Optimization method for electric vehicle participating in photovoltaic consumption considering user response uncertainty
By constructing a response rate model that takes into account both the economic and social attributes of users, and using the Logistic function and NSGA-II algorithm to optimize electric vehicle charging strategies, the uncertainty of electric vehicle user response behavior is solved, thereby improving photovoltaic power consumption efficiency and user satisfaction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NANJING UNIV OF POSTS & TELECOMM
- Filing Date
- 2026-03-27
- Publication Date
- 2026-06-19
AI Technical Summary
Existing technologies fail to accurately characterize the uncertainty of electric vehicle user response behavior, leading to a mismatch between charging scheduling strategies and actual demand. Furthermore, they cannot effectively utilize the diversity of electric vehicle resources on the grid side, impacting grid security and stability, as well as social aspects. The uncertainty of electric vehicle user response behavior in existing technologies also results in low photovoltaic absorption efficiency.
By constructing a response rate model that considers both the economic and social attributes of users, using the Logistic function to characterize the nonlinear mapping relationship between user response rate and charging time, and combining the NSGA-II algorithm to optimize electric vehicle charging strategies, the aggregator's revenue and user satisfaction are synergistically optimized.
It has improved the level of photovoltaic power consumption, achieved synergistic optimization of aggregator operating revenue and user satisfaction, and improved the efficiency of photovoltaic power consumption and the accuracy of user response behavior.
Smart Images
Figure CN122246820A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of power system demand response and distribution network optimization operation technology, and specifically relates to an optimization method for electric vehicles participating in photovoltaic power consumption that takes into account the uncertainty of user response. Background Technology
[0002] Currently, the penetration rate of distributed photovoltaic (PV) power on the distribution network side continues to increase, becoming an important component of the clean energy supply for the distribution network. However, the inherent intermittency, volatility, and randomness of PV output can easily lead to problems such as power backflow and power quality deterioration in the distribution network, severely restricting the efficient local consumption of PV resources and posing a serious challenge to the safe and stable operation of the power grid. Therefore, it is urgent to explore flexible and adjustable resources and optimize energy utilization efficiency. Electric vehicles, as flexible loads adjustable on the user side, have significant advantages such as flexible charging times and large controllability potential, and have received widespread attention in mitigating PV fluctuations and improving the local consumption capacity of PV.
[0003] Demand response technology, as a core means of flexible load dispatching, guides users to adjust their electricity consumption behavior through price signals and incentive mechanisms, playing a crucial role in resource optimization. Among these, incentive-based demand response, due to its direct regulation and high user participation, has become the mainstream application method for electric vehicles to participate in photovoltaic power consumption. However, solely considering the impact of economic incentives on users ignores the behavioral uncertainty caused by users' own willingness to respond and fails to accurately characterize the nonlinear relationship between user response rate and charging behavior, resulting in distorted user response characterization and a disconnect between dispatching strategies and actual demand. In real-world scenarios, the response behavior of electric vehicle users is not only influenced by economic factors such as incentive electricity prices but also by observing and imitating the social behavioral characteristics of other users. The dual effects of economic and social attributes make user response exhibit significant uncertainty. At the same time, traditional methods have not established accurate quantitative models of charging behavior, making it difficult to achieve dynamic mapping between user response rate and charging duration, leading to insufficient accuracy in charging load dispatching.
[0004] Furthermore, with the deepening of power market reform and the gradual improvement of existing incentive mechanisms, differentiated incentive pricing has become an important means to guide electric vehicles to participate in photovoltaic power generation. How to combine price incentives with the multi-attribute characteristics of users, construct a multi-objective optimization model that takes into account the interests of all parties, propose an optimization method for electric vehicles to participate in photovoltaic power generation that takes into account the uncertainty of user response, and achieve a synergistic improvement in the operating income of aggregators and user satisfaction has important practical significance and application value. Summary of the Invention
[0005] The purpose of this invention is to propose an optimization method for electric vehicles participating in photovoltaic power generation that takes into account the uncertainty of user response. This method comprehensively considers both the economic and social attributes of users, accurately quantifies the uncertainty of user response behavior, and constructs a multi-objective optimization model to improve the level of photovoltaic power generation and achieve the optimal balance between aggregator operating revenue and user satisfaction.
[0006] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:
[0007] An optimization method for electric vehicle participation in photovoltaic power generation that takes into account user response uncertainty includes the following steps:
[0008] S1. Input basic resource data such as regional distributed photovoltaic power output data and electricity price parameters, construct an economic response rate model based on incentive intensity, and quantify the differentiated economic attributes of users.
[0009] S2. Integrate economic incentives and user social attributes to establish a user response rate model and quantify the uncertainty of user response behavior.
[0010] S3. Based on user response behavior, establish a Logistic function model for extending electric vehicle charging time and determine the nonlinear mapping relationship between user response rate and charging time.
[0011] S4. Establish a multi-objective optimization model with the goals of maximizing the aggregator's operating revenue and optimizing user satisfaction. Constraints include electric vehicle charging power constraints, state of charge constraints, and off-grid target power constraints. Use the NSGA-II algorithm to solve the electric vehicle charging optimization decision and differentiated incentive pricing strategy.
[0012] The economic response rate model based on incentive intensity is as follows:
[0013] The incentive price intensity model is as follows:
[0014]
[0015] In the formula, The period during which incentivized prices are offered to aggregators. for Time-based incentive pricing, for Electricity price for charging during certain time periods; The discount incentive factor to promote photovoltaic power grid integration can be adjusted by the aggregator based on its own revenue and grid integration situation. This is the difference in electricity prices. Price incentive coefficient to absorb photovoltaic demand.
[0016] To quantify the impact of incentive intensity on user behavior, an EV user economic response intention model is established:
[0017]
[0018] in,
[0019]
[0020] In the formula, The slope of the user responsiveness line. For saturation excitation intensity, For the deviation values of response characteristics of different users, The slope of user responsiveness during initial incentive compensation. To incentivize and compensate for the user response rate slope after it has a strong guiding effect. The intensity of incentives has a strong guiding effect. User response behavior is influenced by multiple factors, including electricity price incentives, risk awareness, and behavioral preferences, which can be specifically divided into two dimensions: social attributes and economic attributes. Among them, social attributes refer to the inherent and relatively stable social characteristics of users, such as risk preferences and the "herd effect"; economic attributes refer to economic factors related to the level of return, including electricity price incentive strategies and user response costs.
[0021] The above steps consider the impact of economic incentives on users. First, a time-varying incentive mechanism oriented towards photovoltaic power consumption is designed from the perspective of aggregators. This takes into account user differences and establishes an economic response rate model that quantifies user willingness to respond, taking into account the intensity of economic incentives. This approach provides a more refined quantification of user economic response willingness compared to existing technologies, rather than relying on macro-level regulation of electric vehicles.
[0022] Furthermore, an economic response rate model based on incentive intensity is established:
[0023]
[0024] In the formula, For user economic response rate, This represents the upper limit of user economic responsiveness. The lower limit of user responsiveness. This is the response threshold.
[0025] The user response rate model that integrates social attributes and economic incentives is as follows:
[0026] During the process of user participation in demand response, a "herd effect" phenomenon occurs, where users observe others and learn by imitation. The social response rate after being influenced by this phenomenon is as follows:
[0027]
[0028] In the formula, For users within the group with varying degrees of willingness, This refers to the social attributes of this type of user. This represents the weight of herd mentality, indicating how users are influenced by the "herd effect".
[0029] Considering both social and economic impacts, a user response rate model is established:
[0030]
[0031] In the formula, The social attribute weighting coefficient. This represents the weighting coefficient for economic attributes.
[0032] Based on the user response rate model, user response data is obtained, thereby quantifying the uncertainty of user response behavior:
[0033]
[0034] In the formula, This indicates that the user is not participating in the response. This indicates user participation and response. Influenced by economic factors, each user exhibits differentiated responses due to varying social factors such as risk aversion; this defines the user's initial social attributes. This study considers the "herd effect" phenomenon—where users observe and learn from others—in the process of responding to user demands. A social response rate model is established, and then, taking into account the combined influence of users' social attributes and economic incentives, a user response rate model considering the coupling of these two attributes is developed to quantify the uncertainty of user response behavior. This method characterizes user response rates under bounded rationality, overcoming the limitations of traditional single-factor economic analysis.
[0035] The electric vehicle charging time extension model is as follows:
[0036]
[0037] In the formula, Actual charging time for users Base charging time for users. To maximize the extendable time, To extend the time response coefficient, Adjustment coefficients for user response rate and extension time.
[0038] To clarify the time window of the charging behavior model within the scheduling cycle, state variables were introduced. Describe the grid connection status of electric vehicles:
[0039]
[0040] In the formula, Indicates the first individual users Time period status, Indicates EV grid connection, Indicates EV disconnection from the network. For the first Users arrive at the charging station during their designated time period. For the first Each user plans to leave the charging station during a specific period. In reality, considering the differences in user response rates, the extended time mapping will not show a single linear relationship; it usually follows an "S"-shaped or exponential trend.
[0041] Therefore, the Logistic function, which has a good "S"-shaped fitting effect, is used here, based on the user response rate. The user's choice to extend the charging time is represented, with 0 indicating no extension and 1 indicating the maximum extension time. A function for extending the charging time of the EV is established, and state variables are introduced. Describe the grid connection status of electric vehicles.
[0042] The multi-objective optimization model, which aims to maximize the aggregator's operational revenue and optimize user satisfaction, is as follows:
[0043] Objective 1: Maximize revenue for aggregators, including charging revenue. Reduce waste of solar power revenue Grid-side electricity purchase cost Three parts:
[0044]
[0045] Charging revenue function:
[0046]
[0047] In the formula, For the first individual users EV charging power during specific time periods This is the simulated step size.
[0048] Reduce wastelight revenue function:
[0049]
[0050] In the formula, for Time period of light abandonment and power output To absorb the additional revenue generated by each unit of renewable energy. for Solar power output during specific time periods for Time period base load
[0051] Grid-side electricity purchase cost function:
[0052]
[0053] In the formula, for The power purchased by time-period aggregators from the grid side.
[0054] Objective 2: Optimal user satisfaction, including time satisfaction and cost satisfaction.
[0055]
[0056] Time-based satisfaction function:
[0057]
[0058] In the formula, The time sensitivity coefficient represents the user's psychological perception of charging delay.
[0059] Cost satisfaction function:
[0060]
[0061] In the formula, This represents the economic sensitivity coefficient, indicating how well users perceive the reduction in charging costs. To save on psychological expectations, the above steps construct a multi-objective optimization function aimed at maximizing aggregator revenue and optimizing user satisfaction, thereby achieving synergistic optimization of aggregator revenue and user satisfaction.
[0062] The constraints of the multi-objective optimization model, which aims to maximize the aggregator's operating revenue and optimize user satisfaction, include: electric vehicle behavior model constraints, electric vehicle charge constraints, electric vehicle off-grid target state of charge constraints, electric vehicle charging power constraints, grid power purchase constraints, and power balance constraints.
[0063] Electric vehicle behavior model constraints:
[0064] First, it is influenced by users' willingness to participate; second, it needs to consider whether users' travel needs are met. If users' willingness to participate is met, but the planned parking time is insufficient to charge to the expected level, users will abandon the response; otherwise, the system has the capability to optimize charging strategies. The expected basic charging time required for electric vehicles is as follows:
[0065]
[0066] In the formula, For the first Target state of charge of individual user electric vehicles For the first The state of charge of each user when joining the network For the first Individual user electric vehicle battery capacity, Maximum charging power for electric vehicles To improve the charging efficiency of electric vehicles.
[0067] Electric vehicle charge constraint:
[0068]
[0069] In the formula, For the first individual users The state of charge of electric vehicles during the time period For the first individual users The state of charge of electric vehicles during a given time period.
[0070] Off-grid target SOC constraint for electric vehicles:
[0071]
[0072] Electric vehicle charging power constraints:
[0073]
[0074] Power grid purchase constraints:
[0075]
[0076] In the formula, It is the maximum power that the aggregator can purchase from the grid.
[0077] Power balance constraints:
[0078]
[0079] In the formula, This represents the total number of electric vehicles.
[0080] The NSGA-II algorithm is used to solve the multi-objective optimization model, yielding the optimal incentive electricity price scheme and electric vehicle charging strategy, specifically:
[0081] Step 1: Set algorithm parameters and basic data, including population size, maximum number of iterations, crossover probability, and mutation probability;
[0082] Step 2: Using the discount incentive factor and electric vehicle charging power for each time period as decision variables, chromosome coding is performed using real number coding, and an initial population group p is randomly generated;
[0083] Step 3: Construct a response model based on the user's economic and social dual attribute information. For each group of decision variables in the population, the incentive electricity price is calculated and the user's social behavior habits are considered to obtain the user response rate and determine whether the user participates in the response.
[0084] Step 4: Based on user response, update the electric vehicle charging time plan in conjunction with the initial charging duration, and calculate the dual-objective fitness value of each individual in the population, namely, the aggregator's operating revenue and user satisfaction.
[0085] Step 5: Perform a fast non-dominated sort on the population and calculate the crowding distance between individuals. Generate a new generation of population through selection, crossover, and mutation operations.
[0086] Step 6: Repeat steps 3-5 until the maximum number of iterations K is reached, and output the Pareto front solution set;
[0087] Step 7: Use the LINMAP method to select the comprehensive optimal solution from the Pareto front solution set, and output the corresponding optimal electricity price discount incentive scheme and EV charging plan.
[0088] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: when the processor executes the program, it implements the method for optimizing the participation of electric vehicles in photovoltaic power consumption, taking into account user response uncertainty, as described in any one of claims 1 to 7.
[0089] A computer-readable storage medium storing computer instructions that, when executed by a processor, implement the aforementioned method for optimizing the participation of electric vehicles in photovoltaic power consumption, taking into account user response uncertainty.
[0090] Compared with the prior art, the advantages of the present invention are as follows:
[0091] By adopting the above technical solution, the electric vehicle participation in photovoltaic power consumption optimization method considering user response uncertainty provided by this invention has the following advantages compared with the prior art: Taking photovoltaic and electric vehicles as typical resources, it combines the spatiotemporal matching characteristics of photovoltaic output and charging demand, considers the uncertainty of users' economic and social attributes, and constructs a time-varying incentive intensity model based on price discounts. Introducing the influence of the "herding effect," it proposes a user response rate model that considers the influence of economic incentives and social attributes, characterizing the uncertainty of user response behavior under bounded rationality, and uses the Logistic function to represent the nonlinear mapping relationship between user response rate and charging extension time. Focusing on aggregator operation and user charging experience, it establishes an electric vehicle charging optimization model with the goal of maximizing aggregator revenue and optimizing user satisfaction, and uses the NSGA-II multi-objective optimization algorithm to solve it, effectively improving the distributed photovoltaic power consumption level and achieving synergistic optimization of aggregator operating revenue and user charging satisfaction. Specifically, this is manifested in:
[0092] First, improve the accuracy of user response behavior modeling: integrate economic incentives and user social attributes to build an electric vehicle user response rate model that combines economic incentives and social attributes, breaking through the limitations of traditional single economic factor analysis and accurately quantifying user response behavior.
[0093] Second, improve the feasibility of charging optimization strategies: Construct an electric vehicle charging time extension model based on the Logistic function to characterize the nonlinear mapping relationship between user response rate and charging time adjustment magnitude, and depict the grid connection status and schedulable time window of electric vehicles, effectively reducing the traditional linear adjustment strategy from exceeding the user's acceptable willingness.
[0094] Third, it achieves a synergistic balance of interests among multiple stakeholders: focusing on the operation of aggregators and the user charging experience, an EV charging optimization model is constructed with the goal of maximizing aggregator revenue and optimizing user satisfaction. The model is solved based on the NSGA-II multi-objective optimization algorithm, effectively resolving the nonlinear conflict between aggregator revenue and user satisfaction.
[0095] Fourth, improve the level of distributed photovoltaic power consumption: take the discount incentive factors of each time period as decision variables, combine the differentiated response characteristics of users, and combine the spatiotemporal matching characteristics of photovoltaic power output and charging demand to optimize incentive pricing and charging plans. Through time-varying discount incentive factors, customize differentiated economic incentive schemes to effectively guide electric vehicle users' charging behavior to shift to the photovoltaic surplus period. Attached Figure Description
[0096] Figure 1 This is a flowchart of the method of the present invention.
[0097] Figure 2 This is a graph showing the relationship between photovoltaic power output and base load.
[0098] Figure 3 This is a timeline showing the arrival times of electric vehicles at charging stations.
[0099] Figure 4 This is the solution set diagram of the Pareto front obtained by solving the problem.
[0100] Figure 5 It is a user participation response graph.
[0101] Figure 6 This is a graph showing the user's extended time.
[0102] Figure 7 This is a load transfer diagram for electric vehicles. Detailed Implementation
[0103] The following detailed description, in conjunction with specific embodiments, provides further details.
[0104] Example: An optimization method for electric vehicle participation in photovoltaic power consumption considering user response uncertainty, comprising the following steps:
[0105] S1: Taking a certain photovoltaic power station area as an example, 100 electric vehicles are randomly selected from that area. The simulation period covers 24 hours a day, with a time step of 15 minutes. The vehicle battery capacity is [50, 80] kWh, the maximum charging power is 7 kW, and the charging efficiency is 95%. The EV grid-connected SOC is 15%-25%, and the expected off-grid SOC is 0.85. Photovoltaic output and base load data are as follows: Figure 2 As shown in Table 1, the model correlation coefficient, response threshold, and other parameter settings are shown in Table 2. The price incentive coefficient for photovoltaic demand is 1.3. Electricity price data is shown in Table 2. The time for electric vehicles to arrive at the charging station is shown in Table 3. Figure 3 As shown, its probability distribution is .
[0106] Table 1 Parameter Settings
[0107]
[0108] Table 2 Time-of-use electricity prices (Unit: Yuan / (kWh))
[0109]
[0110] Based on electricity price data, the economic response rate model based on incentive intensity is as follows:
[0111] The incentive price intensity model is as follows:
[0112]
[0113] In the formula, The period during which incentivized prices are offered to aggregators. for Time-based incentive pricing, for Electricity price for charging during certain time periods; The discount incentive factor to promote photovoltaic power grid integration can be adjusted by the aggregator based on its own revenue and grid integration situation. This is the difference in electricity prices. Price incentive coefficient to absorb photovoltaic demand.
[0114] To quantify the impact of incentive intensity on user behavior, an EV user economic response intention model is established:
[0115]
[0116] in,
[0117]
[0118] In the formula, The slope of the user responsiveness line. For saturation excitation intensity, For the deviation values of response characteristics of different users, The slope of user responsiveness during initial incentive compensation. To incentivize and compensate for the user response rate slope after it has a strong guiding effect. The intensity of incentives has a strong guiding effect.
[0119] Furthermore, an economic response rate model based on incentive intensity is established:
[0120]
[0121] In the formula, For user economic response rate, This represents the upper limit of user economic responsiveness. The lower limit of user responsiveness. This is the response threshold.
[0122] S2. Integrate economic incentives and user social attributes to establish a user response rate model and quantify the uncertainty of user response behavior.
[0123] During the process of user participation in demand response, a "herd effect" phenomenon occurs, where users observe others and learn by imitation. The social response rate after being influenced by this phenomenon is as follows:
[0124]
[0125] In the formula, For users within the group with varying degrees of willingness, This refers to the social attributes of this type of user. This represents the weight of herd mentality, indicating how users are influenced by the "herd effect".
[0126] Considering both social and economic impacts, a user response rate model is established:
[0127]
[0128] In the formula, The social attribute weighting coefficient. This represents the weighting coefficient for economic attributes.
[0129] Based on the user response rate model, user response data is obtained, thereby quantifying the uncertainty of user response behavior:
[0130]
[0131] In the formula, This indicates that the user is not participating in the response. This indicates that the user participated in the response.
[0132] S3. Based on user response behavior, establish a Logistic function model for extending electric vehicle charging time and determine the nonlinear mapping relationship between user response rate and charging time.
[0133] The electric vehicle charging time extension model is as follows:
[0134]
[0135] In the formula, Actual charging time for users Base charging time for users. To maximize the extendable time, To extend the time response coefficient, Adjustment coefficients for user response rate and extension time.
[0136] To clarify the time window of the charging behavior model within the scheduling cycle, state variables were introduced. Describe the grid connection status of electric vehicles:
[0137]
[0138] In the formula, Indicates the first individual users Time period status, Indicates EV grid connection, Indicates EV disconnection from the network. For the first Users arrive at the charging station during their designated time period. For the first Each user plans to leave the charging station during their designated time period.
[0139] S4. Establish a multi-objective optimization model with the goals of maximizing the aggregator's operating revenue and optimizing user satisfaction. Constraints include electric vehicle charging power constraints, state of charge constraints, and off-grid target power constraints. Use the NSGA-II algorithm to solve the electric vehicle charging optimization decision and differentiated incentive pricing strategy.
[0140] The multi-objective optimization model is specifically designed to achieve objective 1: aggregator operating revenue and objective 2: user satisfaction.
[0141] Objective 1: Maximize revenue for aggregators, including charging revenue. Reduce waste of solar power revenue Grid-side electricity purchase cost Three parts:
[0142]
[0143] Charging revenue function:
[0144]
[0145] In the formula, For the first individual users EV charging power during specific time periods This is the simulated step size.
[0146] Reduce wastelight revenue function:
[0147]
[0148] In the formula, for Time period of light abandonment and power output To absorb the additional revenue generated by each unit of renewable energy. for Solar power output during specific time periods for Time period base load
[0149] Grid-side electricity purchase cost function:
[0150]
[0151] In the formula, for The power purchased by time-period aggregators from the grid side.
[0152] Objective 2: Optimal user satisfaction, including time satisfaction and cost satisfaction.
[0153]
[0154] Time-based satisfaction function:
[0155]
[0156] In the formula, The time sensitivity coefficient represents the user's psychological perception of charging delay.
[0157] Cost satisfaction function:
[0158]
[0159] In the formula, This represents the economic sensitivity coefficient, indicating how well users perceive the reduction in charging costs. The psychological expectation coefficient for economic savings.
[0160] The constraints of the multi-objective optimization model, which aims to maximize the aggregator's operating revenue and optimize user satisfaction, include: electric vehicle behavior model constraints, electric vehicle charge constraints, electric vehicle off-grid target state of charge constraints, electric vehicle charging power constraints, grid power purchase constraints, and power balance constraints.
[0161] Electric vehicle behavior model constraints:
[0162] First, it is influenced by users' willingness to participate; second, it needs to consider whether users' travel needs are met. If users' willingness to participate is met, but the planned parking time is insufficient to charge to the expected level, users will abandon the response; otherwise, the system has the capability to optimize charging strategies. The expected basic charging time required for electric vehicles is as follows:
[0163]
[0164] In the formula, For the first Target state of charge of individual user electric vehicles For the first The state of charge of each user when joining the network For the first Individual user electric vehicle battery capacity, Maximum charging power for electric vehicles To improve the charging efficiency of electric vehicles.
[0165] Electric vehicle charge constraint:
[0166]
[0167] In the formula, For the first individual users The state of charge of electric vehicles during the time period For the first individual users The state of charge of electric vehicles during a given time period.
[0168] Off-grid target SOC constraint for electric vehicles:
[0169]
[0170] Electric vehicle charging power constraints:
[0171]
[0172] Power grid purchase constraints:
[0173]
[0174] In the formula, It is the maximum power that the aggregator can purchase from the grid.
[0175] Power balance constraints:
[0176]
[0177] In the formula, This represents the total number of electric vehicles.
[0178] The NSGA-II algorithm is used to solve the multi-objective optimization model, yielding the optimal incentive electricity price scheme and electric vehicle charging strategy, specifically:
[0179] Step 1: Set the algorithm parameters and basic data, including population size 100, maximum number of iterations 200, crossover probability 0.6, and mutation probability 0.1;
[0180] Step 2: Using the discount incentive factor and electric vehicle charging power for each time period as decision variables, chromosome encoding is performed using real number encoding, and 100 initial population groups are randomly generated.
[0181] Step 3: Construct a response model based on the user's economic and social dual attribute information. For each group of decision variables in the population, the incentive electricity price is calculated and the user's social behavior habits are considered to obtain the user response rate and determine whether the user participates in the response.
[0182] Step 4: Based on user response, update the electric vehicle charging time plan in conjunction with the initial charging duration, and calculate the dual-objective fitness value of each individual in the population, namely, the aggregator's operating revenue and user satisfaction.
[0183] Step 5: Perform a fast non-dominated sort on the population and calculate the crowding distance between individuals. Generate a new generation of population through selection, crossover, and mutation operations.
[0184] Step 6: Repeat steps 3-5 until the maximum number of iterations (200) is reached, and output the Pareto front solution set;
[0185] Step 7: Use the LINMAP method to select the comprehensive optimal solution from the Pareto front solution set, and output the corresponding optimal electricity price discount incentive scheme and EV charging plan.
[0186] The solution set of the Pareto front obtained by solving is as follows Figure 4 As shown.
[0187] The LINMAP method is used to consider the weights and preferences of multiple objective functions. After eliminating the dimensional differences between the two objectives through Min-Max normalization, the solution with the best overall performance is selected from the Pareto front, corresponding to a profit of 3926.61 yuan and a satisfaction rate of 0.793. Compared with the optimal solution prioritizing satisfaction, the compromise solution improves the profit by 1.05% while sacrificing only about 0.63% of satisfaction; compared with the optimal solution prioritizing profit, the compromise solution improves satisfaction by 4.64% while losing only 2.5% of profit, achieving a good balance between aggregator profit and user satisfaction.
[0188] The optimal solution obtained is shown in Table 3. The total operating revenue of the aggregator is 3926.61 yuan, and the revenue from photovoltaic power consumption by the aggregator is 2907.63 yuan. The grid purchase cost is 1043.96 yuan, the charging revenue is 2062.93 yuan, and the user satisfaction rate is 0.79.
[0189] Table 3 shows the optimal solution results (unit: yuan).
[0190]
[0191] Under this optimal solution, the user participation response is as follows: Figure 5 As shown, the aggregator exhibits significant clustering characteristics due to the "herd effect." After considering its own interests, the aggregator tends to choose the discount incentive factor of 0.4 during the pricing period in order to attract users to participate in the response and obtain more EV scheduling time.
[0192] Electric vehicle charging behavior optimization, such as Figure 6 As shown, the time-varying demand incentive strategy employed allows for more flexible discount incentive options from the aggregator, resulting in a total extended charging time of 424 time slots for users, with an average extension of 7.85 time slots per user, approximately 1.96 hours. Electric vehicle charging load optimization is as follows: Figure 7 As shown in the figure, compared to the disordered electric vehicle load results shown in blue, the electric vehicle charging load achieved varying degrees of time shifting. Specifically, during peak solar PV periods, charging was performed at maximum power to maximize solar PV utilization; while during other periods, a more economical charging ratio was chosen, considering both charging demand and electricity prices. Therefore, an additional 479.25 kWh of solar PV was utilized, increasing the solar PV utilization rate by 13.04%. Electric vehicle users saved 337.97 yuan in charging costs by charging at a lower price, averaging 6.26 yuan per user.
[0193] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A method for optimizing the participation of electric vehicles in photovoltaic power consumption, taking into account user response uncertainty, characterized in that, Includes the following steps: S1. Input regional distributed photovoltaic power output data, electricity price parameters, and basic resource data to construct an economic response rate model based on incentive intensity, quantifying the differentiated economic attributes of users. S2. Integrate economic incentives and user social attributes to establish a user response rate model and quantify the uncertainty of user response behavior. S3. Based on user response behavior, establish a Logistic function-based model for extending electric vehicle charging time, and determine the nonlinear mapping relationship between user response rate and charging time. S4. Establish a multi-objective optimization model with the goals of maximizing the aggregator's operating revenue and optimizing user satisfaction. The constraints include electric vehicle charging power constraints, state of charge constraints, and off-grid target power constraints. The NSGA-II algorithm is used to solve the electric vehicle charging optimization decision and differentiated incentive pricing strategy.
2. The method for optimizing electric vehicle participation in photovoltaic power consumption considering user response uncertainty according to claim 1, characterized in that, In S1, the economic response rate model based on incentive intensity is as follows: The incentive price intensity model is as follows: In the formula, For time-varying excitation intensity, Electricity price intensity factor, The period during which incentivized prices are offered to aggregators. for Time-based incentive pricing, for Electricity price for charging during certain time periods; The discount incentive factor to promote photovoltaic power grid integration can be adjusted by the aggregator based on its own revenue and grid integration situation. This is the difference in electricity prices. To absorb the price incentive coefficient for photovoltaic demand, To quantify the impact of incentive intensity on user behavior, a model of electric vehicle users' willingness to respond economically is established: in, In the formula, For users Economic responsiveness For users Time-varying stimulus received, The slope of the user responsiveness line. For saturation excitation intensity, For the deviation values of response characteristics of different users, The slope of user responsiveness during initial incentive compensation. To incentivize and compensate for the user response rate slope after it has a strong guiding effect. The intensity of the incentive has a strong guiding effect. Establish an economic response rate model based on incentive intensity: In the formula, For user economic response rate, This represents the upper limit of user economic responsiveness. The lower limit of user responsiveness, This is the response threshold.
3. The method for optimizing electric vehicle participation in photovoltaic power consumption considering user response uncertainty according to claim 1, characterized in that, The user response rate model is established in S2 as follows: During the process of user participation in demand response, there is a "herd effect" phenomenon where users observe others and learn by imitation. The social response rate after users are affected: In the formula, For the first Social responsiveness of individual users For users within the group with varying degrees of willingness, For the first Unaffected social attributes of individual users For the first The social attributes of users with high or low willingness within a group. This represents the weighting of herd mentality, indicating how much users are influenced by the "herd effect." Considering both social and economic impacts, establish a user response rate. Model: In the formula, The social attribute weighting coefficient. For the first Economic response rate per user This represents the weighting coefficient for economic attributes. Based on the user response rate model, user response data is obtained, thereby quantifying the uncertainty of user response behavior: In the formula, For the first Individual user response intentions This indicates that the user is not participating in the response. This indicates that the user participated in the response. This represents the average response rate for users.
4. The method for optimizing electric vehicle participation in photovoltaic power consumption considering user response uncertainty according to claim 1, characterized in that, In S3, the electric vehicle charging time extension model is as follows: In the formula, For the first Actual charging time for each user For the first Basic charging time for each user To maximize the extendable time, To extend the time response coefficient, Adjustment factor for user response rate and extension time. To clarify the time window of the above behavioral model within the scheduling cycle, state variables are introduced. Describe the grid connection status of electric vehicles: In the formula, Indicates the first individual users Time period status, Indicates EV grid connection, Indicates EV disconnection from the network. For the first Users arrive at the charging station during their designated time period. For the first Each user plans to leave the charging station during their designated time period.
5. The method for optimizing electric vehicle participation in photovoltaic power consumption considering user response uncertainty according to claim 1, characterized in that, In step S4, a multi-objective optimization model is established with the goals of maximizing the aggregator's operational revenue and optimizing user satisfaction. Specifically: Objective 1: Maximize revenue for aggregators, including charging revenue. Reduce waste of solar power revenue Grid-side electricity purchase cost Three parts In the formula, The objective function is the aggregator's operating revenue. Charging revenue function: In the formula, For the first individual users EV charging power during specific time periods Total time period The total number of electric vehicle users, This is the simulated step size. Reduce wastelight revenue function: In the formula, for Time-of-use light output To absorb the additional revenue generated by each unit of renewable energy. for Solar power output during specific time periods for Time period base load, for Solar power output is absorbed during specific time periods. Grid-side electricity purchase cost function: In the formula, for The power purchased by time-period aggregators from the grid side. Objective 2: Optimal user satisfaction, including time satisfaction and cost satisfaction. In the formula, The objective function is user satisfaction. For user time satisfaction, User satisfaction with charging costs. Time-based satisfaction function: In the formula, The time sensitivity coefficient represents the user's psychological perception of charging delay. For the first Actual charging time for each user For the first Basic charging time for each user The maximum extendable time has already been explained above. Charging cost satisfaction function: In the formula, This represents the economic sensitivity coefficient, indicating how well users perceive the reduction in charging costs. The psychological expectation coefficient for economic savings.
6. The method for optimizing electric vehicle participation in photovoltaic power consumption considering user response uncertainty according to claim 5, characterized in that, The constraints of the multi-objective optimization model in S4, which aims to maximize the aggregator's operating revenue and optimize user satisfaction, include: electric vehicle behavior model constraints, electric vehicle charge constraints, electric vehicle off-grid target state of charge constraints, electric vehicle charging power constraints, grid power purchase constraints, and power balance constraints, specifically: Electric vehicle behavior model constraints: First, it is influenced by users' willingness to participate; second, it needs to consider whether users' travel needs are met. If users' willingness to participate is met, but the planned parking time is insufficient to charge to the expected level, users will abandon the response; otherwise, it has the ability to optimize charging strategies. Among these, the expected basic charging time required for electric vehicles is: In the formula, For the first Target state of charge of individual user electric vehicles For the first The state of charge of each user when joining the network. For the first Individual user electric vehicle battery capacity, Maximum charging power for electric vehicles To improve the charging efficiency of electric vehicles. Electric vehicle charge constraint: In the formula, For the first individual users The state of charge of electric vehicles during the time period For the first individual users The state of charge of electric vehicles during a given time period. Off-grid target SOC constraint for electric vehicles: Electric vehicle charging power constraints: Power grid purchase constraints: In the formula, It is the maximum power that the aggregator can purchase from the grid. Power balance constraints: In the formula, This represents the total number of electric vehicle users.
7. The method for optimizing electric vehicle participation in photovoltaic power consumption considering user response uncertainty according to claim 1, characterized in that, The NSGA-II algorithm is used to solve the multi-objective optimization model, yielding the optimal incentive electricity price scheme and electric vehicle charging strategy, specifically: Step 1: Set algorithm parameters and basic data, including population size, maximum number of iterations, crossover probability, and mutation probability; Step 2: Using the discount incentive factor and electric vehicle charging power for each time period as decision variables, chromosome coding is performed using real number coding, and an initial population group p is randomly generated; Step 3: Construct a response model based on the user's economic and social dual attribute information. For each group of decision variables in the population, the incentive electricity price is calculated and the user's social behavior habits are considered to obtain the user response rate and determine whether the user participates in the response. Step 4: Based on user response, update the electric vehicle charging time plan in conjunction with the initial charging duration, and calculate the dual-objective fitness value of each individual in the population, namely, the aggregator's operating revenue and user satisfaction. Step 5: Perform a fast non-dominated sort on the population and calculate the crowding distance between individuals. Generate a new generation of population through selection, crossover, and mutation operations. Step 6: Repeat steps 3-5 until the maximum number of iterations K is reached, and output the Pareto front solution set; Step 7: Use the LINMAP method to select the optimal solution from the Pareto front solution set, and output the corresponding optimal electricity price discount incentive scheme and EV charging plan.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements the method for optimizing the participation of electric vehicles in photovoltaic power consumption, taking into account user response uncertainty, as described in any one of claims 1 to 7.
9. A computer-readable storage medium storing computer instructions thereon, characterized in that: When executed by the processor, the computer instructions implement the method for optimizing the participation of electric vehicles in photovoltaic power consumption, as described in any one of claims 1-7, taking into account user response uncertainty.