Electric vehicle charging and discharging scheduling optimization method and device, medium and program

Abstract

The invention provides an electric vehicle charging and discharging scheduling optimization method and device, a medium and a program, and the method takes the minimum operation cost of a power distribution network and the charging cost of an EV user as targets, and obtains a scheduling method of an electric vehicle through optimization. The operation cost of the power distribution network includes subtracting electricity selling income from electricity purchasing cost; the charging cost of the EV users is obtained by subtracting the dispatching income from the battery loss cost of each EV. According to the electric vehicle charging and discharging scheduling optimization method, the operation cost of the power distribution network and the charging cost of the EV user are comprehensively considered, and the electric vehicle charging and discharging scheduling scheme is optimized, so that dynamic balance of benefits on both sides of supply and demand is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of interaction between electric vehicles and power grids, and specifically relates to a method, device, medium, and program for optimizing charging and discharging scheduling of electric vehicles. Background Art

[0002] The large-scale integration of electric vehicles (EVs) has further exacerbated load fluctuations in distribution networks. Demand response, as a key means of balancing supply and demand, can effectively guide EV users to optimize their charging and discharging behaviors. During periods of high grid load, distribution networks need to implement strategies to alleviate the pressure of peak loads on the grid. A common approach is to increase charging and discharging prices to encourage EV users to reduce their charging needs, or even to incentivize them to discharge battery power back into the grid during certain periods. On the other hand, during off-peak periods, the grid can lower charging prices to encourage users to charge, thereby achieving load balancing, avoiding overloading, and improving grid efficiency.

[0003] However, scheduling the charging and discharging of connected electric vehicles inevitably results in additional charging and discharging time. This not only increases charging time for drivers, but also increases battery wear. The resulting battery wear and charging time loss impacts the willingness of electric vehicle users to respond. Existing research has failed to fully consider user response intentions and their variability, making efficient scheduling difficult to achieve objectively. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention provides an electric vehicle charging and discharging scheduling optimization method, device, medium, and program, which are used to optimize EV charging and discharging scheduling while taking into account user response willingness, achieve a balance of interests on both the supply and demand sides, and enhance user response enthusiasm and the economy of the power grid.

[0005] The present invention achieves the above technical objectives through the following technical means.

[0006] An electric vehicle charging and discharging scheduling optimization method is proposed, which takes the minimum distribution network operating cost and the EV user charging cost as the goal and optimizes the electric vehicle scheduling method; wherein:

[0007] The cost of operating a distribution network includes the cost of purchasing electricity less the revenue from selling electricity;

[0008] The EV user charging cost includes the battery loss cost of each EV minus the dispatch revenue.

[0009] Furthermore, the distribution network operation cost is:

[0010] f1=f buy -f sell +fLS +f R

[0011]

[0012]

[0013] Where, f buy is the electricity purchase cost of the distribution network, f sell is the electricity sales revenue of the distribution network, f LS is the transmission loss cost of the distribution network, f R is the load fluctuation penalty fee of the distribution network; l represents the time period, L is the maximum time period, and are the unit price and quantity of electricity purchased in the first period respectively; and are the charging price and charging amount of the i-th EV in time period l; ω f1 is the network loss cost coefficient, is the amount of power lost in the distribution network during period l; f2 is the load fluctuation penalty coefficient, D Lv is the load fluctuation of the distribution network in a day, is the total load of the distribution network during period l, D avg is the average load of the distribution network in a day, j represents the node (charging station) number in the distribution network, J is the number of nodes in the distribution network, are respectively the conventional load, distributed generation output, and charging load of the jth node in the distribution network in period l;

[0014] The EV user charging cost is:

[0015]

[0016] Where, is the battery loss cost of the i-th EV, is the dispatching benefit of the i-th EV, and N is the number of EV clusters; and They are the unit battery loss generated by EV during charging and discharging, and are the time when the i-th EV enters and leaves the grid, and are the additional charging power and additional discharging power generated by the i-th EV at time t, respectively; and are the charging power and discharging power of the i-th EV at time t, is the incentive electricity price for the i-th EV at time t.

[0017] Furthermore, the following constraints are included in the optimization:

[0018] Power flow constraints:

[0019] V j,min ≤V j ≤V j,max

[0020] I k ≤I k,max

[0021] Where V j Represents the voltage of node j in the distribution network, V j,min and V j,max are the lower and upper limits of the node j voltage, I k is the current on branch k in the distribution network, I k,max is the upper limit of the current on branch k;

[0022] Scheduling capacity constraints:

[0023]

[0024] Where, and are the charging power and discharging power dispatched by the distribution network at time t, and are the real-time charging capacity and discharging capacity of the EV cluster at time t, respectively;

[0025] Electric vehicle charging and discharging power constraints:

[0026]

[0027] Where, and are the lower and upper limits of the charging power for the i-th EV respectively;

[0028] Off-grid SOC constraints:

[0029]

[0030] Where, is the time when the i-th EV is off the grid SOC, S i,end is the off-grid SOC requirement of the i-th EV;

[0031] Remaining power constraint:

[0032] S i,min ≤S t,i ≤S i,max

[0033] Where S i,min and S i,max are the lower and upper limits of the SOC of the t-th electric vehicle respectively;

[0034] Charging and discharging electricity price constraints:

[0035]

[0036] Where, and are the lower and upper limit coefficients of electricity prices, is the time-of-use electricity price at time t.

[0037] Furthermore, the real-time charging capacity and discharging capacity of the EV cluster are:

[0038]

[0039] Where, and are the maximum charging power and maximum discharging power of the t-th EV, and are the maximum and minimum values ​​of all charging responsiveness in the EV cluster, and are the maximum and minimum values ​​of all discharge responsiveness in the EV cluster, respectively;

[0040] The charge and discharge responsiveness are:

[0041]

[0042] Where S t,i is the SOC of the i-th EV at time t, and are the charging responsiveness and discharging responsiveness of the i-th EV user, respectively, and a c 、a d 、b c 、b d is the response coefficient, c c and c d is the response constant.

[0043] Furthermore, the incentive electricity price is:

[0044]

[0045] Where τ is the incentive coefficient for scheduling, Ψ is the random disturbance factor, the value range is [0, 0.005], and the unit is yuan / (hour*kWh).

[0046] Furthermore, the particle swarm algorithm is used to optimize the scheduling by minimizing the following objective function:

[0047]

[0048] Where F is the comprehensive cost, λ1 and λ2 are the weights of the distribution network operation cost and EV user charging cost, respectively, λ1+λ2=1, f 1,min and f 1,max are the minimum and maximum values ​​of f1, respectively. 2,min and f 2,max are the minimum and maximum values ​​of f2 respectively.

[0049] Furthermore, during optimization, EV user clusters are classified by calculating the comprehensive anxiety cost, where:

[0050] The combined anxiety cost is:

[0051]

[0052] Where, is the comprehensive anxiety cost of the i-th EV user, is the battery loss anxiety cost of the i-th EV user, is the time loss anxiety cost of the i-th EV user, and ε is the conversion factor; is the battery loss cost benchmark value, To take all The maximum value in the binomial distribution, δ is The probability of occurrence; γ is the conversion factor, α∈[0,1] and β∈(-∞,0)∪(0,+∞) are decision factors, T max is the maximum on-grid duration in the EV cluster;

[0053] EV user clusters are divided into the following three categories:

[0054]

[0055] S1 is a high willingness group, S2 is a medium willingness group, S3 is a low willingness group, and C S1 、C S2 is the threshold;

[0056] During the optimization process, the number of groups in the three categories is controlled by adjusting the values ​​of λ1 and λ2.

[0057] A computer device comprising a memory and a processor;

[0058] The memory is used to store computer programs;

[0059] The processor is used to execute the computer program and implement the above-mentioned electric vehicle charging and discharging scheduling optimization method when executing the computer program.

[0060] A computer-readable storage medium stores a computer program, which, when executed by a processor, causes the processor to execute the above-mentioned electric vehicle charging and discharging scheduling optimization method.

[0061] A computer program product includes a computer program, which implements the above-mentioned electric vehicle charging and discharging scheduling optimization method when executed by a processor.

[0062] The beneficial effects of the present invention are:

[0063] (1) The present invention provides a method, device, medium, and program for optimizing the charging and discharging scheduling of electric vehicles, which comprehensively considers the operating costs of the distribution network and the charging costs of EV users to optimize the charging and discharging scheduling plan of electric vehicles, thereby achieving a dynamic balance of interests on both the supply and demand sides.

[0064] (2) In the present invention, when calculating the operating cost of the distribution network, the load fluctuation penalty fee of the distribution network is included, thereby further promoting the effect of peak shaving and valley filling of the power grid load.

[0065] (3) The dynamic optimization framework proposed in this paper improves user participation through differentiated incentive strategies, enhances the dispatchability potential of EV clusters, and provides a scalable solution for the economic and safe operation of distribution networks under high EV penetration.

[0066] (4) During the optimization process, the present invention takes into account the scheduling capacity constraints of the EV cluster and establishes the influence relationship between user responsiveness and electricity price and SOC to achieve an accurate estimate of the actual dispatchable capacity of the EV cluster, so that the optimized scheduling plan can be closer to the actual user response willingness and can be actually implemented.

[0067] (5) The present invention quantifies the anxiety of EV users about the additional battery loss and time loss caused by charging and discharging scheduling by giving a comprehensive anxiety cost, and classifies the EV cluster based on this to form feedback on the optimization results; among them, the group with higher comprehensive anxiety cost has lower willingness to respond to scheduling, and the corresponding scheduling plan is more difficult to implement; therefore, by controlling the proportion of the three groups, the implementation ability of the optimized scheduling plan can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 Comparison of SOC curves for those participating in distribution network dispatching;

[0069] Figure 2 This is the IEEE33-node distribution network system diagram in the test case;

[0070] Figure 3 is the distributed generation output curve in the test case;

[0071] Figure 4 is the cluster response ratio under different cluster sizes;

[0072] Figure 5 This is a graph showing how charging responsiveness changes with electricity price and SOC;

[0073] Figure 6 This is a graph showing how discharge responsiveness changes with electricity price and SOC;

[0074] Figure 7 Comparison of load curves under three test scenarios in the test case;

[0075] Figure 8 This is a comparison of EV charging and discharging under three test scenarios in the test case;

[0076] Figure 9 is the average incentive electricity price curve at each moment in scenario 3 in the test case. DETAILED DESCRIPTION

[0077] The embodiments of the present invention are described in detail below. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, but should not be understood as limiting the present invention.

[0078] 1. Technical Solution

[0079] 1. User cluster classification

[0080] (1) Battery loss anxiety cost

[0081] When EV (electric vehicle) participates in additional charging and discharging, its battery life may be affected to a certain extent, thereby affecting the enthusiasm of EV users to participate in distribution network scheduling. Figure 1 The figure shows the SOC changes of the EV under two scenarios: whether the EV participates in charge and discharge scheduling. Among them, curve Ma shows the case when the user does not participate in scheduling, and the EV is only charged normally after being connected to the grid, and the SOC remains unchanged after rising to the full charge state. Curve Mb describes the SOC change characteristics of the EV after being connected to the grid due to the implementation of charge and discharge regulation when the user participates in scheduling.

[0082] Taking curve Ma as a reference, for curve Mb, when the user participates in the regulation, after the EV completes the basic charging, at t a The additional charge and discharge phase begins at time t b End this regulation behavior at the moment, and then return to the normal charging state until it is fully charged and disconnected from the grid. a Time and t bThe SOC value at the moment is the same, so t a to t b The time period is regarded as the additional charging and discharging time performed by EVs during the grid dispatch process.

[0083] For EVs, additional battery loss occurs when participating in grid dispatch, which is expressed as:

[0084]

[0085] Where, is the battery loss cost (expense) of the i-th EV, and The unit battery loss (cost) generated by EV during charging and discharging, and are the time when the i-th EV enters and leaves the grid, and are the additional charging power and additional discharging power generated by the i-th EV at time t (during the entire charging cycle), that is, as in the example above, a to t b If t is not within the above additional charge and discharge period, the corresponding and is 0.

[0086] Cost of battery loss The resulting emotional rejection of EV users can be quantified as the following battery loss anxiety costs:

[0087]

[0088] Where, is the battery loss anxiety cost of the i-th EV user, λ b >1 is the rejection coefficient, H is a 0-1 variable that obeys the binomial distribution. When H = 0, the battery loss anxiety cost is equal to the battery loss cost. When H = 1, the battery loss anxiety cost will be magnified by λ on the basis of the battery loss cost. b times, thus reflecting the characteristics of a user group with a strong rejection mentality.

[0089] When considering the battery loss anxiety cost, the binomial distribution is usually used to describe the loss events that each user may face. However, since this paper focuses on the willingness of the entire user group rather than the specific circumstances of each individual, the binomial distribution is converted to a group-level representation. The battery loss anxiety cost of the i-th EV user can be expressed as:

[0090]

[0091] Where, is the battery loss cost benchmark value, To take all The maximum value in the binomial distribution, δ is Probability of occurrence.

[0092] This model introduces the rejection psychology coefficient based on traditional battery loss to better quantify the user's anxiety, and combines SOC and time information to provide a more comprehensive analysis.

[0093] (2) Time loss anxiety cost

[0094] When EVs participate in distribution network regulation, the charging mode shifts from "charging at any time" to "charging at a limited time." This change reduces users' flexibility in vehicle usage. Therefore, a new model is proposed to measure the anxiety caused by this time restriction. This anxiety is quantified by the EV's on-grid time and is defined as follows:

[0095]

[0096] Where, is the time loss anxiety cost of the i-th EV user, is the time that the i-th EV is online, γ is the conversion factor, α∈[0,1] and β∈(-∞,0)∪(0,+∞) are decision factors, T max is the maximum online time in the EV cluster.

[0097] As the time EVs are online increases, the control of charging time reduces users' flexible car usage time, thereby causing higher anxiety costs.

[0098] This model introduces decision factors α and β to simulate different users' sensitivity to network duration. The anxiety costs of different user groups vary depending on the combination of these factors. The model shows how the anxiety costs change under different decision factors, indicating that the maximum anxiety cost is correlated with the maximum network duration. The changes in decision factors reflect users' sensitivity to different charging times.

[0099] (3) Comprehensive anxiety costs

[0100] Based on the anxiety cost of battery loss and the anxiety cost of time loss, the following comprehensive anxiety cost is proposed to quantify the willingness of EV users to participate in distribution network scheduling:

[0101]

[0102] Where, is the comprehensive anxiety cost of the i-th EV user, f i poweris the economic benefit brought by the dispatch of the i-th EV, specifically the discharge subsidy minus the charging cost, and ε is the conversion factor.

[0103] This model considers the benefits of scheduling and the additional battery loss and time anxiety that users may experience during the scheduling process, helping to determine users' willingness to participate in scheduling. A larger value indicates a stronger user's willingness to participate.

[0104] (4) User cluster classification

[0105] As the comprehensive anxiety cost increases, the responsiveness (willingness to participate) of EV users decreases. Therefore, in this embodiment, the comprehensive anxiety cost The user clusters are divided into three categories based on the level of

[0106]

[0107] Among them, S1 is a high willingness group, S2 is a medium willingness group, S3 is a low willingness group, and the threshold C S1 、C s2 It is set by humans based on actual survey conditions.

[0108] 2. Electricity Price-Energy Consolidation Response Surface

[0109] The traditional user responsiveness model only uses electricity price as the only variable to calculate the responsiveness, assuming it to be a fixed value. However, in reality, the user's response behavior is not only affected by the charging and discharging electricity price, but is also deeply restricted by EV and SOC. The interaction of these multiple factors makes the responsiveness more complicated, and it is necessary to consider the dual effects of battery power and electricity price changes. Based on the Weber-Fechner law, the present invention establishes an electricity price-power (State of Charge, SOC) collaborative responsiveness surface, analyzes the nonlinear effects of charging and discharging prices and SOC on user response behavior, and thereby establishes the charging and discharging capacity constraints of electric vehicle clusters.

[0110] (1) Incentive electricity prices

[0111] To encourage EV users to respond to dispatch, the following incentive electricity prices are set:

[0112]

[0113] Where, is the incentive electricity price for the i-th EV at time t. The incentive electricity price varies according to the charging and discharging behavior. It can be divided into charging incentive price and discharging incentive price, among which the incentive price is applied when EV is charging. Charge EV users and apply this incentive electricity price when EV is discharged Subsidize EV users; is the time-of-use electricity price at time t, τ is the incentive coefficient for scheduling, Ψ is the random disturbance factor, the value range is [0, 0.005], and the unit is yuan / (hour*kWh).

[0114] Note: Both charging incentive price and discharging incentive price are in the form of time-of-use electricity price. When performing calculations, two different sets of parameters τ and Ψ can be set according to actual needs, so that the charging incentive electricity price and the discharging incentive electricity price for the same EV at the same time are differentiated.

[0115] (2) Responsiveness

[0116] Electricity price and SOC are two key factors in modeling charge and discharge responsiveness. First, changes in electricity prices directly influence user charging behavior: high prices encourage users to charge less, while low prices encourage more charging. Similarly, changes in discharge prices can significantly influence user discharge responsiveness: high prices encourage discharge, while low prices reduce discharge.

[0117] In addition to electricity price, SOC level is also a significant factor. When SOC is higher, users are more inclined to discharge rather than charge; conversely, when SOC is lower, the demand for charging is stronger. Therefore, the user's charge and discharge responsiveness is not only a function of electricity price but also closely related to SOC.

[0118] According to the Weber-Fechner law, users perceive price and SOC fluctuations relative, not absolute. This means that users tend to react to changes based on the percentage of change rather than the absolute price change. Based on this principle, user responsiveness can be quantified using the following formula:

[0119]

[0120] Where S t,i is the SOC of the i-th EV at time t, and are the charging responsiveness and discharging responsiveness of the i-th EV user, respectively, and a c 、a d 、b c 、b d is the response coefficient, c c and c d is the response constant.

[0121] Then, based on the EV connection and disconnection time, the SOC at each moment and the set electricity price, the real-time charging capacity of the EV cluster at a certain moment (time t) can be calculated. and discharge capacity :

[0122]

[0123] Where, and are the maximum charging power and maximum discharging power of the i-th EV, and All in the EV cluster The maximum and minimum values ​​are used to indicate the upper and lower limits of EV user charging responsiveness. and All in the EV cluster The maximum and minimum values ​​in are used to represent the upper and lower limits of EV user discharge responsiveness; N is the number of dispatchable EVs (the number of EV clusters).

[0124] 3. Optimization of charging and discharging scheduling

[0125] In order to achieve the dual optimization of distribution network operation cost and EV user economic cost, a multi-objective optimization model was constructed and solved by particle swarm optimization.

[0126] Disorderly charging of large-scale electric vehicles can exacerbate grid load peaks and impact system stability. To address this issue, a multi-objective optimization model is proposed to coordinate distribution network operational requirements with the interests of EV users, achieving a dynamic balance between the two. This model aims to minimize both distribution network operating costs and EV user charging costs, thereby balancing grid stability with user economics, achieving a coordinated optimization of both interests.

[0127] (1) Distribution network operating costs

[0128] f1=f buy -f sell +f LS +f R

[0129] Where, f buy is the electricity purchase cost of the distribution network, f sell is the electricity sales revenue of the distribution network, f LS is the transmission loss cost of the distribution network, f R is the load fluctuation penalty fee of the distribution network; their respective expressions are:

[0130]

[0131] In the formula, l represents the time period, and L is the maximum time period. For example, if a day is divided into time periods at intervals of 1 hour, then L = 24. and are the unit price and quantity of electricity purchased in the first period respectively;

[0132]

[0133] Where, and are the charging price and charging amount of the i-th EV in time period l respectively;

[0134] Note: and The relationship between them is: when time t falls within the range of time period l, then l a and l b are the starting and ending times of period l, is the charging power of the i-th EV at time t;

[0135]

[0136] Where, ω f1 is the network loss cost coefficient, is the amount of power lost in the distribution network during period l, which can be calculated based on the current and resistance on the transmission lines between nodes in the distribution network during each period;

[0137]

[0138]

[0139] Where, ω f2 is the load fluctuation penalty coefficient, D Lv is the load fluctuation of the distribution network in a day, is the total load of the distribution network during period l, D avg is the average load of the distribution network in one day, j represents the node number in the distribution network, J is the number of nodes in the distribution network, are respectively the conventional load, distributed power generation output, and charging load of the j-th node in the distribution network in period l, where distributed power generation refers to new energy power generation such as photovoltaic power generation and wind power generation.

[0140] The above distribution network operation cost model aims to optimize the dispatching strategy and improve the economy and stability of the power grid operation by balancing the network losses and load fluctuations of the distribution network.

[0141] (2) EV user charging costs

[0142] To reduce the economic burden on EV users and encourage them to actively participate in grid dispatch, the optimization goal not only focuses on maximizing user benefits, but also comprehensively considers battery loss to ensure long-term economic efficiency. The optimization strategy is developed from the following perspectives:

[0143]

[0144] Where, is the battery loss cost of the i-th EV, and its expression is as described above; f i power is the dispatch benefit of the i-th EV, which is expressed as follows:

[0145]

[0146] Where, and where is the charging power and discharging power of the i-th EV at time t, respectively. Note: EVs participate in grid scheduling. When the grid load is high, a higher incentive price is given, and users discharge electricity during this period to earn electricity price benefits. When the grid load is low, a lower incentive price is set, and users charge during this low-price period.

[0147] (3) Multi-objective optimization

[0148] Since the EV user charging cost and the distribution network operation cost may differ greatly in value, the maximum-minimum normalization method is introduced to obtain the following comprehensive objective function:

[0149]

[0150] Where F is the comprehensive cost, λ1 and λ2 are the weights of the distribution network operation cost f1 and the EV user charging cost f2, respectively, λ1+λ2=1, f 1,min and f 1,max are the minimum and maximum values ​​of f1, respectively. 2,min and f 2,max are the minimum and maximum values ​​of f2 respectively.

[0151] (4) Constraints

[0152] In the process of optimizing the comprehensive objective function, the following constraints are set:

[0153] 1) Power flow constraints:

[0154] V j,min ≤V j ≤V j,max

[0155] I k ≤I k,max

[0156] Where V j Represents the voltage of node j in the distribution network, V j,min and V j,max are the lower and upper limits of the node j voltage, I k is the current on branch k in the distribution network, I k,max is the upper limit of the current on branch k;

[0157] 2) Scheduling capacity constraints:

[0158]

[0159] Where, and They are the charging power dispatched by the distribution network at time t (the charging power of each EV The sum of the discharge power and the discharge power of each EV For the charging and discharging scheduling of EV clusters, the charging and discharging capacity limitations after user response must be considered, that is, to ensure that charging and discharging are within the dispatchable capacity range;

[0160] 3) Electric vehicle charging and discharging power constraints:

[0161]

[0162] Where, and are the lower and upper limits of the charging power for the i-th EV respectively;

[0163] 4) Off-grid SOC constraints:

[0164]

[0165] Where, is the time when the i-th EV is off the grid SOC, S i,end is the off-grid SOC requirement of the i-th EV, usually defaulted to S i,end The value is 0.9, which means that the SOC is at least 0.9 when the vehicle leaves;

[0166] 5) Remaining power constraint:

[0167] S i,min ≤S t,i ≤S i,max

[0168] Where S i,min and S i,max are the lower and upper limits of the SOC of the i-th electric vehicle respectively;

[0169] 6) Charge and discharge price constraints:

[0170]

[0171] Where, and The lower and upper limit coefficients of the electricity price are set to control user costs and ensure the smooth operation of the power grid. In this embodiment, the upper and lower limits of the charging and discharging electricity prices are set. In this embodiment, the values ​​of the two are 0.3 and 0.9 respectively;

[0172] (5) Solution

[0173] This example uses a particle swarm optimization algorithm to minimize the overall cost F and solve for an electric vehicle charging and discharging scheduling plan (the charging and discharging power of each electric vehicle during each time period). During the optimization process, the number of groups S1, S2, and S3 is determined by calculating the comprehensive anxiety cost of each EV user. When the proportion of the S3 group is high, the EV user charging cost weight λ2 is increased to prevent users from dropping out due to high anxiety costs. When the proportion of the S1 group is high, the EV user charging cost weight λ2 is reduced to fully utilize its low anxiety characteristics to maximize grid revenue. Ultimately, the proportions of the three groups are within the required range to ensure that the scheduling plan can be implemented by actual EV users.

[0174] 2. Test and Verification

[0175] like Figure 2 The figure shows the distribution network system diagram used for this simulation test, which includes one EV charging station, four photovoltaic power stations, and two wind power stations.

[0176] like Figure 3 The figure shows the output curve of distributed power generation. In the figure, PV1, PV2, PV3, and PV4 are the outputs of four photovoltaic power stations, and WT1 and WT2 are the outputs of wind power stations.

[0177] The time-of-use electricity prices and distribution network electricity purchase unit prices for each period are shown in Tables 1 and 2 below.

[0178] Table 1: Time-of-use electricity prices

[0179]

[0180] Table 2: Electricity purchase price

[0181]

[0182] like Figure 4 The figure shows the cluster response ratio for different cluster sizes. As the size of the EV cluster increases, the cluster response ratio gradually stabilizes. To eliminate the influence of data point overlap, the variance of 50 experiments was calculated and a variance curve was plotted. The results show that the variance of the response ratio gradually decreases with increasing cluster size. Once the cluster size reaches a certain value, the variance stabilizes, with the range remaining within 0.1. These test results indicate that larger EV clusters can more stably reflect users' willingness to participate in distribution network regulation, and this stability increases with cluster size. In contrast, in smaller clusters, individual decisions have a greater impact on the overall response, resulting in larger fluctuations in the response ratio. Therefore, larger clusters lead to more consistent user decisions regarding participation in regulation, and the system becomes more stable.

[0183] Figure 5 and Figure 6 The synergistic effect surfaces of charging and discharging responsiveness at different charging and discharging electricity prices and SOC levels are presented. Simulation results show that charging responsiveness exhibits a nonlinear decreasing trend with increasing charging price, especially when the SOC value exceeds 0.7, where the rate of responsiveness decreases significantly, indicating that users are significantly less sensitive to electricity price changes at high SOC levels. Discharging responsiveness exhibits an exponential growth characteristic with increasing discharge price, and the growth slope of responsiveness increases significantly when the SOC exceeds 0.5, indicating a significant threshold effect in users' response to higher discharge price incentives. Further analysis shows that charging responsiveness reaches its upper limit in the range where charging price is less than 0.4 yuan / kWh and SOC is less than 0.3, while discharging responsiveness approaches saturation when the discharge price exceeds 0.9 yuan / kWh and SOC is greater than 0.8. The simulation results reveal that user responsiveness to charging and discharging is affected by electricity price and SOC, but the responsiveness varies from user to user. Specifically, the charging responsiveness decreases with increasing charging electricity prices, while the discharging responsiveness increases with increasing discharge prices. Furthermore, the responsiveness changes are subject to clear upper and lower limits, a characteristic that is reflected in both the charging and discharging processes.

[0184] In order to evaluate the technical effect of the charge-discharge scheduling optimization method of the present invention, the following three different charge-discharge scenarios are designed for simulation comparison:

[0185] Scenario 1: EVs are charged in an unordered manner, ignoring the impact of electricity prices (no scheduling of charging and discharging power for each EV);

[0186] Scenario 2: Only time-of-use electricity prices are used for guidance, and the rest of the EVs are charged and discharged in an orderly manner based on the present invention (the charging and discharging power of each EV is scheduled);

[0187] Scenario 3: Using the method of the present invention, incentive electricity prices are set to guide EVs to charge and discharge in an orderly manner (scheduling the charging and discharging power of each EV).

[0188] The simulation results are compared with the load curve and the change of EV charging and discharging amount. Figure 7 and Figure 8 .

[0189] like Figure 7 and Figure 8As shown in the figure, in scenario 1, due to uncontrolled charging, EVs tend to charge during the peak hours of the power grid, resulting in a "peak on peak" phenomenon in the load curve. In contrast, the time-of-use electricity price guidance strategy in scenario 2 enables EVs to charge during off-peak hours and discharge during peak hours, thereby reducing load fluctuations. In scenario 3, an orderly charging and discharging strategy optimized based on the charging and discharging scheduling optimization method of the present invention is adopted. EVs charge during off-peak hours and discharge during peak hours, with minimal fluctuations in the load curve, showing the optimal peak-shaving and valley-filling effect. These results show that the incentive electricity price mechanism can effectively guide EV users to optimize their charging and discharging behaviors and significantly reduce load fluctuations. Compared with other scenarios, the optimized scheduling strategy proposed in this paper shows obvious advantages and can more effectively improve the stability and security of the power grid.

[0190] like Figure 9 The figure shows the average charging incentive price and discharging incentive price (discharge compensation price in the figure) at each time point, as determined by the present invention in Scenario 3. Regulating EV charging and discharging behavior through incentive prices can effectively guide user decisions. During off-peak hours, the EV charging incentive price is lower than the TOU price, attracting a large number of users to charge during these times to reduce charging costs. Because the discharge incentive compensation is at its lowest level, users are reluctant to discharge during these times, avoiding losses caused by the discharge benefit being lower than the charging cost. During stable load periods, the EV charging incentive price remains lower than the TOU price, encouraging some users who originally planned to charge during peak load periods to shift their charging needs to these times. Furthermore, the discharge incentive compensation is slightly higher than the TOU price, prompting some EV users to discharge during these times due to their ample idle time. During peak load periods, the EV discharge incentive compensation reaches its highest level, attracting a large number of users to discharge their vehicles, taking advantage of the difference in charging and discharging prices to offset charging costs. Because charging prices are higher during these times, users, while still meeting their electricity needs, tend to avoid charging during these times to reduce unnecessary charging expenses.

[0191] Table 3 shows the operating costs of the distribution network under different scenarios, including upstream grid power purchase costs, network losses, charging station revenue, and load fluctuation penalties. Compared with Scenario 1, Scenario 2 and Scenario 3 show significant reductions in upstream grid power purchase costs, network losses, and load fluctuation penalties. In particular, the load fluctuation penalty is reduced by 777.4 yuan and 940.4 yuan, respectively. Although charging station revenue in Scenario 2 and Scenario 3 decreases compared to Scenario 1, this phenomenon can be attributed to two factors: first, EVs reduce their own charging costs by adjusting their charging and discharging times; second, the distribution network provides corresponding price discounts to encourage EVs to participate in demand response. Overall, the total operating costs of the distribution network in Scenario 2 and Scenario 3 are reduced by 476.4 yuan and 544.4 yuan, respectively, compared to Scenario 1.

[0192] Table 3: Distribution network operating costs under different scenarios

[0193]

[0194] III. Devices, Storage Media, and Program Products

[0195] 1. Based on the same inventive concept as the above-mentioned electric vehicle charging and discharging scheduling optimization method, the present application also provides an electronic device, which includes a processor and a memory, and the memory stores computer-readable code, wherein, when the computer-readable code is executed by the processor, the electric vehicle charging and discharging scheduling optimization method of the present invention is implemented.

[0196] The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium can store an operating system and computer-readable code. The computer-readable code includes program instructions that, when executed, cause the processor to execute the electric vehicle charging and discharging scheduling optimization method. The processor provides computing and control capabilities, supporting the operation of the entire electronic device. The memory provides an environment for the computer-readable code in the non-volatile storage medium to run. When executed by the processor, the computer-readable code causes the processor to execute the electric vehicle charging and discharging scheduling optimization method.

[0197] It should be understood that the processor may be a central processing unit, other general-purpose processors, digital signal processors, application-specific integrated circuits, field programmable gate arrays or other programmable logic devices, transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or any conventional processor.

[0198] 2. This application also provides a readable storage medium, which can be the internal storage unit of the electronic device described in the aforementioned embodiment, such as the hard disk or memory of the computer device. The readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart memory card, secure digital card, etc. equipped with the electronic device.

[0199] 3. The present application also provides a computer program product, comprising a computer program or instructions, which, when executed by a processor, implements the electric vehicle charging and discharging scheduling optimization method of the present invention.

[0200] The present invention is not limited to the above-mentioned embodiments. Any obvious improvement, replacement or modification that can be made by those skilled in the art without departing from the essence of the present invention shall fall within the scope of protection of the present invention.

Claims

1. A method for optimizing charging and discharging scheduling of electric vehicles, characterized by: Taking the minimization of distribution network operation cost and EV user charging cost as the goal, the dispatching method of electric vehicles is optimized; among them: The cost of operating a distribution network includes the cost of purchasing electricity less the revenue from selling electricity; The EV user charging cost includes the battery loss cost of each EV minus the dispatch revenue.

2. The electric vehicle charging and discharging scheduling optimization method according to claim 1, characterized in that: The distribution network operation cost is: f1=f buy -f sell +f LS +f R f R =ω f2 D Lv Where, f buy is the electricity purchase cost of the distribution network, f sell is the electricity sales revenue of the distribution network, f LS is the transmission loss cost of the distribution network, f R is the load fluctuation penalty fee of the distribution network; l represents the time period, K is the maximum time period, and are the unit price and quantity of electricity purchased in the first period respectively; and are the charging price and charging amount of the i-th EV in time period l; ω f1 is the network loss cost coefficient, is the amount of power lost in the distribution network during period l; f2 is the load fluctuation penalty coefficient, D Lv is the load fluctuation of the distribution network in a day, is the total load of the distribution network during period l, D avg is the average load of the distribution network in a day, j represents the node (charging station) number in the distribution network, J is the number of nodes in the distribution network, are respectively the conventional load, distributed generation output, and charging load of the jth node in the distribution network in period l; The EV user charging cost is: Where, is the battery loss cost of the i-th EV, f i power is the dispatching benefit of the i-th EV, and N is the number of EV clusters; and They are the unit battery loss generated by EV during charging and discharging, and are the time when the i-th EV enters and leaves the grid, and are the additional charging power and additional discharging power generated by the i-th EV at time t, respectively; and are the charging power and discharging power of the i-th EV at time t, is the incentive electricity price for the i-th EV at time t.

3. The electric vehicle charging and discharging scheduling optimization method according to claim 2, characterized in that: The following constraints are included in the optimization: Power flow constraints: In j,min ≤V j ≤V j,max I k ≤I k,max Where V j Represents the voltage of node j in the distribution network, V j,min and V j,max are the lower and upper limits of the node j voltage, I k is the current on branch k in the distribution network, I k,max is the upper limit of the current on branch k; Scheduling capacity constraints: Where, and are the charging power and discharging power dispatched by the distribution network at time t, and are the real-time charging capacity and discharging capacity of the EV cluster at time t, respectively; Electric vehicle charging and discharging power constraints: Where, and are the lower and upper limits of the charging power for the i-th EV respectively; Off-grid SOC constraints: Where, is the time when the i-th EV is off the grid SOC, S i,end is the off-grid SOC requirement of the i-th EV; Remaining power constraint: S i,min ≤S t,i ≤S i,max Where S i,min and S i,max are the lower and upper limits of the SOC of the i-th electric vehicle respectively; Charging and discharging electricity price constraints: Where, and are the lower and upper limit coefficients of electricity prices, is the time-of-use electricity price at time t.

4. The electric vehicle charging and discharging scheduling optimization method according to claim 3, characterized in that: The real-time charging capacity and discharging capacity of the EV cluster are: Where, and are the maximum charging power and maximum discharging power of the i-th EV, and are the maximum and minimum values ​​of all charging responsiveness in the EV cluster, and are the maximum and minimum values ​​of all discharge responsiveness in the EV cluster, respectively; The charge and discharge responsiveness are: Where S t,i is the SOC of the i-th EV at time t, and are the charging responsiveness and discharging responsiveness of the i-th EV user, respectively, and a c 、a d 、b c 、b d is the response coefficient, c c and c d is the response constant.

5. The method for optimizing charging and discharging scheduling of electric vehicles according to any one of claims 2 to 4, characterized in that: The incentive electricity price is: Where τ is the incentive coefficient for scheduling, Ψ is the random disturbance factor, the value range is [0, 0.005], and the unit is yuan / (hour*kWh).

6. The electric vehicle charging and discharging scheduling optimization method according to claim 2, characterized in that: The particle swarm algorithm is used to optimize the scheduling by minimizing the following objective function: Where F is the comprehensive cost, λ1 and λ2 are the weights of the distribution network operation cost and EV user charging cost, respectively, λ1+λ2=1, f 1,min and f 1,max are the minimum and maximum values ​​of f1, respectively. 2,min and f 2,max are the minimum and maximum values ​​of f2 respectively.

7. The electric vehicle charging and discharging scheduling optimization method according to claim 6, characterized in that: During optimization, EV user clusters are classified by calculating the comprehensive anxiety cost, where: The combined anxiety cost is: Where, is the comprehensive anxiety cost of the i-th EV user, is the battery loss anxiety cost of the i-th EV user, is the time loss anxiety cost of the i-th EV user, and ε is the conversion factor; is the battery loss cost benchmark value, To take all The maximum value in the binomial distribution, δ is The probability of occurrence; γ is the conversion factor, α∈[0,1] and β∈(-∞,0)∪(0,+∞) are decision factors, T max is the maximum on-grid duration in the EV cluster; EV user clusters are divided into the following three categories: S1 is a high willingness group, S2 is a medium willingness group, S3 is a low willingness group, and C S1 、C S2 is the threshold; During the optimization process, the number of groups in the three categories is controlled by adjusting the values ​​of λ1 and λ2.

8. A computer device, characterized in that: including memory and processor; The memory is used to store computer programs; The processor is used to execute the computer program and implement the electric vehicle charging and discharging scheduling optimization method according to any one of claims 1 to 7 when executing the computer program.

9. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor is caused to execute the electric vehicle charging and discharging scheduling optimization method according to any one of claims 1 to 7.

10. A computer program product, characterized in that: The method comprises a computer program, which, when executed by a processor, implements the electric vehicle charging and discharging scheduling optimization method according to any one of claims 1 to 7.

Images