key indicators of charging piles: Based on long-distance travel information, determine the baseline route and candidate charging pile set, and calculate the key indicators of each candidate charging pile, including additional driving distance, expected waiting time and total charging cost.

[0027] S3) Construct a combined weighted TOPSIS model: integrate objective and subjective weights to calculate the applicability score of each candidate charging station (e.g., ...). Figure 3 As shown in the figure, the objective weights are calculated based on the entropy weight method, and the subjective weights are based on user preference quantification.

[0028] In a preferred embodiment of the present invention, when a user initiates a long-distance charging planning request, the module first performs positive and standardized preprocessing on the indicators: , Objective weight calculation: The entropy weight method assigns different objective weight factors to each indicator based on the degree of indicator variation. (Three components in total) To objectively weigh the importance of different indicators, a charging pile applicability calculation model is constructed. After preprocessing the parameters, normalized results are obtained. The calculation formula is: Calculate the objective weighting factor for each indicator, involving n samples and m indicators, where... Let j be the standardized proportion of index j in sample i. These are the preprocessed, positively standardized values; First, calculate the information entropy of index j. The calculation formula is: Based on this, the objective weight can be derived. The calculation formula is: ,in The objective weight of the k-th indicator. Let be the information entropy of the k-th indicator.

[0029] In a preferred embodiment of the present invention, the subjective weight calculation in step S3) includes: the system pre-trains a multiple linear regression model and a user classification model using relevant data from the sample set. Based on the current user's historical behavior or preset preferences, the classification model categorizes the user into a specific category (such as "time-priority type A", "cost-priority type B", "distance-priority type C"). Subsequently, the relevant models are invoked to calculate the group subjective weights reflecting the user's preference for each indicator. and individual subjective weight The calculation formula is shown below. ; in , This represents the regression coefficient of group type X with respect to index j. Indicates the subjective weight of the group. Indicates individual subjective weight, These represent the classification probabilities of an individual user belonging to groups A, B, and C, respectively.

[0030] In a preferred embodiment of the present invention, step S3) of fusing objective weights and subjective weights includes: fusing objective weights using the principle of minimum entropy. and subjective weight The combined weights used for TOPSIS evaluation are obtained, and the calculation formula is as follows: ; in To minimize information entropy, Entropy weight is the objective weight. For individual subjective weight, The weighted integral is a comprehensive consideration of both objective and subjective factors.

[0031] In a preferred embodiment of the present invention, the calculation of the applicability score in step S3) includes: calculating the Euclidean distance between each candidate charging pile and the virtual optimal charging pile using a combined weighted TOPSIS model. Euclidean distance to the worst virtual charging station The applicability score is obtained using the following formula: ; in For applicability score, To determine the distance to the virtual optimal charging station. To determine the distance to the virtual worst-case charging station, For normalized matrix elements, and These are virtual positive and negative ideal solutions, respectively.

[0032] S4) Generate charging strategy: Calculate the total charging demand and the number of charging times required. Based on the applicability score of each candidate charging pile, use an improved genetic algorithm to generate a two-stage charging strategy, including the charging pile combination scheme in the first stage and the charging amount allocation scheme of each charging pile in the second stage.

[0033] In a preferred embodiment of the present invention, calculating the total charging demand and the required number of charging cycles specifically includes: The formula for calculating total charging demand is: ,in The expected total charge at the end point. This represents the initial remaining battery power. Baseline energy consumption for the reference path. Energy consumption for detour refueling Total charging demand; The formula for calculating the required number of charging cycles is: ,in For unit charging capacity, The required number of charging cycles.

[0034] As a preferred embodiment of the present invention, the core objective of the first-stage charging pile combination scheme is to select a charging station combination of NC quantity from the candidate charging station set CP. (NC is the number of charging cycles calculated by the system based on travel distance and current battery capacity), achieving multi-objective optimization of "reducing total additional driving distance, reducing total charging waiting time, and improving the overall availability of charging stations," while ensuring power accessibility between charging stations, specifically including: S1.1) Encode the charging pile sequence into a first-stage chromosome and sort it in ascending order according to the relative order of the charging piles with respect to the baseline path.

[0035] In a specific embodiment of the present invention, the population size is set as M. The genetic code of a chromosome is represented as follows: ,in Indicates the first The selected chromosome number Each charging station is assigned an ID number. To facilitate subsequent genetic operations and fitness function calculations, each generated chromosome needs to be sorted in ascending order according to the myindex corresponding to the charging station, such as... Figure 2 As shown, myindex is defined as the relative order of the charging piles with respect to path L.

[0036] S1.2) Design a fitness function with the optimization objectives of minimizing total additional travel distance, minimizing total waiting time, and maximizing total applicability.

[0037] As a specific embodiment of the present invention, the fitness function calculation process is as follows: For each chromosome Its fitness calculation is mainly based on three original target factors: total new paths. Total expected charging wait time and the overall applicability of charging piles .in Chromosomes The Each charging station is relative to Additional driving distance, Chromosomes The Expected waiting time for each charging station Chromosomes The The usability rating of each charging station. (Iteration) : ,in It is the total additional driving distance. This is the total expected waiting time. It is the overall applicability.

[0038] To eliminate the influence of different dimensions of the optimization objective and promote the construction of the fitness function through positive operations, it is necessary to consider the relevant objective factors. Preprocessing is performed. and They represent the original target factors of the chromosome. The upper and lower limits, This indicates that the charging pile set CP is categorized by index. The value of the k-th element in ascending order, for =1, 2, 3: After normalization, each fitness can be achieved through Calculation, traversal ,for =1, 2, 3: , The corresponding weights The calculation formula is as follows: , , To ensure ultimate feasibility, any requirement that cannot be met should have its corresponding fitness set to 0.

[0039] S1.3) Iterative optimization is performed using elite retention, roulette wheel selection operators, two-point crossover operators, and adaptive mutation operators until convergence yields the optimal charging pile combination scheme. In the selection operation, the top 10% of chromosomes with the highest fitness in the population are directly introduced into the next generation to ensure the continuation of high-quality genes; the remaining 90% of individuals are selected using a roulette wheel selection method, with the selection probability of each chromosome individual i proportional to its fitness. The calculation formula is as follows: In the crossover operation, the following steps are performed in a certain proportion: two parent chromosomes A and B are randomly selected, and after being processed by the crossover function, offspring chromosomes C and D are obtained.

[0040] The mutation probability is adaptively adjusted based on chromosome fitness. For chromosomes with fitness higher than the population average, a lower mutation probability is set to reduce damage to high-quality genes; for chromosomes with fitness lower than the average, a higher mutation probability is set to increase diversity, thereby achieving the goal of escaping local optima. During the mutation process, a gene locus is randomly selected, and the corresponding charging pile is replaced with a charging pile in the candidate set CP that has not been selected by the chromosome. The formula for calculating the mutation probability adaptively adjusted based on chromosome fitness is as follows: ; in For the i-th chromosome, For adaptive mutation probability, The average fitness of the population. These represent the maximum and minimum fitness of the population. If the number of iterations reaches a preset threshold... Or the difference in optimal fitness across multiple generations is less than a set threshold. If the condition is met, the iteration terminates and the corresponding optimal combination of charging piles for the chromosome is output; otherwise, return to S1.2 and proceed to the next iteration.

[0041] To obtain the optimal combination of charging stations Afterwards, this stage aims to "reduce charging costs" by combining battery capacity, power continuity, and market balance constraints to solve for the optimal charging capacity allocation strategy for each charging station. ,in Indicates allocation to charging stations The charging capacity ratio.

[0042] As a preferred embodiment of the present invention, the second-stage charging allocation scheme specifically includes: S2.1) encoding the charging allocation ratio as a second-stage chromosome, with the population size set to M. Chromosome The genetic code is represented as: .

[0043] S2.2) For each chromosome Its fitness calculation is primarily based on two original target factors: total charging service fee. And the positive correlation between the charging capacity of NC charging cycles and the energy consumption of the corresponding driving distance. . (A, B) represents the correlation coefficient between vectors A and B. (Generally, the larger the value of an element at a certain position in A, the larger the value of the corresponding element at that position in B, indicating a stronger positive correlation.) The larger the value of (A, B), the better. Indicated by chromosomes The resulting vector, The driving distance corresponding to the number of NC charging cycles is defined by the following formula: ,in Chromosomes The The distance of each charging station relative to L. Indicates projection to The distance between adjacent nodes corresponding to deviation points.

[0044] The design employs a fitness function with the optimization objectives of minimizing charging cost and ensuring a positive correlation between charging amount and driving distance. This is combined with a feasibility factor to ensure that total charging demand is met and that the battery is not depleted between driving segments. The fitness function is: ,in The fitness score of chromosome i in the second stage. This indicates the results are sorted in descending order. This indicates the result of sorting in ascending order. The total service cost for the charging strategy represented by chromosome i. For charging capacity and driving range Correlation coefficient; finally, based on the highest fitness of the current generation, determine whether an update is needed. .

[0045] S2.3) Iterative optimization is performed using selection, crossover, and mutation operators, employing rank normalization to match the relationship between charge amount and distance until convergence yields the optimal charge allocation scheme. Most operations in this step are the same as in S1.3). Accordingly, Replace with .

[0046] If the number of iterations reaches a preset threshold Or the difference in optimal fitness across multiple generations is less than a set threshold. If the iteration terminates, the corresponding optimal chromosome is output. Otherwise, return to S2.2 and proceed to the next iteration.

[0047] S5) Output charging strategy: The generated charging pile combination sequence and corresponding charging amount allocation scheme are returned to the user as the final charging strategy.

[0048] This invention effectively solves the complex charging planning problem in the long-distance travel scenario of electric vehicles by combining genetic algorithms and the combined weight TOPSIS model, achieving efficient and highly personalized intelligent recommendations and significantly improving the user experience.

Claims

1. A method for scheduling electric vehicle charging based on genetic algorithms and multi-objective optimization, characterized in that, The method includes: S1) Travel information acquisition: Acquire long-distance travel information for electric vehicles, including the origin, destination, initial battery level, and user preference information. S2) Calculate key indicators of charging piles: Based on long-distance travel information, determine the baseline route and candidate charging pile set, and calculate the key indicators of each candidate charging pile, including additional driving distance, expected waiting time and total charging cost. S3) Construct a combined weighted TOPSIS model: integrate objective weights and subjective weights to calculate the applicability score of each candidate charging pile. The objective weights are calculated based on the entropy weight method, and the subjective weights are quantified based on user preferences. S4) Generate charging strategy: Calculate the total charging demand and the number of charging times required. Based on the applicability score of each candidate charging pile, use an improved genetic algorithm to generate a two-stage charging strategy, including the charging pile combination scheme in the first stage and the charging amount allocation scheme of each charging pile in the second stage. S5) Output charging strategy: Return the generated charging pile combination sequence and corresponding charging amount allocation scheme to the user as the final charging strategy.

2. The electric vehicle charging scheduling method based on genetic algorithm and multi-objective optimization according to claim 1, characterized in that, The user preference information includes vague language descriptions of additional driving distance, expected waiting time, and total charging cost, corresponding to five levels: not important, not very important, average, relatively important, and very important, which are used to quantify subjective weights.

3. The electric vehicle charging scheduling method based on genetic algorithm and multi-objective optimization according to claim 1, characterized in that, The objective weight calculation based on the entropy weight method in step S3) includes: The objective weights of key indicators are calculated using the entropy weight method. The formula for calculating information entropy is: ； in Let be the information entropy of indicator j, and n be the number of candidate charging piles. Let j be the standardized proportion of index j in sample i. These are the preprocessed, positively standardized values. This is the preprocessed value of the k-th candidate charging pile on the j-th indicator; The objective weight is derived using information entropy, and the calculation formula is as follows: ; Where m represents the number of key indicators. The objective weight of the k-th indicator. Let be the information entropy of the k-th indicator.

4. The electric vehicle charging scheduling method based on genetic algorithm and multi-objective optimization according to claim 1, characterized in that, The subjective weight quantification based on user preference in step S3) includes: Based on the fuzzy preference descriptions provided by users, a multiple linear regression model was used to calculate the group's subjective weights. Fisher classification model is used to calculate individual subjective weights. ; ； in , This represents the regression coefficient of group type X with respect to index j. Indicates the subjective weight of the group. Indicates individual subjective weight, These represent the classification probabilities of an individual user belonging to groups A, B, and C, respectively.

5. The electric vehicle charging scheduling method based on genetic algorithm and multi-objective optimization according to claim 1, characterized in that, The integration of objective weights and subjective weights in step S3) includes: Using the principle of minimum entropy to fuse objective weights and subjective weight The combined weights used for TOPSIS evaluation are obtained, and the calculation formula is as follows: ； in To minimize information entropy, Entropy weight is the objective weight. For individual subjective weight, The weighted integral is a comprehensive consideration of both objective and subjective factors.

6. The electric vehicle charging scheduling method based on genetic algorithm and multi-objective optimization according to claim 1, characterized in that, The calculation of the applicability score in step S3) includes: The Euclidean distance between each candidate charging pile and the virtual optimal charging pile was calculated using a combined weighted TOPSIS model. Euclidean distance to the worst virtual charging station The applicability score is obtained using the following formula: ； in For applicability score, To determine the distance to the virtual optimal charging station. To determine the distance to the virtual worst-case charging station, For normalized matrix elements, and These are virtual positive and negative ideal solutions, respectively.

7. The electric vehicle charging scheduling method based on genetic algorithm and multi-objective optimization according to claim 1, characterized in that, The first phase of the charging pile combination scheme specifically includes: S1.1) Encode the charging pile sequence into a first-stage chromosome and sort it in ascending order according to the relative order of the charging piles relative to the baseline path; S1.2) Design a fitness function with the optimization objectives of minimizing total additional driving distance, minimizing total waiting time, and maximizing total applicability, and combine feasibility factors to ensure the continuity of power supply between charging piles; The fitness function is: ;in The fitness score of chromosome i in the first stage. As weight, Total additional driving distance, Total waiting time For total applicability; S1.3) Iterative optimization is performed using elite retention and roulette wheel selection operators, two-point crossover operators, and adaptive mutation operators until convergence is obtained to obtain the optimal charging pile combination scheme. The mutation probability is adaptively adjusted according to the chromosome fitness, and the calculation formula is as follows: ； in For the i-th chromosome, For adaptive mutation probability, The average fitness of the population. These represent the maximum and minimum fitness of the population.

8. The electric vehicle charging scheduling method based on genetic algorithm and multi-objective optimization according to claim 1, characterized in that, The second phase of the charging allocation scheme specifically includes: S2.1) Encode the charging allocation ratio as a second-stage chromosome; S2.2) The design uses a fitness function with the optimization objectives of minimizing charging cost and ensuring a positive correlation between charging amount and driving distance, and combines feasibility factors to ensure that the total charging demand is met and that the power is not depleted between segments; The fitness function is: ,in The fitness score of chromosome i in the second stage. This indicates the results are sorted in descending order. This indicates the result of sorting in ascending order. The total service cost for the charging strategy represented by chromosome i. For charging capacity and driving range Correlation coefficient; S2.3) Iterative optimization is performed through selection, crossover and mutation operators, and rank normalization is used to match the relationship between the amount of charge and the distance until convergence is obtained to obtain the optimal charge allocation scheme.

9. The electric vehicle charging scheduling method based on genetic algorithm and multi-objective optimization according to claim 1, characterized in that, Step S4) includes: The formula for calculating total charging demand is: ,in The expected total charge at the end point. This represents the initial remaining battery power. Baseline energy consumption for the reference path. Energy consumption for detour refueling Total charging demand; The formula for calculating the required number of charging cycles is: ,in For unit charging capacity, The required number of charging cycles.

Images