A multi-objective optimization-oriented electric vehicle charging scheduling method

By constructing a vehicle access mechanism and a linear weighted objective function, and combining genetic algorithms and simulated annealing algorithms to optimize electric vehicle charging time, the problem of the imbalance between profit and satisfaction in electric vehicle charging scheduling is solved, achieving a win-win situation for parking lot operators and car owners.

CN114742284BActive Publication Date: 2024-11-26FUJIAN NORMAL UNIV +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210306082.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-03-25
Publication Date
2024-11-26
Estimated Expiration
2042-03-25

AI Technical Summary

Technical Problem

Existing electric vehicle charging scheduling strategies fail to effectively balance the profits of parking lot operators and the satisfaction of car owners, which may lead to the impact on power grid security and unmet vehicle charging needs when large-scale electric vehicle access is implemented.

Method used

A multi-objective optimization method for electric vehicle charging scheduling is adopted. By constructing objective functions for maximizing vehicle access mechanism, linear weighted operator profit, and maximizing owner satisfaction, and combining genetic algorithm and simulated annealing algorithm, the charging time allocation is optimized.

Benefits of technology

This approach achieves a win-win situation by increasing parking lot operators' profits and car owner satisfaction while ensuring power grid security, meeting vehicle charging needs, and mitigating power distribution network load fluctuations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114742284B_ABST
    Figure CN114742284B_ABST
Patent Text Reader

Abstract

The present invention provides a multi-objective optimization-oriented electric vehicle charging scheduling method, comprising the steps of: constructing a vehicle access mechanism to restrict vehicles entering a parking lot; linearly weighting the two optimization objectives of maximizing operator profits and maximizing owner satisfaction to obtain the final objective function of each individual vehicle entering the parking lot for charging; randomly allocating the charging time of individual vehicles to obtain the charging allocation plan of individual vehicles; using a genetic algorithm to update the charging allocation plan, and combining a simulated annealing algorithm to determine whether to receive the updated charging allocation plan; and outputting the final charging allocation plan of the individual vehicle. The present invention can balance the profits of parking lot operators and the satisfaction of car owners, obtain the optimal charging allocation plan for vehicles, and can simultaneously improve the profits of operators and the satisfaction of car owners, while effectively alleviating the load fluctuation of the distribution network, achieving a win-win situation for parking lot operators and electric vehicle owners.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging, and in particular to an electric vehicle charging scheduling method oriented to multi-objective optimization. Background Art

[0002] As fossil energy shortages and environmental pollution become increasingly serious, electric vehicles have attracted widespread attention from the government, industry and academia due to their energy-saving and environmentally friendly features. They are gradually becoming popular and widely used in daily life, and are expected to replace traditional fuel vehicles in the future. At the same time, the use of electric vehicles has also put forward new requirements for public charging facilities.

[0003] At present, many electric vehicle owners are unable to install charging infrastructure in residential parking lots, and their charging needs cannot be met. To alleviate this problem, parking lots will likely serve as charging stations in the future, that is, charging their electric vehicles in parking lots near their workplaces. By equipping parking lots with multiple charging piles and providing parking and charging services, car owners can charge during parking. At the same time, since car owners' working hours are usually longer than the charging time of electric vehicles, the charging completion rate can be greatly improved. It is expected that parking lots in office buildings, commercial areas and airports can adopt this model in the future. Therefore, parking lots need to provide convenient and fast charging services for electric vehicles. However, due to the dynamic arrival and random charging characteristics of electric vehicles, if appropriate coordinated charging technology is not used, the access of large-scale electric vehicles will not only affect the safety of the distribution network, such as increased load peaks, overload of distribution network lines and increased distribution network losses, but may also cause the charging needs of vehicles to be unable to be met when leaving, resulting in profit losses for parking lot operators. Therefore, how to improve the charging efficiency of large-scale electric vehicles in parking lot environments is particularly important. In order to alleviate the above adverse effects, it is crucial to design a charging scheduling strategy and apply an efficient scheduling algorithm to ensure the interests of parking lot operators and electric vehicle owners.

[0004] There are two stakeholders involved in the process of electric vehicle charging scheduling, namely, charging service providers and electric vehicle owners. On the one hand, from the perspective of charging service providers, in order to ensure the safety of the power grid, it is necessary to achieve peak load shaving as much as possible, so it is hoped that the rationality of each vehicle's charging period will be considered while achieving maximum profit. On the other hand, from the perspective of car owners, in order to avoid the situation where the expected power cannot be reached before leaving, the enthusiasm of car owners to participate in the coordination of charging should be mobilized, the satisfaction of car owners should be guaranteed, the charging time should be shortened as much as possible, and the charging fees they need to pay should be reduced. However, the scheduling strategies currently used usually take charging profits and car owner satisfaction as the goals of two different stakeholders, and rarely consider these two goals comprehensively. The expansion of the electric vehicle market share has put forward new requirements for the charging scheduling strategy of electric vehicles, and current research work has not yet formed a complete and effective solution for profit and satisfaction. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an electric vehicle charging scheduling method oriented to multi-objective optimization, which balances the profits of parking lot operators and the satisfaction of car owners under the premise of ensuring the safety of the power grid.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a multi-objective optimization-oriented electric vehicle charging scheduling method, comprising the steps of:

[0007] S1. Establish a vehicle access mechanism to restrict vehicles entering the parking lot;

[0008] S2. Linearly weight the two optimization objectives of maximizing operator profit and maximizing owner satisfaction to obtain the final objective function of each individual vehicle entering the parking lot for charging;

[0009] S3. Randomly assigning charging time for the individual vehicles to obtain a charging allocation plan for the individual vehicles;

[0010] S4, using a genetic algorithm to update the charging allocation plan, and combining a simulated annealing algorithm to determine whether to accept the updated charging allocation plan;

[0011] S5. Output the final charging allocation plan for the individual vehicle.

[0012] The beneficial effects of the present invention are as follows: the present invention provides an electric vehicle charging scheduling method for multi-objective optimization. First, in order to ensure that the vehicle can complete the charging task before leaving and meet the charging needs of more vehicles as much as possible, a vehicle access mechanism is proposed. Secondly, linear weighting is used to comprehensively consider the two optimization goals of maximizing the operator's profit and maximizing the owner's satisfaction, and the final objective function of the vehicle individual is obtained. Finally, a genetic algorithm is substituted and combined with a simulated annealing algorithm to balance the profit of the parking lot operator and the satisfaction of the owner, and the optimal charging allocation plan for the vehicle is obtained, which can simultaneously improve the operator's profit and the owner's satisfaction, and effectively alleviate the load fluctuation of the distribution network, so as to achieve a win-win situation for the parking lot operator and the electric vehicle owner. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The flowchart of an electric vehicle charging scheduling method for multi-objective optimization is shown. DETAILED DESCRIPTION

[0014] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.

[0015] Please refer to Figure 1 , a multi-objective optimization-oriented electric vehicle charging scheduling method, comprising the steps of:

[0016] S1. Establish a vehicle access mechanism to restrict vehicles entering the parking lot;

[0017] S2. Linearly weight the two optimization objectives of maximizing operator profit and maximizing owner satisfaction to obtain the final objective function of each individual vehicle entering the parking lot for charging;

[0018] S3. Randomly assigning charging time for the individual vehicles to obtain a charging allocation plan for the individual vehicles;

[0019] S4, using a genetic algorithm to update the charging allocation plan, and combining a simulated annealing algorithm to determine whether to accept the updated charging allocation plan;

[0020] S5. Output the final charging allocation plan for the individual vehicle.

[0021] From the above description, it can be seen that the beneficial effects of the present invention are: first, in order to ensure that the vehicle can complete the charging task before leaving and meet the charging needs of more vehicles as much as possible, a vehicle access mechanism is proposed; secondly, linear weighting is used to comprehensively consider the two optimization goals of maximizing the operator's profit and maximizing the owner's satisfaction, and the final objective function of the vehicle individual is obtained. Finally, the genetic algorithm is substituted and combined with the simulated annealing algorithm to balance the parking lot operator's profit and the owner's satisfaction, and the optimal charging allocation plan for the vehicle is obtained, which can simultaneously improve the operator's profit and the owner's satisfaction, and effectively alleviate the load fluctuation of the distribution network, achieving a win-win situation for parking lot operators and electric vehicle owners.

[0022] Furthermore, the step S1 further includes the following steps:

[0023] S0. There are c parking spaces in the preset parking lot, of which m are equipped with charging piles. The scheduling time interval is set to ΔT, and the total scheduling time T is divided into several time slices. The charging control center of the parking lot adjusts the charging plan after every ΔT.

[0024] From the above description, it can be seen that the total scheduling time is divided into multiple time slices, and the charging plan is adjusted in time within each time slice to ensure that the charging scheduling of electric vehicles always maintains the best efficiency.

[0025] Furthermore, the step S1 specifically includes the following steps:

[0026] S11. Before the individual vehicle enters the parking lot, it is determined according to the vehicle access mechanism whether its charging demand can be completed before leaving the parking lot. If so, the individual vehicle is parked in a parking space with a charging pile, otherwise the individual vehicle is parked in a parking space without a charging pile;

[0027] S12: After the vehicle is connected to the charging pile, the vehicle information of the vehicle is obtained, and the vehicle information includes the arrival time t of the vehicle at the parking lot. i a , expected departure time t i l and daily mileage d i , i represents the i-th vehicle that enters the parking lot and parks at a parking space with a charging pile;

[0028] S13, the arrival time t i a and the estimated departure time t i l Convert to the corresponding arrival time slice n i a and the expected departure time slice n i l, the daily mileage is converted into the vehicle's power consumption, that is, the desired charge before leaving r i , then n i a 、n i l and r i Satisfies the following formula:

[0029]

[0030]

[0031] r i =λd i (3);

[0032] λ=e h / d h (4);

[0033] Among them, e h is the power consumption of the vehicle in a time slice ΔT, d h is the mileage of the individual vehicle in a time slice ΔT;

[0034] S14, let X i represents the charging status of the i-th vehicle within the total scheduling time T, then X i =[X i 1 X i 2 X i 3 ...X i T ] T , which can be further expressed as:

[0035]

[0036] Among them, 1 indicates the charging state, 0 indicates the uncharging state, and X i n The value of X is 0 before the i-th vehicle arrives at the parking lot or leaves the parking lot. During the parking charging period, i n The value of is represented as a binary variable, and its value of 1 or 0 is determined by the vehicle admission mechanism and scheduling algorithm.

[0037] As can be seen from the above description, the charging status of the electric vehicle in each parking time slice is represented by the binary simplification of 0 and 1, which is convenient for scheduling and unified management of the charging of each vehicle.

[0038] Furthermore, the vehicle access mechanism satisfies the following restrictions:

[0039]

[0040] Among them, n i ri The number of time slices required to complete charging.

[0041] From the above description, it can be seen that in order to ensure the satisfaction of car owners, all vehicles entering the parking lot for charging must complete charging before leaving the parking lot. Therefore, a vehicle access mechanism is adopted, and vehicles allowed to enter for charging are regarded as dispatchable vehicles. In this way, the charging period for each car can be allocated based on the data of all vehicles in the parking lot. At the same time, under the premise of meeting the needs of car owners, part of the charging can be transferred to the low-peak period of electricity consumption, so as to achieve orderly charging of vehicles.

[0042] Furthermore, the step S2 specifically includes the following steps:

[0043] S21. Solve the objective function of maximizing the operator's profit and obtain the first objective function, where the operator's profit is the difference between the charging fee paid by the car owner to the operator and the electricity purchase fee paid by the operator to the power grid. Assume that there is currently v s The individual vehicles enter the parking lot for charging, then the first objective function is as shown in Formula 7:

[0044]

[0045]

[0046]

[0047]

[0048] Among them, P c and P g They represent the charging fees paid by the car owner to the operator and the electricity purchase fees paid by the operator to the grid in a single time slice during the peak and valley periods, j=1, j=2 and j=3 represent the peak, flat and valley periods respectively, N i r represents the number of time slices that the i-th vehicle charges during the peak, flat and valley periods;

[0049] S22, solving the objective function of maximizing the owner's satisfaction to obtain a second objective function, wherein the owner's satisfaction S includes the satisfaction with the charging time and the satisfaction with the charging cost, and the second objective function is as shown in Formula 11:

[0050]

[0051] Where e is the charge amount in a time slice ΔT, is the satisfaction penalty function, σ is the preset penalty coefficient, then:

[0052]

[0053] S23, linearly weighting the first objective function and the second objective function to obtain the final objective function, as shown in Formula 13:

[0054]

[0055] Among them, P now and S now Represent the current profit and satisfaction respectively, P max and S max They represent the maximum possible values ​​of profit and satisfaction respectively, which are specifically expressed as formulas 14 and 15:

[0056]

[0057] S max =2·v (15).

[0058] From the above description, it can be seen that since two optimization objectives need to be met - maximizing operator profits and maximizing owner satisfaction, these two objectives need to be linearly weighted to transform the multi-objective problem into a single-objective problem.

[0059] Furthermore, the step S2 also includes a constraint condition, and the constraint condition is specifically:

[0060] The charge amount e of the vehicle in a time slice ΔT satisfies:

[0061]

[0062] Where u is the rated charging power and η is the charging efficiency;

[0063] The vehicle can only be charged during the time period when it is connected to the charging pile, and the time to start charging cannot be earlier than the time when the vehicle arrives, and the time to end charging cannot be later than the time when the vehicle is expected to leave, resulting in Formula 17:

[0064]

[0065] The number of the individual vehicles charged simultaneously in the time slice t cannot exceed the number of charging piles M, and formula 18 is obtained:

[0066]

[0067] The final charging amount of the individual vehicle parked in the parking space with the charging pile is allowed to meet its charging demand, and formula 19 is obtained:

[0068]

[0069] From the above description, it can be seen that the constraints are restricted to ensure the effectiveness of the final objective function after linear weighting of the two objectives of maximizing operator profit and maximizing car owner satisfaction.

[0070] Furthermore, the step S3 is specifically as follows:

[0071] The available charging time slice for each individual vehicle is obtained by formula 17. i a and the expected departure time slice n i l Randomly select n i r time slices as the charging time of the individual vehicle, and each individual vehicle represents a charging time allocation scheme for all vehicles. Then, a number of individual vehicles are randomly generated according to all the individual vehicles as the initial population. Then, the position of the individual vehicle α at the βth iteration is as shown in Formula 20, that is, the charging allocation scheme of the individual vehicle is obtained:

[0072]

[0073] Among them, α has v s Percentile, represents the charging time of the kth vehicle at the βth iteration, and the charging time satisfies formulas 21 and 22:

[0074]

[0075]

[0076] From the above description, it can be seen that by obtaining the charging time slices of a single vehicle during parking and charging, it is possible to freely allocate which time slices to charge within these time slices so as to meet the required power before the vehicle leaves the parking lot, that is, to obtain the vehicle's charging allocation plan. At the same time, the charging time is limited to satisfy formulas 21 and 22 to ensure the correctness of the charging time allocation, that is, no repeated time slices can occur.

[0077] Furthermore, the step S4 specifically includes the following steps:

[0078] S41, initializing the vehicle individual, and obtaining a fitness function value F1 from the final objective function of the vehicle individual, where F1 satisfies:

[0079]

[0080] And output the current charging allocation plan Z1;

[0081] S42, introducing the fitness function value into a genetic algorithm, randomly selecting the vehicle individuals in the initial population, and combining them through selection operation, crossover operation or mutation operation to generate a new next generation, obtaining a new fitness function value F2, and obtaining a new optimal solution Z2;

[0082] S43. If F2≥F1, the optimal function is Z2; otherwise, the optimal function is Z1.

[0083] From the above description, it can be seen that the present invention aims to balance the profits of parking lot operators and the satisfaction of car owners. Therefore, the larger the final objective function value, the better the result. Therefore, the final objective function of each allocation scheme is regarded as the fitness function value to obtain a charging allocation scheme that satisfies the fitness function value. At the same time, for the accuracy of the scheme, the randomly initialized vehicle individuals are calculated to obtain the fitness function value, and a genetic algorithm is used to select some vehicle individuals based on the fitness function value. Generation after generation is generated through selection, crossover or mutation operations, and gradually evolves towards the optimal solution. At the same time, in order to make up for the deficiency of "premature" convergence of the genetic algorithm, the evolution mechanism of the simulated annealing algorithm is introduced to realize the combination of the good local search capability of the simulated annealing algorithm and the global search capability of the genetic algorithm.

[0084] Furthermore, the step S43 further includes:

[0085] If F2<F1, the simulated annealing algorithm is introduced to calculate the Metropolis probability P M , whose size is affected by the current temperature and the difference between the fitness function values ​​before and after the mutation, then P M for:

[0086]

[0087] Among them, Tem is the temperature control parameter, and the cooling process can be expressed as:

[0088] Tem=tem0 / (1+q) (25);

[0089] Among them, tem0 is the initial temperature, q is the number of iterations;

[0090] Randomly generate a random number in the range of (0,1). When it is greater than P M When the optimal function is Z1, it is less than or equal to P M When , the optimal function is Z2.

[0091] From the above description, we can see that by introducing the simulated annealing algorithm into the genetic algorithm for operation, that is, choosing whether to accept the new solution after inheritance, we can not only increase the diversity of genes, but also avoid the premature maturity of the optimal solution to a certain extent.

[0092] Furthermore, the selection operation includes: adopting a tournament selection strategy to select a number of relatively good vehicle individuals from the initial population of the previous generation and inherit them to the next generation;

[0093] The crossover operation includes: setting the crossover probability to 0.9, randomly generating a random number in the range of [0,1), and if it is less than the preset crossover probability, performing a two-point crossover operation;

[0094] The mutation operation includes: setting the mutation probability to 0.1, randomly generating a random number in the range of [0,1), and if it is less than the preset mutation probability, performing a mutation operation. If there are unsolvable time slices, selecting the quantile with the most unsolvable time slices in the vehicle individual and changing the value of the quantile, while ensuring that the changed value satisfies the constraint condition.

[0095] From the above description, we can see that the three basic operations of genetic algorithms are selection, crossover and mutation. The selection operation has the advantages of low complexity and easy parallel processing. The crossover operation can quickly obtain the optimal individual from two randomly selected individuals, and the mutation operation can enrich the diversity of individuals and increase the accuracy of the optimal solution. In the application, the required genetic operation can be selected according to actual needs.

[0096] The present invention provides an electric vehicle charging scheduling method for target optimization, which is applicable to a parking lot scenario for allocating charging time for electric vehicles taking into account the profits of parking lot operators and the satisfaction of car owners.

[0097] Please refer to Figure 1 , Embodiment 1 of the present invention is:

[0098] The charging scheduling problem in a parking lot environment can be described as follows: when a vehicle enters and connects to a charging pile, the charging control center will allocate the charging time for each vehicle based on the acquired vehicle information (vehicle arrival time, expected departure time, and daily mileage), charging pile information (number of remaining available charging piles, rated charging power), access mechanism, and distribution network-side electricity prices, with the goal of balancing the profits of parking lot operators and the satisfaction of car owners.

[0099] The present embodiment provides a multi-objective optimization-oriented electric vehicle charging scheduling method, such as Figure 1 As shown, the steps include:

[0100] S1. Establish a vehicle access mechanism to restrict vehicles entering the parking lot;

[0101] S2. Linearly weight the two optimization objectives of maximizing operator profit and maximizing owner satisfaction to obtain the final objective function of each individual vehicle entering the parking lot for charging;

[0102] S3. Randomly assign charging time to individual vehicles to obtain a charging allocation plan for the individual vehicles;

[0103] S4, using a genetic algorithm to update the charging allocation plan, and combining it with a simulated annealing algorithm to determine whether to accept the updated charging allocation plan;

[0104] S5. Output the final charging allocation plan for each vehicle.

[0105] That is, in this embodiment, firstly, in order to ensure that the vehicle can complete the charging task before leaving and meet the charging needs of more vehicles as much as possible, a vehicle access mechanism is proposed. Secondly, the two optimization goals of maximizing the operator's profit and maximizing the owner's satisfaction are comprehensively considered by linear weighting to obtain the final objective function of the individual vehicle. Finally, the genetic algorithm is substituted and combined with the simulated annealing algorithm to balance the profit of the parking lot operator and the satisfaction of the owner, and the optimal charging allocation plan for the vehicle is obtained, which can simultaneously improve the operator's profit and the owner's satisfaction, and effectively alleviate the load fluctuation of the distribution network, so as to achieve a win-win situation for the parking lot operator and the electric vehicle owner.

[0106] Embodiment 2 of the present invention is:

[0107] A multi-objective optimization-oriented electric vehicle charging scheduling method, based on the above-mentioned embodiment 1, in this embodiment, before step S1, further includes the following steps:

[0108] S0. There are c parking spaces in the preset parking lot, of which m are equipped with charging piles. The scheduling time interval is set to ΔT, and the total scheduling time T is divided into several time slices. The charging control center of the parking lot adjusts the charging plan after every ΔT.

[0109] That is, the total scheduling time is divided into multiple time slices, and the charging plan is adjusted in time within each time slice to ensure that the electric vehicle charging scheduling always maintains the best efficiency.

[0110] Wherein, step S1 specifically includes the following steps:

[0111] S11. Before an individual vehicle enters the parking lot, it is determined based on the vehicle access mechanism whether its charging needs can be completed before leaving the parking lot. If so, the individual vehicle is parked in a parking space with a charging pile. Otherwise, the individual vehicle is parked in a parking space without a charging pile.

[0112] The charging control center will decide whether to accept the owner's charging request and manage the power supply based on the scheduling strategy, thereby starting or pausing the charging equipment in the field. The entire charging process is operated by the circuit switching control unit, so there is no need to move the vehicle during the charging process. In order to ensure the satisfaction of the owner, that is, all vehicles entering can complete charging before leaving, this embodiment adopts a vehicle access mechanism, in which the vehicle access mechanism needs to meet the following restrictions:

[0113]

[0114] Among them, n i ri is the number of time slices required to complete charging.

[0115] That is, in order to ensure the satisfaction of car owners, all vehicles entering the parking lot for charging must complete charging before leaving the parking lot. Therefore, a vehicle access mechanism is adopted, and vehicles allowed to enter for charging are regarded as dispatchable vehicles. In this way, the charging time period for each car can be allocated based on the data of all vehicles in the parking lot. At the same time, under the premise of meeting the needs of car owners, part of the charging time can be transferred to the low-peak period of electricity consumption, so as to achieve orderly charging of vehicles.

[0116] When electric cars are allowed to enter the parking lot, the parking lot can manage the electric cars in a certain area in a unified manner, that is, the vehicle information of the car can be obtained after the vehicle is plugged into the charging pile.

[0117] S12: After the vehicle is connected to the charging pile, the vehicle information of the vehicle is obtained, including the arrival time t of the vehicle at the parking lot. i a , expected departure time t i l and daily mileage d i , i represents the i-th vehicle that enters the parking lot and parks in a parking space with a charging pile.

[0118] S13, the arrival time t i a and the estimated departure time t i l Convert to the corresponding arrival time slice n i a and the expected departure time slice n i l , the daily mileage is converted into the vehicle's power consumption, that is, the desired charge before leaving r i , then n i a 、n i l and r i Satisfies the following formula:

[0119]

[0120]

[0121] r i =λd i (3);

[0122] λ=e h / d h (4);

[0123] Among them, e h is the power consumption of an individual vehicle in a time slice ΔT, d h is the mileage of an individual vehicle within a time slice ΔT.

[0124] S14, let X i represents the charging status of the i-th vehicle within the total scheduling time T, then X i =[X i 1 X i 2 X i 3 ...X i T ] T , which can be further expressed as:

[0125]

[0126] Among them, 1 indicates the charging state, 0 indicates the uncharging state, and X i n The value of is 0 before the i-th vehicle arrives at the parking lot or leaves the parking lot. During the parking charging period, i n The value of is represented as a binary variable, and its value of 1 or 0 is determined by the vehicle access mechanism and scheduling algorithm. That is, the charging status of the electric vehicle in each parking time slice is represented by a binary simplification of 0 and 1, which is convenient for scheduling and unified management of the charging of each vehicle.

[0127] Then, considering the two optimization objectives, namely, operator profit and owner satisfaction, step S2 specifically includes the following steps:

[0128] S21. Solve the objective function of maximizing the operator's profit and obtain the first objective function, where the operator's profit is the difference between the charging fee paid by the car owner to the operator and the electricity purchase fee paid by the operator to the power grid. Assume that there is currently v s If a vehicle enters the parking lot to charge, the first objective function is as follows:

[0129]

[0130]

[0131]

[0132]

[0133] Among them, P c and P g They represent the charging fees paid by the car owner to the operator and the electricity purchase fees paid by the operator to the grid in a single time slice during the peak and valley periods, j=1, j=2 and j=3 represent the peak, flat and valley periods respectively, N i r It represents the number of time slices that the i-th vehicle charges during the peak, flat and valley periods.

[0134] For example, in this embodiment, P g =(P c 1 , P c 2 , P c 3 )=(3,2,1), which means that when the electricity price is in the peak period, the operator pays 3 yuan per unit time slice to the grid, when the electricity price is in the normal period, the operator pays 2 yuan per unit time slice to the grid, and when the electricity price is in the valley period, the operator pays 1 yuan per unit time slice to the grid; similarly, for example, when P c =(P c 1 , P c 2 , P c 3 )=(6,5,4), which means that when the electricity price is in the peak period, the charging fee paid by the car owner to the operator is 6 yuan per unit time slice, when the electricity price is in the normal period, the charging fee paid by the car owner to the operator is 5 yuan per unit time slice, and when the electricity price is in the valley period, the charging fee paid by the car owner to the operator is 4 yuan per unit time slice; for example, N i r =(7, 8, 9) means that the i-th vehicle charged for 7 time slices during the peak electricity price period, 8 time slices during the normal electricity price period, and 9 time slices during the valley electricity price period.

[0135] S22, solving the objective function of maximizing the owner's satisfaction, and obtaining the second objective function, wherein the owner's satisfaction S includes the satisfaction with the charging time and the satisfaction with the charging cost, and the second objective function is as shown in Formula 11:

[0136]

[0137] Where e is the charge amount in a time slice ΔT, is the satisfaction penalty function, σ is the preset penalty coefficient, then:

[0138]

[0139] In this embodiment, n in the above formula (11) i r2 and n i r3 They represent the number of charging time slices of the vehicle during the flat electricity price period and the valley electricity price period. Since the charging fee paid to the operator is lower when the vehicle is charged during the non-peak electricity price period, (n i r2 +n i r3 ) / n i r (The ratio of the number of time slots for charging during off-peak periods to the total number of time slots required to complete charging) indicates the vehicle's satisfaction with charging costs without adding the number of charging time slots during peak periods. i r1 .

[0140] S23, linearly weighting the first objective function and the second objective function to obtain a final objective function, as shown in Formula 13:

[0141]

[0142] Among them, P now and S now Represent the current profit and satisfaction respectively, P max and S max They represent the maximum possible values ​​of profit and satisfaction respectively, which are specifically expressed as formulas 14 and 15:

[0143]

[0144] S max =2·v (15).

[0145] Since two optimization objectives need to be met simultaneously, the two objectives need to be linearly weighted to transform the multi-objective problem into a single-objective problem.

[0146] At the same time, step S2 also includes constraints, specifically:

[0147] The charge amount e of a vehicle in a time slice ΔT satisfies:

[0148]

[0149] Where u is the rated charging power and η is the charging efficiency;

[0150] The vehicle can only be charged during the time slice when it is connected to the charging pile, and the time to start charging cannot be earlier than the time when the vehicle arrives, and the time to end charging cannot be later than the time when the vehicle is expected to leave, resulting in Formula 17:

[0151]

[0152] The number of vehicles that are charged simultaneously in time slice t cannot exceed the number of charging piles M, and we get formula 18:

[0153]

[0154] The final charging amount of the individual vehicle parked in the parking space with the charging pile is allowed to meet its charging demand, and formula 19 is obtained:

[0155]

[0156] That is, the constraints are limited to ensure the effectiveness of the final objective function after linear weighting of the two objectives of maximizing operator profit and maximizing car owner satisfaction.

[0157] Wherein, step S3 is specifically as follows:

[0158] The available charging time slice for each vehicle can be obtained from the above formula 17. Therefore, when the vehicle reaches time slice n i a and the expected departure time slice n i l Randomly select n i r Assuming that each vehicle individual represents a charging time allocation scheme for all vehicles, a number of vehicle individuals are randomly generated based on all vehicle individuals as the initial population. The position of vehicle individual α at the βth iteration is as shown in Formula 20, and the charging allocation scheme for the vehicle individual is obtained:

[0159]

[0160] Among them, α has v s Percentile, represents the charging time of the kth vehicle at the βth iteration, which satisfies formulas 21 and 22:

[0161]

[0162]

[0163] Since the charging time slices of individual vehicles during parking and charging can be obtained, the charging time slices can be freely allocated within these time slices to meet the required power before the vehicle leaves the parking lot, thereby obtaining the vehicle's charging distribution plan. At the same time, the charging time is limited to satisfy formulas 21 and 22 to ensure the correctness of the charging time distribution, that is, no repeated time slices can occur.

[0164] Wherein, step S4 specifically includes the following steps:

[0165] S41, initialize the vehicle individual, and obtain the fitness function value F1 from the final objective function of the vehicle individual, then F1 satisfies:

[0166]

[0167] And output the current charging allocation plan Z1;

[0168] S42, introducing the fitness function value into the genetic algorithm, randomly selecting vehicle individuals in the initial population, and combining them through selection operation, crossover operation or mutation operation to generate a new next generation, obtaining a new fitness function value F2, and obtaining a new optimal solution Z2;

[0169] S43. If F2≥F1, the optimal function is Z2, otherwise the optimal function is Z1.

[0170] Since the present invention aims to balance the profits of parking lot operators and the satisfaction of car owners, the larger the final objective function value, the better the result. Therefore, the final objective function of each allocation scheme is regarded as the fitness function value to obtain a charging allocation scheme that satisfies the fitness function value. At the same time, for the accuracy of the scheme, the randomly initialized vehicle individuals are calculated to obtain the fitness function value, and a genetic algorithm is used to select some vehicle individuals based on the fitness function value to generate generations through selection, crossover or mutation operations, and gradually evolve towards the optimal solution. At the same time, in order to make up for the deficiency of "premature" convergence of the genetic algorithm, the evolution mechanism of the simulated annealing algorithm is introduced to realize the combination of the good local search ability of the simulated annealing algorithm and the global search ability of the genetic algorithm.

[0171] It is particularly important to note that since the problem encoding may result in unsolvable solutions, that is, the number of electric vehicles charged in a certain time slice exceeds the number of charging piles, when two vehicle individuals are unsolvable, the vehicle individual with a larger fitness function value is more likely to obtain a better solution through subsequent selection, crossover or mutation operations, so the vehicle individual with a larger fitness function value is also selected, and the definition of the fitness function is the same as the above formula.

[0172] Wherein, step S43 further includes:

[0173] If F2<F1, the simulated annealing algorithm is introduced to calculate the Metropolis probability P M , whose size is affected by the current temperature and the difference in fitness function value before and after the mutation, then P M for:

[0174]

[0175] Among them, Tem is the temperature control parameter, and the cooling process can be expressed as:

[0176] Tem=tem0 / (1+q) (25);

[0177] Among them, tem0 is the initial temperature, q is the number of iterations;

[0178] Randomly generate a random number in the range of (0,1). When it is greater than P M When Z1 is less than or equal to P M When , the optimal function is Z2.

[0179] That is, if the fitness function value of the vehicle individual after inheritance is greater than the fitness function value of the vehicle individual before mutation, then inheritance is accepted; otherwise, a random number with a value range of (0,1) is randomly generated, and when it is greater than P M When it is less than P, it does not accept inheritance. M When, accept inheritance.

[0180] That is, in this embodiment, by introducing the simulated annealing algorithm into the genetic algorithm for operation, that is, choosing whether to accept the new solution after inheritance, not only can the diversity of genes be increased, but also the premature maturity of the optimal solution can be avoided to a certain extent.

[0181] Among them, in this embodiment, the selection operation includes: adopting the tournament selection strategy, selecting several relatively good vehicle individuals from the previous generation of the initial population to be inherited to the next generation, which has the advantages of low complexity and easy parallel processing; the crossover operation includes: setting the crossover probability to 0.9, randomly generating a random number in the range of [0,1), if it is less than the pre-set crossover probability, then performing a two-point crossover operation, and the optimal individual can be quickly obtained from two randomly selected individuals; the mutation operation includes: setting the mutation probability to 0.1, randomly generating a random number in the range of [0,1), if it is less than the pre-set mutation probability, then performing a mutation operation, if there is an unsolvable time slice, then selecting the quantile with the most unsolvable time slices in the vehicle individual and changing the value of the quantile, while ensuring that the changed value meets the constraint conditions, which can enrich the diversity of individuals and increase the accuracy of the optimal solution. In the application, the required genetic operation can be selected according to actual needs.

[0182] Embodiment 3 of the present invention is:

[0183] A multi-objective optimization-oriented electric vehicle charging scheduling method is disclosed. Based on the above second embodiment, in this embodiment, the steps of the genetic simulated annealing algorithm are exemplified.

[0184] First, each individual vehicle corresponds to a charging time allocation plan for the vehicle. The following Algorithm 1 is a pseudo code for the specific implementation of the Genetic Simulated Annealing Algorithm (GASA).

[0185] Algorithm 1:

[0186] 1. Input: total number of vehicles, number of charging stations, vehicle arrival time, estimated departure time, daily mileage

[0187] 2. Initialization: Set the relevant parameters of the algorithm: population size N, maximum number of iterations of the genetic algorithm (GA) β max , crossover probability P c , mutation probability P M , SA initial temperature Tem0, let β=1

[0188] 3. Determine the number of vehicles v that can enter the parking lot and charge according to formula (6): s

[0189] 4. Randomly assign the charging time of vehicles according to formulas (1) and (2), generate N individuals, and initialize the population

[0190] 5.whileβ≤β max do

[0191] 6.for j=1 to j=v s

[0192] 7. Calculate the number of vehicles charging simultaneously in each time slice, and calculate the operator profit, owner satisfaction and fitness function value of the charging time allocation scheme according to formulas (7), (11) and (23).

[0193] 8.if x j n >m

[0194] 9.Return the time slice of the current vehicle that exceeds the charging pile usage limit

[0195] 10.end if

[0196] 11. end for

[0197] 12. Perform selection, crossover and mutation operations on the initialized population, setting β = β + 1

[0198] 13. Output the current optimal solution Z1 and its fitness function value F1, and set tem = Tem0

[0199] 14. After mutation, we get the new solution Z2 and its fitness function value F2

[0200] 15. If F2>F1, then Z2 is the current optimal solution, otherwise go to step 16;

[0201] 16. Calculate P according to formula (25) M

[0202] 17.if P M >random(0,1)

[0203] 18.Z2 is the current optimal solution;

[0204] 19.else

[0205] 20.Z1 is the current optimal solution;

[0206] 21.end if

[0207] 22.tem=tem / (1+β)

[0208] 23.end while

[0209] 24. Output profit, satisfaction, fitness function value and corresponding charging time allocation plan

[0210] The specific explanation of Algorithm 1 is as follows:

[0211] 1) In step 1 to step 4, the relevant parameters of the algorithm are initialized, and the initial population is randomly generated according to the vehicle information.

[0212] 2) In step 5 to step 11, the relevant information of the current vehicle individual is calculated.

[0213] 3) In step 12 to step 14, the selection, crossover, and mutation operations in the genetic algorithm and the Metropolis criterion in the simulated annealing algorithm are introduced to update the current vehicle individual and recalculate its related information.

[0214] 4) In step 15 to step 23, the evolution mechanism of the simulated annealing algorithm is combined to decide whether to accept the new solution. If the fitness function value of the new solution is greater than the previous fitness function value, the new solution is accepted, otherwise the new solution is accepted with a certain probability.

[0215] 5) In step 24, the relevant information of the individual vehicle and its corresponding charging time allocation plan are output.

[0216] In summary, the present invention provides an electric vehicle charging scheduling method for multi-objective optimization, which considers the influence of different indicators on the charging scheduling strategy in the parking lot scenario, and proposes a charging time allocation strategy based on a simulated genetic algorithm for multi-objective optimization, with the aim of balancing the profits of parking lot operators and the satisfaction of car owners under the premise of ensuring the safety of the power grid. The method comprehensively considers the influence of operator profits and car owner satisfaction on charging scheduling, adopts a linear weighting method, and introduces the evolutionary mechanism of simulated annealing into the genetic algorithm, balancing the exploration and development of the algorithm, that is, judging whether to accept the new solution generated after the mutation operation of the genetic algorithm, thereby expanding the local search capability and effectively balancing the profits of parking lot operators and the satisfaction of car owners.

[0217] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A multi-objective optimization-oriented electric vehicle charging scheduling method, characterized in that: Includes steps: S1. Establish a vehicle access mechanism to restrict vehicles entering the parking lot; S2. Linearly weight the two optimization objectives of maximizing operator profit and maximizing owner satisfaction to obtain the final objective function of each individual vehicle entering the parking lot for charging; S3. Randomly assigning charging time for the individual vehicles to obtain a charging allocation plan for the individual vehicles; S4, using a genetic algorithm to update the charging allocation plan, and combining a simulated annealing algorithm to determine whether to accept the updated charging allocation plan; S5. Outputting the final charging allocation plan for the individual vehicle; The step S1 also includes the following steps: S0, there are c parking spaces in the preset parking lot, of which there are m parking spaces with charging piles. Set the scheduling time interval to ΔT, and divide the total scheduling time T into several time slices. Then the charging control center of the parking lot adjusts the charging plan after every ΔT; The step S1 specifically includes the following steps: S11. Before the individual vehicle enters the parking lot, it is determined according to the vehicle access mechanism whether its charging demand can be completed before leaving the parking lot. If so, the individual vehicle is parked in a parking space with a charging pile, otherwise the individual vehicle is parked in a parking space without a charging pile; S12: After the vehicle is connected to the charging pile, the vehicle information of the vehicle is obtained, and the vehicle information includes the arrival time t of the vehicle at the parking lot. i a , expected departure time t i l and daily mileage d i , i represents the i-th vehicle that enters the parking lot and parks at a parking space with a charging pile; S13, the arrival time t i a and the estimated departure time t i l Convert to the corresponding arrival time slice n i a and the expected departure time slice n i l , the daily mileage is converted into the vehicle's power consumption, that is, the desired charge amount ri before leaving, then n i a 、n i l and ri satisfy the following formula: (1); (2); (3); (4); Among them, e h is the power consumption of the vehicle in a time slice ΔT, d h is the mileage of the individual vehicle in a time slice ΔT; S14, let X i represents the charging status of the i-th vehicle within the total scheduling time T, then X i =[X i 1 , X i 2 , X i 3 , ..., X i T ] T , further expressed as: (5); Among them, 1 indicates the charging state, 0 indicates the uncharging state, and X i n The value of is 0 before the i-th vehicle arrives at the parking lot or leaves the parking lot. During the parking charging period, i n The value of is represented as a binary variable, and the value of 1 or 0 is determined by the vehicle admission mechanism and the scheduling algorithm; The vehicle admission mechanism meets the following constraints: (6); Among them, n i ri The number of time slices required to complete charging; The step S2 specifically includes the following steps: S21. Solve the objective function of maximizing the operator's profit and obtain the first objective function, where the operator's profit is the difference between the charging fee paid by the car owner to the operator and the electricity purchase fee paid by the operator to the power grid. Assume that there is currently v s The individual vehicles enter the parking lot for charging, then the first objective function is as shown in Formula 7: (7); (8); (9); (10); Among them, P c and P g They represent the charging fees paid by the car owner to the operator and the electricity purchase fees paid by the operator to the grid in a single time slice during the peak and valley periods, j=1, j=2 and j=3 represent the peak, flat and valley periods respectively, and N i r represents the number of time slices that the i-th vehicle charges during the peak, flat and valley periods; S22, solving the objective function of maximizing the owner's satisfaction to obtain a second objective function, wherein the owner's satisfaction S includes the satisfaction with the charging time and the satisfaction with the charging cost, and the second objective function is as shown in Formula 11: (11); in, is the satisfaction penalty function, σ is the preset penalty coefficient, then: (12); S23, linearly weighting the first objective function and the second objective function to obtain the final objective function, as shown in Formula 13: (13); Among them, P now and S now Represent the current profit and satisfaction respectively, P max and S max They represent the maximum possible values ​​of profit and satisfaction respectively, which are specifically expressed as formulas 14 and 15: (14); (15); The step S2 also includes constraints, which are specifically: The charge amount e of the vehicle in a time slice ΔT satisfies: (16); Where u is the rated charging power and η is the charging efficiency; The vehicle can only be charged during the time period when it is connected to the charging pile, and the time to start charging cannot be earlier than the time when the vehicle arrives, and the time to end charging cannot be later than the time when the vehicle is expected to leave, resulting in Formula 17: (17); The number of the individual vehicles charged simultaneously in the time slice t cannot exceed the number of charging piles M, and formula 18 is obtained: (18); The final charging amount of the individual vehicle parked in the parking space with the charging pile is allowed to meet its charging demand, and formula 19 is obtained: (19)。 2. The multi-objective optimization-oriented electric vehicle charging scheduling method according to claim 1 is characterized in that: The step S3 is specifically as follows: The available charging time slice for each vehicle is obtained by formula 17, where the vehicle reaches time slice n i a and the expected departure time slice n i l Randomly select n i r time slices as the charging time of the individual vehicle, and each individual vehicle represents a distribution scheme for the charging time of all vehicles. Then, a number of individual vehicles are randomly generated according to all the individual vehicles as the initial population. Then, the position of the individual vehicle α at the βth iteration is as shown in Formula 20, that is, the charging distribution scheme of the individual vehicle is obtained: (20); Among them, α has v s Percentile, represents the charging time of the kth vehicle at the βth iteration, and the charging time satisfies formulas 21 and 22: (21); (22)。 3. The multi-objective optimization-oriented electric vehicle charging scheduling method according to claim 2 is characterized in that: The step S4 specifically comprises the following steps: S41, initializing the vehicle individual, and obtaining a fitness function value F1 from the final objective function of the vehicle individual, where F1 satisfies: (23); And output the current optimal solution Z1; S42, introducing the fitness function value into the genetic algorithm, randomly selecting the vehicle individuals in the initial population, and combining them through selection operation, crossover operation or mutation operation to generate a new next generation, obtaining a new fitness function value F2, and obtaining a new optimal solution Z2; S43. If F2 ≥ F1, the optimal solution is Z2, otherwise the optimal solution is Z1.

4. The multi-objective optimization-oriented electric vehicle charging scheduling method according to claim 3 is characterized in that: The step S43 further includes: If F2<F1, the simulated annealing algorithm is introduced to calculate the Metropolis probability P M , whose size is affected by the current temperature and the difference between the fitness function values ​​before and after the mutation, then P M for: (24); Among them, Tem is the temperature control parameter, and the cooling process is expressed as: (25); Among them, tem0 is the initial temperature, q is the number of iterations; Randomly generate a random number in the range of (0,1). When it is greater than P M When Z1 is less than or equal to P M When , the optimal solution is Z2.

5. The multi-objective optimization-oriented electric vehicle charging scheduling method according to claim 3 is characterized in that: The selection operation includes: adopting a tournament selection strategy to select a number of relatively good vehicle individuals from the initial population of the previous generation and inherit them to the next generation; The crossover operation includes: setting the crossover probability to 0.9, randomly generating a random number in the range of [0,1), and if it is less than the preset crossover probability, performing a two-point crossover operation; The mutation operation includes: setting the mutation probability to 0.1, randomly generating a random number in the range of [0,1), and if it is less than the preset mutation probability, performing a mutation operation. If there are unsolvable time slices, selecting the quantile with the most unsolvable time slices in the vehicle individual and changing the value of the quantile, while ensuring that the changed value satisfies the constraint condition.