A Charge / Discharge Planning Method for Electric Vehicles in Large-Scale V2G Scenarios Based on Constrained Evolution

By employing the EPR-COE method and utilizing population restart strategy and guided initialization, the evolutionary dilemma in EV charging and discharging planning in large-scale V2G scenarios is solved, improving solution efficiency and planning quality.

CN120493725BActive Publication Date: 2026-01-30HUNAN UNIV OF SCI & TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510585279.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-05-08
Publication Date
2026-01-30
Estimated Expiration
2045-05-08

AI Technical Summary

Technical Problem

In large-scale V2G scenarios, existing evolutionary algorithms are prone to getting stuck in evolutionary dilemmas, resulting in low efficiency in solving trolley charging and discharging planning, making it difficult to meet the requirements of high reliability and stability.

Method used

We adopt the EPR-COE (Evolutionary Constraint-Based Planning) method for trolley charging and discharging in large-scale V2G scenarios. By constructing an optimization model and constraints, we optimize the trolley charging and discharging planning using a population restart strategy, multi-level restart timing estimation, and guided initialization strategy.

Benefits of technology

It improves the algorithm's solution efficiency in large-scale V2G systems, enhances the quality and search efficiency of trolley charging and discharging planning, and reduces the impact of evolutionary dilemmas.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120493725B_ABST
    Figure CN120493725B_ABST
Patent Text Reader

Abstract

This invention addresses the problem of planning the charging and discharging behavior of electric vehicles in large-scale V2G scenarios, and designs a constrained evolution-based method for this purpose. The method employs a population restart strategy that includes a multi-level restart timing estimation method and a multi-scenario restart population generation method, while also incorporating a guided initialization strategy. The guided initialization strategy accelerates the population's search within the feasible region, while the population restart strategy enables the algorithm to accurately estimate restart timing at a lower cost and provides targeted solutions for different types of evolutionary dilemmas, thereby improving the algorithm's solution efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of energy management, and in particular relates to a method for planning the charging and discharging of electric vehicles in a large-scale V2G scenario based on constrained evolution. Technical Background

[0002] With the rapid increase in social electricity demand, various renewable energy technologies have been widely applied. However, new energy technologies, such as photovoltaic, wind, and tidal power, are limited by natural resources, resulting in large fluctuations and low flexibility in power supply. Furthermore, they lack energy storage capabilities, making it difficult to meet the high reliability and stability requirements of smart grids. Thanks to the widespread adoption of electric vehicles, vehicle-to-grid (V2G) technology has received considerable attention and research. V2G technology essentially transforms large-scale electric vehicles (EVs) from "load terminals" into distributed "energy storage units," supporting bidirectional energy exchange between EVs and the grid, thereby achieving goals such as peak shaving and valley filling, and enhancing grid regulation capabilities.

[0003] With the increasing number of EVs connected to the power grid, the scheduling objects of V2G systems exhibit characteristics such as large scale, high dimensionality, and strong constraints. Solving the overall charging and discharging planning problem for EVs faces challenges such as feasible region shrinkage and discretization, significant objective conflicts, and high computational complexity. Evolutionary algorithms, due to their excellent global search capabilities and adaptability to complex optimization problems, are widely used in V2G scenarios for EV charging and discharging planning optimization. However, when performing charging and discharging planning for scenarios with large EV scales, such as large logistics parks, stations, and ports, evolutionary algorithms are prone to getting stuck in infeasible regions or encountering the "evolutionary dilemma" of premature convergence and search stagnation in the feasible region, seriously affecting the solution efficiency and the quality of the planning scheme. Existing research lacks attention to the algorithm mechanism itself, and still mainly relies on traditional optimization algorithms for solving the problem, making it difficult to respond quickly and improve the evolutionary dilemma. Therefore, it is urgent to propose an optimization algorithm that can effectively cope with the evolutionary dilemma and improve search efficiency to adapt to the charging and discharging planning needs of EVs in large-scale V2G scenarios. Summary of the Invention

[0004] To achieve the above objectives, this invention employs a constrained evolution-based electric vehicle charging and discharging planning method (EPR-COE) for large-scale V2G scenarios, comprising the following steps:

[0005] Step 1: Construct an optimization model for the electric vehicle charging and discharging planning problem in a large-scale V2G scenario;

[0006] Step 2: Use the constrained evolutionary algorithm based on population restart to solve the problem model and calculate the start time and duration of charging and discharging for each trolleybus.

[0007] The optimization model in step 1 specifically includes:

[0008] The objective function for optimizing the model is:

[0009]

[0010] Where f1 represents the overall charging cost of the electric vehicle, f2 represents the mean square error of the total load on the power grid after charging and discharging in each time period, ω represents the weighting coefficient; K represents the number of electric vehicles, T represents the number of time periods; Pc k and Pd k Let ct represent the charging power and discharging power of the k-th trolley. k,t and dt k,t p represents the charging and discharging durations of the k-th t-th trolley during the t-th time period. t P represents the electricity price for the t-th time period; t PB represents the total load after charging and discharging in the t-th time period. t This represents the base load before the summation in the t-th time period. ηd represents the average load after summing up all time periods. k This represents the discharge efficiency of the kth trolley.

[0011] Constraints of the optimization model:

[0012]

[0013] Among them, SOC k,min and SOC k,max This represents the lower and upper limits of the allowable battery capacity of the k-th electric vehicle, SOC. k,cs and SOC k,ds ct represents the charge level of the k-th trolley before charging and discharging; k and dt k Let ηc represent the total charging time and total discharging time of the kth trolley. k Represents the charging efficiency of the k-th trolley; SOC k,d The value of l represents the charge level of the kth trolley when it leaves the charging area. k and u k This represents the distance traveled by the kth trolley and its energy consumption per kilometer; ta k and td k cs represents the time when the k-th trolley arrives at and leaves the charging area. k and ds k Indicates the time when the k-th trolley begins charging and discharging; state k This represents the charging state of the kth trolley, and state k ∈{0,1}, a value of 0 indicates that the kth trolley is not charging, and a value of 1 indicates that it is charging; P max N represents the maximum allowable charging load of the charging area. maxThis indicates the number of charging stations in the charging area.

[0014] Step 2 specifically includes the following steps:

[0015] Step 2.1: Input the basic parameters of the problem model and the basic parameters of the constrained evolutionary algorithm;

[0016] Step 2.2: Encode the solution variables;

[0017] Step 2.3: Generate the initial population P = (x1,...,x) using a guided initialization strategy. NP )';

[0018] Step 2.4: Calculate the fitness value f(x) of all individuals in the current population. i ) and default degree G(x) i );

[0019] Step 2.5: Perform mutation and crossover operations to generate a progeny population P' = (y1,...,y NP )', calculate the fitness value f(y) of all individuals in the offspring population. i ) and default rate G(y) i );

[0020] Step 2.6: Perform the selection operation to obtain the next generation population and update the optimal charging and discharging plan x for the electric vehicle. best ;

[0021] Step 2.7: Record the population search traces and calculate the feasible margin probability, elite margin probability, and unexplored margin probability for each region for each decision variable;

[0022] Step 2.8: Execute the multi-level restart timing estimation method;

[0023] Step 2.9: Based on the type of evolutionary dilemma determined in 2.8, perform the corresponding operation in the multi-scenario restart population generation method;

[0024] Step 2.10: Determine if the maximum number of iterations has been reached. If so, end the algorithm and output the optimal planning scheme for charging and discharging the electric vehicle; otherwise, proceed to step 2.4.

[0025] Step 2.7 specifically includes the following steps:

[0026] Step 2.7.1: Divide the search range of each decision variable into s regions;

[0027] Step 2.7.2: Record the number of times each region of each decision variable is explored, the number of times a feasible solution appears, and the number of times an elite solution appears;

[0028] Step 2.7.3: Calculate the feasible marginal probability, elite marginal probability, and unexplored marginal probability for each region of each decision variable.

[0029] The multi-level restart timing estimation method in step 2.8 specifically includes the following steps:

[0030] Step 2.8.1, calculate the current iteration number g and the boundary point G. b The relationship between the sizes of g and G; if g < G b If the condition is met, then the algorithm proceeds to step 2.8.2; otherwise, step 2.10 is executed.

[0031] Step 2.8.2, calculate the expected value ρ for whether the population has been trapped in the feasible or infeasible region for a long time. The formula is as follows:

[0032]

[0033] Where FR∈{0,1}, FR=0 indicates that all planning schemes are infeasible, and FR=1 indicates that all planning schemes are feasible; N1 represents the consecutive algebras with FR=0, and N2 represents the consecutive algebras with FR=1; if the restart expectation is met and FR=0, the algorithm executes the restart population generation method for the population trapped in the infeasibility region dilemma; if the restart expectation is met and FR=1, the algorithm executes step 2.8.3; if the restart expectation is not met or the value of FR is neither 0 nor 1, step 2.10 is executed.

[0034] Step 2.8.3, calculate the evolutionary rate OR(g) of the population's optimal solution, which is calculated as follows:

[0035]

[0036] Compare OR(g) with the evolution rate OR(g-1) of the previous generation's optimal solution, and use N3 to record the number of consecutive generations in which the evolution rate decreases; when N3 reaches a threshold If the condition is met, then the algorithm proceeds to step 2.8.4; otherwise, step 2.10 is executed.

[0037] Step 2.8.4, calculate the population density PD, the formula is:

[0038]

[0039]

[0040] in, Let represent the mean of all individuals for the j-th decision variable; when PD is less than the threshold λ, the algorithm executes the restart population generation method to address the premature convergence dilemma in the feasible region; otherwise, it executes the restart population generation method to address the search stagnation dilemma in the feasible region.

[0041] The beneficial effects of this invention are reflected in the following aspects: Based on the differential evolution algorithm and multi-objective constraint processing technology, a population restart strategy is used, which includes a multi-level restart timing estimation method and a multi-scenario restart population generation method, while a guided initialization strategy is also employed. The multi-level restart timing estimation method balances the accuracy and computational cost of restart timing estimation, while the multi-scenario restart population generation method utilizes population search traces to implement targeted population restart measures for three evolutionary dilemmas, thereby improving the algorithm's efficiency in solving the trolley charging and discharging planning problem in large-scale V2G systems. Compared to random initialization, the initial population generated by the guided initialization strategy better matches the characteristics of the trolley charging and discharging planning problem in V2G systems and accelerates the population's entry into the feasible region search. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of a scenario according to an embodiment of the present invention;

[0043] Figure 2 This is a diagram of the EPR-COE algorithm architecture according to an embodiment of the present invention;

[0044] Figure 3 This is a graph showing the convergence curve of the default degree in the ablation experiment of the guided initialization strategy according to an embodiment of the present invention.

[0045] Figure 4 This is a graph showing the convergence curve of the fitness value of the comparison algorithm in an embodiment of the present invention;

[0046] Figure 5 This is a comparison chart of the load optimization results of the comparison algorithm in an embodiment of the present invention. Detailed Implementation

[0047] To provide a clearer understanding of the technical features, objectives, and beneficial effects of this invention, the following detailed description of the technical solution is provided. Obviously, the described embodiments are only a portion of the embodiments of this invention, not all of them, and should not be construed as limiting the scope of implementation of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the protection scope of this invention.

[0048] This invention addresses the problem that existing evolutionary algorithms for solving trolley charging and discharging planning in V2G scenarios are prone to evolutionary dilemmas due to the large number of trolleys and numerous constraints. It proposes a large-scale trolley charging and discharging planning method based on constraint evolution, EPR-COE.

[0049] A schematic diagram of the research problem scenario is shown below. Figure 1 As shown in the diagram, the algorithm architecture is as follows: Figure 2As shown. The specific implementation method is described in three parts: the first part is the description of the simulation experiment scenario and the setting of parameters; the second part is the specific application process of the method; and the third part is the experimental results.

[0050] The first part is: description of the simulation experiment scenario and parameter settings.

[0051] The experimental scenario of this invention is a logistics park charging area equipped with 100 V2G charging piles, with a maximum charging power of 2000kW allowed at the same time. It needs to provide charging and discharging services for 300 electric vehicles from 8:00 to 19:00 and for 200 electric trucks from 19:00 to 8:00 the next day.

[0052] In the experimental scenario of this invention, the performance parameters of the electric vehicle and the electric truck are shown in the table below:

[0053]

[0054] The second part describes the specific application process of the method, including the following steps:

[0055] Step 1: Construct an optimization model for the electric vehicle charging and discharging planning problem in a large-scale V2G scenario, specifically including the objective function and constraints:

[0056] The objective function for optimizing the model is:

[0057]

[0058] Where f1 represents the overall charging cost of the electric vehicle, f2 represents the mean square error of the total load after charging and discharging of the power grid in each time period, ω represents the weighting coefficient, and ω∈(0,1); K represents the number of electric vehicles, T represents the number of time periods in a day, which is set to 24 in this invention; Pc k and Pd k Let ct represent the charging power and discharging power of the k-th trolley. k,t and dt k,t p represents the charging and discharging durations of the k-th t-th trolley during the t-th time period. t P represents the electricity price for the t-th time period; t PB represents the total load after charging and discharging in the t-th time period. t This represents the base load before the summation in the t-th time period. ηd represents the average load after summing up all time periods. k This represents the discharge efficiency of the kth trolley.

[0059] Constraints of the optimization model:

[0060]

[0061] Among them, SOCk,min and SOC k,max This represents the lower and upper limits of the allowable battery capacity of the k-th electric vehicle, SOC. k,cs and SOC k,ds ct represents the charge level of the k-th trolley before charging and discharging; k and dt k Let ηc represent the total charging time and total discharging time of the kth trolley. k Represents the charging efficiency of the k-th trolley; SOC k,d The value of l represents the charge level of the kth trolley when it leaves the charging area. k and u k This represents the distance traveled by the kth trolley and its energy consumption per kilometer; ta k and td k cs represents the time when the k-th trolley arrives at and leaves the charging area. k and ds k Indicates the time when the k-th trolley begins charging and discharging; state k This represents the charging state of the kth trolley, and state k ∈{0,1}, a value of 0 indicates that the kth trolley is not charging, and a value of 1 indicates that it is charging; P max N represents the maximum allowable charging load of the charging area. max This indicates the number of charging stations in the charging area.

[0062] Step 2: Solve the optimization model using a constrained evolutionary algorithm based on population restart, calculating the start time and duration of charging and discharging for each trolleybus. For example... Figure 2 The provided algorithm architecture diagram includes the following steps:

[0063] Step 2.1: Input the basic parameters of the problem model and the basic parameters of the constrained evolutionary algorithm;

[0064] Step 2.2: Encode the solution variables;

[0065] Step 2.3: Generate the initial population P = (x1,...,x) using a guided initialization strategy. NP )';

[0066] Step 2.4: Calculate the fitness value f(x) of all individuals in the current population. i ) and default degree G(x) i );

[0067] Step 2.5: Perform mutation and crossover operations using the differential evolution algorithm to generate the offspring population P' = (y1,...,y NP )', calculate the fitness value f(y) of all individuals in the offspring population. i ) and default rate G(y) i );

[0068] Step 2.6: Perform a selection operation using multi-objective constraint processing techniques to obtain the next generation of the population and update the optimal trolley charging and discharging plan x. best ;

[0069] Step 2.7: Record the population search traces and calculate the feasible margin probability, elite margin probability, and unexplored margin probability for each region for each decision variable;

[0070] Step 2.8: Execute the multi-level restart timing estimation method;

[0071] Step 2.9: Based on the type of evolutionary dilemma determined in 2.8, perform the corresponding operation in the multi-scenario restart population generation method;

[0072] Step 2.10: Determine if the maximum number of iterations has been reached. If so, end the algorithm and output the optimal trolley charging and discharging plan. Otherwise, proceed to step 2.4.

[0073] In section 2.1, input all the parameters required for the problem model; input the population size NP and the maximum number of iterations g. max The maximum number of evaluations is MaxFEs. Here, the population size is NP = 100, and the maximum number of iterations is g. max =20000, Maximum number of evaluations MaxFEs=2000000.

[0074] In section 2.2, the solution variables are encoded, x i It can be represented as x i =(cs i,1 ,ct i,1 ,ds i,1 ,dt i,1 ,...cs i,K ,ct i,K ,ds i,K ,dt i,K This represents a charging and discharging planning scheme for electric vehicles. Where cs... i,k ,ct i,k ,ds i,k ,dt i,k Let c and s represent the start time of charging, total charging time, start time of discharging, and total discharging time of the kth trolley in the i-th individual, respectively, with values ​​ranging from c to s. i,k ∈[ta k ,td k ), ct i,k ∈[0,td k -ta k ), ds i,k ∈[ta k ,td k ), dti,k ∈[0,td k -ta k ).

[0075] In section 2.3, the current iteration number g = 1 is set, and an initial population P = (x1,...,x) is generated using a guided initialization strategy. NP In this strategy, the charging and discharging order of the electric vehicles is first determined. If the electric vehicle arrives during off-peak hours, it is charged first and then discharged; if it arrives during peak hours, it is discharged first and then charged. If the k-th electric vehicle in the i-th individual is charged first and then discharged: firstly, two times are randomly selected from the current off-peak period and the subsequent adjacent off-peak periods as the charging start times cs. i,k The total charging time (ct) is the time between the charging completion time and the charging end time. i,k Then, two moments are randomly selected from the longest peak segment as the discharge start time ds. i,k The total discharge duration dt is the time between the discharge end and the discharge termination time. i,k For the case of discharging before charging: two moments are randomly selected during this peak period as the discharge start time ds. i,k The total discharge duration dt is the time between the discharge end and the discharge termination time. i,k Then, within the longest consecutive off-peak period, two moments are randomly selected as the charging start times (cs). i,k The total charging time (ct) is the time between the charging completion time and the charging end time. i,k .

[0076] In section 2.4, the x value for each individual in the population is calculated. i The fitness value f(x) in the electric vehicle charging and discharging planning problem model i ) and default rate G(x) i ).

[0077] In 2.5, for each individual x i Generate a random number between 0 and 1, rand. If rand < g / G max The DE / rand-to-best / 1 / bin operator is used to perform population mutation and crossover operations, generating a progeny population P' = (y1,...,y NP Otherwise, use the DE / current-to-rand / 1 operator to perform population mutation and crossover operations, generating a progeny population P' = (y1,...,y) NP )'. Calculate the y of each individual in the offspring population. i The fitness value f(y) in the electric vehicle charging and discharging planning problem model i ) and default rate G(y) i ).

[0078] In 2.6, for each individual x in the parent population i and the corresponding individual y in the offspring population i Through formula g ws Calculate g(x) = λf(x) + (1-λ)G(x) respectively. ws (x i |λ i ) and g ws (y i |λ i If g ws (y i |λ i )≤g ws (x i |λ i If ), then use y i Replace x i ,f(y i ) replace f(x) i ), G(y i ) replace G(x i ). Where λ i =i×η / NP, The values ​​of Γ and α are 30 and 0.75, respectively. If g ws (y i |λ i )≤g ws (x best |λ i If ), then use y. i Replace x best ,f(y i ) replace f(x best ), G(y i ) replace G(x best ).

[0079] In section 2.7, the calculation of the marginal probabilities for each region of each decision variable includes the following three steps:

[0080] In section 2.7.1, the search range for each decision variable is first uniformly divided into s regions. l max and u max These represent the lower and upper bounds of the search, respectively, corresponding to the decision variable with the broadest search scope.

[0081] In section 2.7.2, variables RF(j,k) and RI(j,k) represent the number of times the j-th feasible and infeasible variables appear in the k-th region of the dimension up to the present generation, respectively; RE(j,k) represents the number of times the j-th variable appears in the k-th region, RE(j,k) = RI(j,k) + RF(j,k); RB(j,k) represents the number of times the j-th variable appears in the k-th region among the top 25% of elite individuals in the present generation.

[0082] In section 2.7.3, the feasible edge probability FP(j,k), elite edge probability BP(j,k), and unexplored edge probability UP(j,k) of the j-th dimension and k-th region are calculated as follows:

[0083]

[0084]

[0085]

[0086]

[0087] In version 2.8, the multi-level restart timing estimation method includes the following four steps:

[0088] In 2.8.1, the first-level estimation is performed, calculating the current iteration number g and the boundary point G. b The relationship between the sizes of G, where G b =0.8×g max When g < G b If the condition is met, algorithm 2.8.2 is executed; otherwise, 2.10 is executed.

[0089] In section 2.8.2, the second-level estimation is performed. First, the expected value ρ for determining whether the population has been trapped in the feasible or infeasible region for a long period is calculated. The formula is as follows:

[0090]

[0091] Where FR∈{0,1}, FR=0 indicates that all planning schemes are infeasible, and FR=1 indicates that all planning schemes are feasible; N1 represents the consecutive algebras with FR values ​​of 0, N2 represents the consecutive algebras with FR values ​​of 1, and α=0.05; a random number rand between 0 and 1 is generated. If rand<ρ and FR is 0, the algorithm immediately executes the restart population generation method to deal with the population being trapped in the infeasibility region dilemma; if rand<ρ and FR is 1, the algorithm executes 2.8.3; if the restart expectation is not met or the value of FR is neither 0 nor 1, then 2.10 is executed.

[0092] In section 2.8.3, the third-level estimation is performed by first calculating the evolutionary rate OR(g) of the population optimum, which is calculated as follows:

[0093]

[0094] Compare OR(g) with the evolution rate OR(g-1) of the previous generation's optimal solution, and use N3 to record the number of consecutive generations in which the evolution rate decreases. When N3 reaches a threshold... When that happens, algorithm 2.8.4 is executed, where... The value is equal to the population size NP; if N3 does not reach the threshold, then execute 2.10.

[0095] In section 2.8.4, the fourth-level estimation is performed to calculate the population density PD, which is calculated as follows:

[0096]

[0097]

[0098] in, Let represent the mean of the j-th individual. When PD is less than the threshold λ, the algorithm immediately executes the method to restart population generation to address the premature convergence dilemma in the feasible region; otherwise, the algorithm immediately executes the method to restart population generation to address the search stagnation dilemma in the feasible region. Where λ = 0.5.

[0099] In version 2.9, based on the type of predicament the algorithm is stuck in, as determined in version 2.8, the corresponding method for restarting population generation is executed. Specifically, there are three types:

[0100] When a population is trapped in an infeasible region, the probability p(j,k) of the j-th dimension variable of the restarted individual appearing in the k-th region is calculated as p(j,k)=UP(j,k);

[0101] For the population to fall into the premature convergence dilemma within the feasible region, the probability p(j,k) of the j-th dimension variable of the restarted individual appearing in the k-th region is calculated as follows:

[0102] For a population stuck in a search stagnation dilemma within the feasible region, the probability p(j,k) of the j-th dimension variable of the restarted individual appearing in the k-th region is calculated as follows:

[0103] In 2.10, if the current iteration number g = g max Then exit the algorithm and input the optimal planning solution x. best Otherwise, increment the value of g by 1 and proceed to step 2.4.

[0104] The third part is: display of simulation experiment results and result analysis.

[0105] Simulation results (see attached figures) Figure 3 , Figure 4 , Figure 5 .

[0106] exist Figure 3 In the initial population generated by EPR-COE in this invention, the optimal default value is much lower than the optimal default value of EPR-COE-rand using random initialization. Furthermore, in the 15000th generation (FEs are 1.5 × 10⁻⁶), EPR-COE's default value is significantly lower. 6 A feasible solution has been found around 25000 generations (FEs are 2.5 × 10⁻⁶). 6 It took approximately 10 minutes to find a feasible solution. Compared to random initialization, the guided initialization strategy proposed in this invention can accelerate the population's entry into the feasible region for searching.

[0107] exist Figure 4 In this study, the EPR-COE algorithm outperformed other comparative algorithms in terms of optimal fitness. Specifically, it improved by 58.7% compared to IUDE; by 69.2% compared to ECO-HCT; and by 55.5% compared to DeCODE.

[0108] exist Figure 5 In this study, the EPR-COE algorithm outperformed other comparative algorithms in terms of load variance across all time periods. Specifically, it improved by 68.9% compared to the baseline load; 69.0% compared to IUDE; 68.7% compared to ECO-HCT; and 44.2% compared to DeCODE.

[0109] This invention proposes a method called EPR-COE and successfully applies it to the electric vehicle charging and discharging planning problem in large-scale V2G scenarios. Comparison with several advanced algorithms demonstrates that the EPR-COE algorithm has excellent capabilities in handling such problems.

[0110] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for electric vehicle charging and discharging planning in a large-scale V2G scenario based on constrained evolution, characterized in that, The method comprises the following two steps: Step 1, constructing an optimization model of electric vehicle charging and discharging planning under a large-scale V2G scenario; Step 2, solving the problem model by using a constraint evolutionary algorithm based on population restart to calculate the start time and duration of charging and discharging of each electric vehicle; The objective function of the optimization model in the step 1 comprises: wherein f1 represents the overall charging cost of the electric vehicles, f2 represents the mean square deviation of the aggregated charging and discharging load of the power grid in each time period, ω represents the weight coefficient; K represents the number of electric vehicles, T represents the number of time periods; Pc k and Pd k represent the charging power and discharging power of the kth electric vehicle, ct k,t and dt k,t represent the charging duration and discharging duration of the kth electric vehicle in the tth time period, p t represents the electricity price in the tth time period; P t represents the aggregated charging and discharging load in the tth time period, PB t represents the basic load before the aggregation of charging and discharging in the tth time period, represents the mean value of the aggregated charging and discharging load in all time periods, ηd k represents the discharging efficiency of the kth electric vehicle; The constraint conditions of the optimization model comprise: where SOC k,min and SOC k,max denote the lower and upper limits of the allowable electric quantity of the kth electric vehicle, SOC k,cs and SOC k,ds denote the electric quantity of the kth electric vehicle before charging and discharging; ct k and dt k denote the total charging time and total discharging time of the kth electric vehicle, ηc k denotes the charging efficiency of the kth electric vehicle; SOC k,d denotes the electric quantity of the kth electric vehicle when leaving the charging area, l k and u k denote the driving distance and power consumption per kilometer of the kth electric vehicle; ta k and td k denote the time when the kth electric vehicle arrives at and leaves the charging area, cs k and ds k denote the time when the kth electric vehicle starts charging and discharging; state k denotes the charging state of the kth electric vehicle, and state k ∈{0, 1}, with a value of 0 indicating that the kth electric vehicle does not perform charging, and a value of 1 indicating that it is charging; P max denotes the maximum charging load allowed by the charging area, N max denotes the number of charging piles of the charging area.

2. The method according to claim 1, wherein, The step 2 specifically comprises the following sub-steps: Step 2.1, inputting basic parameters of the problem model and basic parameters of the constraint evolutionary algorithm; Step 2.2, encoding the solving variables; Step 2.

3. Generate initial population P = (x1,..., x NP ) T ; Step 2.4, calculate the fitness f(x i ) and the degree of violation G(x i ) of all individuals in the current population; Step 2.

5. Perform mutation and crossover operations to generate offspring population P' = (y1,...,y NP ) T , compute fitness value f(y i ) and violation degree G(y i ) for all individuals in offspring population P'; Step 2.6, perform selection operation to obtain next generation population, update the optimal planning scheme of electric vehicle charging and discharging x best ; Step 2.7, recording the search traces of the population, and calculating the feasible edge probability, elite edge probability and unexplored edge probability of each decision variable in each region; Step 2.8, executing a multi-level restart opportunity estimation method; Step 2.9, according to the evolution difficulty type judged in 2.8, executing corresponding operations in the multi-scenario restart population generation method; Step 2.10, judging whether the maximum iteration number is reached, if yes, ending the algorithm and outputting the optimal charging and discharging planning scheme of the electric vehicle, otherwise, executing 2.

4.

3. The method of claim 2, wherein the method is based on constrained evolution. The step 2.7 specifically comprises the following sub-steps: Step 2.7.1, uniformly dividing the search range of each decision variable into s regions; Step 2.7.2, recording the number of times each region of each decision variable is explored, the number of times a feasible solution appears, and the number of times an elite solution appears; Step 2.7.3, calculating the feasible edge probability, elite edge probability and unexplored edge probability of each region of each decision variable.

4. The method according to claim 3, wherein, The multi-level restart opportunity estimation method in the step 2.8 specifically comprises the following sub-steps: Step 2.8.1, compute the size relationship between the current iteration number g and the break point G b If g < G b then the algorithm performs 2.8.2, otherwise it performs step 2.10; Step 2.8.2, calculating whether the population has been trapped in the feasible domain or the infeasible domain for a long time, and the calculation formula is as follows: Wherein, FR is 0, FR is 1, N1 represents the continuous algebra of FR value being 0, and N2 represents the continuous algebra of FR value being 1; if the restart expectation is met and FR is 0, the algorithm executes the restart population generation method for the population trapped in the infeasible domain; if the restart expectation is met and FR is 1, the algorithm executes 2.8.3; if the restart expectation is not met or the value of FR is neither 0 nor 1, step 2.10 is executed; Step 2.8.3, calculating the evolution rate OR(g) of the population optimal solution, and the calculation formula is as follows: Compare OR(g) with the size of the optimal solution evolutionary rate of the last generation OR(g-1), and record the number of consecutive generations of evolutionary rate decline with N3; when N3 reaches the threshold , then the algorithm performs 2.8.4, otherwise step 2.10 is performed; Step 2.8.4, calculating the population density PD, and the calculation formula is as follows: wherein, represents the mean of all individuals of the jth decision variable; when PD is less than a threshold value λ, the algorithm performs a restart population generation method for a population trapped in a premature convergence dilemma in the feasible region; otherwise, it performs a restart population generation method for a population trapped in a search stagnation dilemma in the feasible region.

Citation Information

Patent Citations

  • Asynchronous parallel hyper-heuristic method for large-scale optimization problem

    CN117056039A

  • Ordered charging and discharging control method for electric vehicle based on improved DBO algorithm

    CN117996754A