Charging station regulation and control method and system considering ordered charging and demand response

Abstract

The invention relates to the technical field of charging station regulation and control, and discloses a charging station regulation and control method and system considering ordered charging and demand response. The method comprises the steps of processing uncertainty of vehicle arrival time and user departure time through a demand prediction model, constructing a scheduling optimization model to maximize charging station income, and comprehensively considering charging income, V2G discharge income, electricity purchase cost and user waiting penalty. After a power grid demand response signal is received, two-stage decoupling optimization is carried out, the first stage is to optimize a charging pile shutdown strategy, and the second stage is to optimize a vehicle scheduling scheme. The time uncertainty is adaptively adjusted through an affine decision rule, a mixed integer linear programming solver is utilized to obtain an optimal regulation and control instruction of the charging station, and robust scheduling execution is realized. According to the method, the problem that the charging station cannot realize multi-target collaborative optimization scheduling under double challenges of demand prediction uncertainty and power grid demand response is solved, and the robustness and the economic benefit of a charging station scheduling scheme are improved.

Description

Technical Field

[0001] This application relates to the field of charging station control technology, and in particular to a charging station control method and system that takes into account orderly charging and demand response. Background Technology

[0002] With the rapid growth of electric vehicle ownership, charging stations, as a key infrastructure for electric vehicle energy replenishment, have seen their intelligent scheduling and management technologies become crucial support for industry development. Existing charging station scheduling technologies primarily rely on discrete-time models for static optimization, allocating services to arriving vehicles through pre-set charging strategies. Some advanced systems can consider peak-valley electricity price differences for cost optimization and possess basic load balancing capabilities. Simultaneously, with the promotion of grid-side demand response mechanisms, some charging stations are beginning to participate in grid peak-shaving services, responding to grid dispatch instructions by temporarily adjusting charging power. Furthermore, the rise of V2G technology has provided charging stations with a new model for bidirectional energy exchange.

[0003] Existing technologies have revealed several shortcomings in practical applications: First, traditional scheduling methods typically optimize only a single objective, making it difficult to balance electricity purchase costs, charging revenue, user waiting experience, and grid constraints, resulting in poor overall system efficiency. Second, existing prediction models often employ simple historical averaging or linear regression methods, failing to accurately capture the time-varying characteristics and randomness of vehicle arrivals, thus limiting prediction accuracy. Third, users' expected departure times often deviate significantly, and traditional scheduling strategies do not systematically model this uncertainty, easily leading to scheduling scheme failure when users leave early or late. Finally, in demand response scenarios, existing systems lack mechanisms to protect the service continuity of already charged vehicles; simple power reduction may interrupt ongoing charging services, impacting user experience. Summary of the Invention

[0004] This application provides a charging station control method and system that considers orderly charging and demand response, which is used to solve the problem that charging stations cannot achieve multi-objective collaborative optimization scheduling under the dual challenges of demand forecast uncertainty and grid demand response, thereby improving the robustness and economic efficiency of the charging station scheduling scheme.

[0005] Firstly, this application provides a charging station control method that considers orderly charging and demand response, the charging station control method considering orderly charging and demand response including: Feature identification and pattern analysis were performed on the original charging station operation data. A vehicle arrival demand prediction model was established using a non-homogeneous Poisson process, and a user departure time uncertainty model was established using a truncated normal distribution, resulting in demand prediction data that includes time uncertainty handling capabilities. Based on the demand forecast data, a new user arrival scheduling optimization model is constructed. Taking the maximization of charging station revenue as the objective function, and comprehensively considering charging revenue, V2G discharge revenue, electricity purchase cost and user waiting penalty, a normal operation mode scheduling scheme is obtained. When a grid demand response signal is received, the normal operation mode scheduling scheme is decoupled and optimized in two stages. In the first stage, the optimal charging pile shutdown strategy is calculated under the premise of protecting the already charged vehicles. In the second stage, the vehicle scheduling scheme is re-optimized under the shutdown constraint to obtain the demand response mode scheduling scheme. The vehicle-charging station matching relationship of already charged vehicles is subject to immutable constraints, and time uncertainty is adaptively adjusted through affine decision rules to obtain a robust scheduling and execution scheme. The robust scheduling execution scheme is input into a mixed-integer linear programming solver for optimization, supporting V2G bidirectional energy flow control and obtaining the optimal control command for the charging station.

[0006] Secondly, this application provides a charging station control system that considers orderly charging and demand response, the charging station control system considering orderly charging and demand response includes: The demand forecasting module is used to perform feature identification and pattern analysis on the raw charging station operation data. It uses a non-homogeneous Poisson process to establish a vehicle arrival demand forecasting model and a truncated normal distribution to establish a user departure time uncertainty model, thus obtaining demand forecasting data that includes time uncertainty processing capabilities. The scheme formulation module is used to construct a new user arrival scheduling optimization model based on the demand forecast data. With the maximization of charging station revenue as the objective function, it comprehensively considers charging revenue, V2G discharge revenue, electricity purchase cost and user waiting penalty to obtain a normal operation mode scheduling scheme. The response adjustment module is used to perform a two-stage decoupling optimization process on the normal operation mode scheduling scheme when a grid demand response signal is received. The first stage calculates the optimal charging pile shutdown strategy under the premise of protecting the already charged vehicles. The second stage re-optimizes the vehicle scheduling scheme under the shutdown constraint to obtain the demand response mode scheduling scheme. The constraint processing module is used to perform immutable constraint processing on the vehicle-charging station matching relationship of charged vehicles, and to adaptively adjust the time uncertainty through affine decision rules to obtain a robust scheduling execution scheme. The instruction generation module is used to input the robust scheduling execution scheme into a mixed integer linear programming solver for optimization, supporting V2G bidirectional energy flow control, and obtaining the optimal control instructions for the charging station.

[0007] Thirdly, a charging station control device that considers orderly charging and demand response is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the charging station control device that considers orderly charging and demand response to execute the above-described charging station control method that considers orderly charging and demand response.

[0008] Fourthly, a computer-readable storage medium is provided, wherein instructions are stored therein, which, when executed on a computer, cause the computer to perform the above-described charging station control method that takes into account orderly charging and demand response.

[0009] The technical solution provided in this application establishes a vehicle arrival demand prediction model using a non-homogeneous Poisson process, which can accurately capture the changes in vehicle arrival patterns at different times, on weekdays, and holidays. Compared with traditional static prediction methods, this significantly improves prediction accuracy and adaptability. Simultaneously, a truncated normal distribution is used to establish a user departure time uncertainty model, quantifying the randomness of user behavior through probability distribution, effectively solving the scheduling failure problem caused by neglecting user behavior uncertainty in traditional scheduling methods. The scheduling optimization model built based on demand prediction data achieves multi-objective collaborative optimization by comprehensively considering charging revenue, V2G discharge revenue, electricity purchase cost, and user waiting penalties, avoiding the local optimum problem caused by single-objective optimization. When a grid demand response signal is received, the two-stage decoupling optimization strategy, through the first stage's decision to protect already charged vehicles from deactivation and the second stage's rescheduling optimization, satisfies both grid peak-shaving needs and ensures user service quality, resolving the contradiction between demand response and user experience.

[0010] In specific application areas of charging station regulation, the introduction of affine decision rules provides adaptive adjustment capabilities for handling time uncertainties. When actual user behavior deviates from expectations, the system can automatically adjust the charging strategy without resolving the entire optimization problem, significantly improving the system's real-time response and robustness. Immutable constraints on the vehicle-charging station matching relationship of already charged vehicles ensure the executability of the scheduling scheme and avoid the disconnect between the theoretical optimal solution and actual operation. The application of a mixed-integer linear programming solver enables complex multi-constraint optimization problems to obtain optimal solutions within a reasonable time. Furthermore, the support for V2G bidirectional energy flow control not only provides charging stations with additional revenue sources but also enhances the grid's regulation capabilities, realizing bidirectional interaction between charging stations and the grid. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1 This is a schematic diagram of an embodiment of a charging station control method that considers orderly charging and demand response in this application. Figure 2 This is a graph illustrating the uncertainty analysis of user departure time in the embodiments of this application; Figure 3 This is a comparison diagram of charging station power scheduling in the embodiments of this application; Figure 4 This is a flowchart of the two-stage optimized scheduling process for charging stations in this application embodiment; Figure 5 This is a schematic diagram of an embodiment of a charging station control system that considers orderly charging and demand response in this application. Figure 6 This is a schematic block diagram of the structure of a charging station control device that considers orderly charging and demand response in an embodiment of the present invention. Detailed Implementation

[0012] This application provides a charging station control method and system that considers orderly charging and demand response. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0013] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the charging station control method considering orderly charging and demand response in this application includes: Step S1: Perform feature identification and pattern analysis on the original charging station operation data, establish a vehicle arrival demand prediction model using a non-homogeneous Poisson process, and establish a user departure time uncertainty model using a truncated normal distribution to obtain demand prediction data that includes time uncertainty handling capabilities.

[0014] Specifically, based on the raw operational data of charging stations, key features of vehicle arrival and user departure behaviors are extracted through feature recognition and pattern analysis. A vehicle arrival demand prediction model is then established using a non-homogeneous Poisson process. The non-homogeneous Poisson process can flexibly handle the changing frequency of vehicle arrivals across different time periods, making it particularly suitable for highly dynamic environments like charging stations, where vehicle arrivals exhibit significant time dependence. This model accurately captures arrival demand at different times, thereby predicting future charging demand. Furthermore, considering the uncertainty of user departure times, a truncated normal distribution model is used to simulate this behavior. User departure times typically exhibit some randomness and may be influenced by various factors, such as charging duration and waiting time, causing irregular fluctuations in departure times. The truncated normal distribution model, by reasonably limiting departure times, effectively avoids outliers that are too long or too short, preserving the randomness and volatility of departure times, thus providing more accurate data for demand forecasting. By combining these two methods, charging stations can generate demand forecast data that includes both dynamic changes in vehicle arrival demand and addresses the uncertainty of user departure times. Such forecast data is more timely and accurate, providing a foundation for scheduling optimization, resource allocation, and the formulation of charging station operation strategies.

[0015] Step S2: Construct a scheduling optimization model for new user arrival based on demand forecast data. With the goal of maximizing charging station revenue, the model comprehensively considers charging revenue, V2G discharge revenue, electricity purchase cost, and user waiting penalty to obtain a normal operation mode scheduling scheme.

[0016] Specifically, based on demand forecast data, a scheduling optimization model was constructed for the arrival of new users, aiming to maximize the overall revenue of charging stations. This model comprehensively considers multiple factors, including charging revenue, V2G discharge revenue, electricity purchase cost, and user waiting penalty. Charging revenue refers to the income a charging station receives when providing electricity services to users. V2G discharge revenue refers to the revenue obtained by discharging electricity from vehicles to the grid (V2G). This revenue not only improves the profitability of charging stations but also helps regulate load during peak grid demand, thereby optimizing energy allocation. Electricity purchase cost refers to the expenditure incurred by charging stations when purchasing electricity from the grid, directly affecting operating costs and therefore must be included in cost calculations. User waiting penalty measures the negative effects of excessively long wait times, including potential user churn and a decline in user experience. Excessive waiting times may lead users to choose other charging stations or abandon charging services altogether. By considering all these factors, the resulting scheduling scheme optimizes resource allocation while ensuring user demand, thereby maximizing charging station revenue and maintaining a good user experience and operational efficiency.

[0017] Step S3: When the grid demand response signal is received, the normal operation mode scheduling scheme is decoupled and optimized in two stages. In the first stage, the optimal charging pile shutdown strategy is calculated under the premise of protecting the charged vehicles. In the second stage, the vehicle scheduling scheme is re-optimized under the shutdown constraint to obtain the demand response mode scheduling scheme.

[0018] Specifically, upon receiving a grid demand response signal, the scheduling scheme for normal operation mode undergoes a two-stage decoupled optimization process. Grid demand response typically refers to the grid requiring users to reduce electricity consumption or adjust their electricity consumption patterns during peak hours or when power supply is tight. The first stage optimization objective is to calculate the optimal charging pile deactivation strategy while protecting already charged vehicles. This stage needs to consider the matching relationship between already charged vehicles and charging piles, and reasonably deactivate some charging piles to reduce the operating load on charging piles and reduce the burden on the grid. The second stage, under the deactivation constraint, re-optimizes the vehicle scheduling scheme to ensure that, under the background of grid demand response, the remaining charging piles can efficiently meet the charging needs of newly arriving vehicles and minimize user waiting time. This stage needs to consider the idle status of charging piles and dynamically adjust charging priorities to ensure that critical or high-demand vehicles can charge in a timely manner. Through this two-stage decoupled optimization process, the operational efficiency of charging stations can be maximized while meeting grid demand response, ensuring both user charging needs and charging station revenue.

[0019] Step S4: Apply immutable constraints to the vehicle-charging station matching relationship of the charged vehicles, and adaptively adjust the time uncertainty through affine decision rules to obtain a robust scheduling execution scheme.

[0020] Specifically, to ensure the vehicle-charging station matching relationship of already charged vehicles remains unchanged, the usage status of charging stations is constrained to guarantee that once charging is complete, the matching relationship between the vehicle and the charging station will not be altered. This constraint ensures that vehicles that have completed charging will not affect the efficiency of charging stations or the charging needs of other users due to scheduling adjustments. To address the time uncertainties during the charging process, such as the uncertainty of charging duration or user departure, an affine decision rule is used for adaptive adjustment. The affine decision rule introduces a real-time feedback mechanism into the model, dynamically adjusting charging time and resource allocation based on actual operating conditions to reduce the impact of uncertainties. This rule can flexibly adjust the scheduling scheme according to changes in demand under different circumstances, improving the robustness of the system and ensuring that charging stations can maintain efficient operation, minimize resource waste, and reduce user waiting time even when facing various external or internal uncertainties.

[0021] Step S5: Input the robust scheduling execution scheme into the mixed integer linear programming solver for optimization, support V2G bidirectional energy flow control, and obtain the optimal control command for the charging station.

[0022] Specifically, the robust scheduling execution scheme is input into a mixed-integer linear programming solver for optimization. The solver models various constraints of the charging station and automatically optimizes the scheduling and resource allocation of charging piles to ensure efficient system operation. During this process, the solver considers multiple factors such as the number of charging piles, charging demand, vehicle arrival and departure times, and V2G bidirectional energy flow. V2G (Vehicle-to-Grid) bidirectional energy flow control allows charging stations to not only obtain power from the grid but also discharge power back to the grid, achieving bidirectional energy exchange. Through V2G technology, charging stations can feed energy from vehicle batteries back to the grid during peak demand periods, reducing grid load, while simultaneously obtaining power from the grid during off-peak periods, optimizing charging costs. The mixed-integer linear programming solver can adjust the activation and deactivation strategies of charging piles according to the specific needs of the charging station, dynamically adjusting the charging process to maximize benefits and achieve optimal energy allocation. In the final stage, the solver outputs the optimal control instructions for the charging station. These instructions include the optimal charging pile usage plan, V2G energy flow arrangement, and interaction strategy with the power grid, ensuring that the charging station can maintain efficient and stable operation under changing grid demand and vehicle charging demand.

[0023] It is understood that the implementing entity of this application can be a charging station control system that considers orderly charging and demand response, or it can be a terminal or a server; the specific implementation is not limited here. This application's embodiments use a server as an example for illustration.

[0024] In one specific embodiment, the process of performing feature identification and pattern analysis on the original charging station operation data, establishing a vehicle arrival demand prediction model using a non-homogeneous Poisson process, and establishing a user departure time uncertainty model using a truncated normal distribution can specifically include the following steps: (1) Perform data cleaning and format standardization on the original charging station operation data, and extract the vehicle arrival time series, user expected departure time series and actual departure time series to obtain a standardized historical dataset; (2) Based on the standardized historical dataset, time pattern mining is performed to establish a non-homogeneous Poisson process model containing basic arrival rate parameters, daily periodicity coefficient parameters, weekday correction factors and holiday correction factors. The parameters are fitted by the maximum likelihood estimation algorithm, and the prediction bias is corrected by combining the ARIMA time series model to obtain the corrected vehicle arrival prediction model. (3) Perform deviation statistical analysis on the expected departure time and actual departure time of users in the standardized historical dataset, establish a truncated normal distribution model and set a 3-times standard deviation cutoff boundary to obtain the uncertainty model of user departure time; (4) Integrate the corrected vehicle arrival prediction model and the user departure time uncertainty model to obtain demand prediction data that includes time uncertainty processing capabilities.

[0025] Specifically, the raw charging station operation data undergoes data cleaning and format standardization to extract vehicle arrival time series, user expected departure time series, and actual departure time series, resulting in a standardized historical dataset. Based on this standardized dataset, time pattern mining is performed to establish a non-homogeneous Poisson process model, incorporating key factors such as the base arrival rate parameter, daily periodicity coefficient parameter, weekday correction factor, and holiday correction factor. Maximum likelihood estimation is used for parameter fitting, which more accurately describes the dynamic characteristics of vehicle arrival times. An ARIMA (Autoregressive Integral Moving Average) time series model is combined to correct prediction biases, further improving the accuracy of arrival predictions, resulting in a corrected vehicle arrival prediction model. Statistical analysis of the deviations between user expected departure times and actual departure times in the standardized dataset is conducted, establishing a truncated normal distribution model with a cutoff boundary of 3 standard deviations to effectively handle the uncertainty of departure times. This model can effectively describe the randomness and uncertainty of user departure behavior, thus providing more accurate time predictions for charging station operations. By integrating the corrected vehicle arrival prediction model with the user departure time uncertainty model, a demand forecast data with time uncertainty handling capabilities was obtained, providing a basis for scheduling optimization and resource allocation.

[0026] Specifically, the system models and predicts the arrival demand of electric vehicles. A non-homogeneous Poisson process is used to describe vehicle arrival patterns and arrival rates. The change with time t is expressed as: in Based on arrival rate, It is the daily periodicity coefficient. For phase shift, and Correction factors for weekdays and holidays are respectively. and This is the corresponding indicator function. Based on historical data, parameter values ​​are determined through maximum likelihood estimation, and short-term forecast bias is corrected by combining the ARIMA time series model.

[0027] To model the uncertainty of user departure time, let user... The stated expected departure time is Actual departure time Follows a truncated normal distribution: in It was estimated by analyzing historical deviation data. Let be the standard deviation. To handle this uncertainty, a robust optimization method is used, introducing a set of uncertainties with a 95% confidence level into the constraints: in, Let i be the set of uncertainties. The actual departure time variable for user i.

[0028] The tone of the new user's arrival is described as follows: Objective function: The optimization model aims to maximize the revenue of charging stations. in, Indicates vehicle At the charging station and time The charging revenue, It is a vehicle At the charging station and time The charging power, It is a vehicle At the charging station and time V2G discharge benefits, It is a vehicle At the charging station and time V2G discharge power, It is a charging station At any moment The unit price of electricity purchased from the grid. It is a charging station At any moment Total electricity purchased, It is the penalty coefficient for users waiting per unit of time. It is a vehicle The waiting time is N, where N is the total number of vehicles, S is the total number of charging stations, T is the total time period, and F is the total revenue of the charging stations.

[0029] Taking a charging station in a city's commercial district as an example, the implementation process of this invention is described in detail. The charging station is equipped with 30 charging piles, including 12 AC piles (rated power 7kW), 10 DC piles (rated power 60kW), and 8 supercharging piles (rated power 120kW, supporting V2G function), with a total installed capacity of 1644kW. The charging station operates using a peak-valley time-of-use pricing strategy, with an average charging service fee of 1.6-2.2 yuan / kWh. At 9:30 AM one morning, system monitoring showed that 18 vehicles were currently charging: 10 using AC piles (70kW power), 5 using DC piles (300kW power), and 3 using supercharging piles (360kW power), with a total power consumption of 730kW. At this time, a new electric vehicle arrives with a battery capacity of 85.4kWh, a current SOC of 20%, and the user expects to charge to 85%, requiring approximately 55.5kWh of charging. The user is expected to leave at 12:00 PM. Based on historical data analysis, the system prediction module predicts that 12-15 more vehicles will arrive between 10:00 and 14:00, of which approximately 40% are for fast charging. The system models the departure time of these newly arrived vehicles as a normal distribution N(12:00, 25min²), meaning the probability of their actual departure time being between 11:10 and 12:50 is 95%. Figure 2 As shown.

[0030] In one specific embodiment, the process of constructing a scheduling optimization model based on demand forecast data when new users arrive may specifically include the following steps: (1) Based on the demand forecast data, design a multi-objective revenue function that includes charging service revenue, V2G discharge revenue, electricity purchase cost and user waiting time penalty. Set binary decision variables for vehicle charging pile allocation, continuous decision variables for charging power, continuous decision variables for V2G discharge power and continuous decision variables for vehicle waiting time to obtain the optimization objective and decision variable system. (2) Establish a constraint system based on the physical limitations of charging piles and the service requirements of vehicles, including charging pile capacity constraints to ensure that each charging pile is allocated to a maximum of 1 vehicle at each time, vehicle allocation constraints to ensure that each vehicle is allocated to a maximum of 1 charging pile at each time, and power limit constraints to limit the charging and discharging power within the rated range of the equipment, thus obtaining the basic constraint set; (3) Establish robust charging demand constraints based on user charging needs, combine user departure time uncertainty model to ensure that the vehicle completes the target charging amount before the actual departure time, establish battery capacity constraints to ensure that the battery charge is kept within a safe range during charging and discharging, and obtain a set of safety constraints. (4) The optimization target and decision variable system, basic constraint set and safety constraint set are integrated and processed to obtain the normal operation mode scheduling scheme.

[0031] Specifically, the physical limitations of charging stations and the service demands of vehicles establish a system of constraints. The charging station capacity constraint ensures that each charging station provides service to at most one vehicle at any given time, expressed as: in, It is a binary decision variable, representing the vehicle. At any moment Whether to be assigned to a charging station Vehicle allocation constraints ensure that each vehicle can be allocated to at most one charging station at any given time, as shown below: Power limiting constraints restrict the charging and discharging power within the rated range of the device, expressed as: in, It is a vehicle At the charging station and time The charging power, It is a vehicle At the charging station and time V2G discharge power, It is a charging station Maximum charging power, It is a charging station The maximum discharge power.

[0032] Based on user charging demand, a robust charging demand constraint was established. Combined with a user departure time uncertainty model, this ensures that vehicles complete the target charging amount before their actual departure time. This constraint ensures that the charging station can adapt to fluctuations in user departure time, and is expressed as: in, It is a vehicle At the charging station and time The charging power, It is a vehicle The target charging amount User The actual departure time Indicates vehicle The start time of charging. Battery capacity constraints ensure that the battery level remains within a safe range during charging and discharging, expressed as: in, and These are the minimum and maximum safe battery capacity, respectively. It is a vehicle At any moment Battery capacity.

[0033] By integrating the optimization objectives and decision variables, the set of basic constraints, and the set of safety constraints, a scheduling scheme for normal operation is obtained. This scheme will dynamically schedule charging based on the actual needs of the charging station, resource constraints, and battery safety requirements, ensuring efficient and economical charging and maximizing the total revenue of the charging station.

[0034] In one specific embodiment, the process of performing a two-stage decoupling optimization of the normal operation mode scheduling scheme may specifically include the following steps: (1) When the grid demand response signal is received, the charging pile occupancy status at the start of the demand response is analyzed and processed, the set of vehicles that have started charging but have not finished charging is identified and the corresponding charging piles are set to an unusable state, and the idle charging piles and the charging piles that are expected to finish charging before the start of the demand response are set to an unusable state, so as to obtain the charging pile availability classification result. (2) Based on the classification results of charging pile availability, a first-stage shutdown decision optimization model is established. The objective function is to minimize the shutdown cost weighted by the shutdown priority weight and the opportunity cost coefficient of the time period. The total rated power of the shutdownable charging piles is constrained to be greater than or equal to the peak demand of the power grid. The shutdown strategy of the charging piles is obtained by solving the problem through a greedy algorithm. (3) Based on the charging pile deactivation strategy, a second-stage rescheduling optimization model is established. Power robustness constraints are set to ensure that user charging demand is met, physical constraints of deactivated charging piles are set to 0 power of deactivated charging piles, and constraints of already charged vehicles are set to keep the vehicle-pile matching relationship unchanged. The rescheduling scheme is obtained by solving the mixed integer linear programming problem. (4) The charging pile deactivation strategy and rescheduling scheme are integrated and processed to obtain the demand response mode scheduling scheme.

[0035] Specifically, upon receiving a grid demand response signal, the system analyzes the charging pile occupancy status at the start of the demand response, identifies the set of vehicles that have started charging but have not yet completed charging, and sets these charging piles to an unusable state to ensure that these vehicles can continue charging without service interruption due to the demand response. Simultaneously, the system sets idle charging piles and those expected to complete charging before the start of the demand response to an unusable state, thus obtaining a charging pile availability classification result, clarifying which charging piles can be deactivated and which need to continue operating. Next, based on the charging pile availability classification result, the system establishes a first-stage deactivation decision optimization model, aiming to determine which charging piles should be deactivated by minimizing the deactivation cost. The objective function of this optimization model can be expressed as: in, and These are the weights of charging revenue and discharging revenue, respectively. It's the charging power. It is the V2G discharge power. The electricity was purchased from the power grid. It is a moment The waiting time and This is a coefficient representing the grid's electricity purchase cost and the user's waiting time penalty. Using this model, the system optimizes the charging pile deactivation strategy and uses a greedy algorithm to determine which charging piles need to be deactivated. Based on the deactivation strategy in the first stage, the system establishes a rescheduling optimization model for the second stage. This ensures that charging demand is met, while setting the power of deactivated charging piles to 0 ensures that deactivated charging piles no longer provide service. The vehicle-to-charging pile matching relationship of already charged vehicles remains unchanged. A rescheduling scheme is obtained by solving a mixed-integer linear programming problem. Integrating the charging pile deactivation strategy and the rescheduling scheme, a demand-response mode scheduling scheme is obtained, ensuring that charging stations can operate efficiently and maximize their overall revenue while guaranteeing user experience.

[0036] In one specific embodiment, the process of adaptively adjusting time uncertainty using affine decision rules may specifically include the following steps: (1) Establish an affine decision rule model, and express the charging power as a linear combination of the product of the basic charging power, the affine decision coefficient and the user's actual departure time deviation. The affine decision coefficient matrix is ​​obtained by training with historical data, and the adaptive adjustment rule is obtained. (2) When it is detected that a user leaves early, the charging power is increased according to the adaptive adjustment rule to ensure that the target charging amount is completed before the actual departure time. At the same time, the local re-optimization process is triggered to release the charging pile resources and obtain the early departure response plan. (3) When a user is detected to be leaving late, the original charging power is maintained or reduced according to the adaptive adjustment rules, and the charging needs are completed in the extra time to obtain a solution for the delayed departure. (4) The adaptive adjustment rules, early departure response plan and delayed departure response plan are integrated and processed to obtain a robust scheduling execution plan.

[0037] Specifically, the system establishes an affine decision rule model, where charging power is represented as a linear combination of the base charging power, the affine decision coefficients, and the deviation from the user's actual departure time. Through training with historical data, the system obtains the affine decision coefficient matrix, thus forming an adaptive adjustment rule. This rule dynamically adjusts the charging power based on the user's actual departure time deviation. When a user is detected leaving early, the system increases the charging power according to the adaptive adjustment rule to ensure the user completes the target charging amount before the actual departure time. The system triggers local re-optimization to release charging pile resources to provide charging services for arriving users, thus obtaining an early departure response plan. Conversely, when a user is detected leaving late, the system maintains the original charging power or decreases it according to the adaptive adjustment rule, utilizing the extra time to complete the charging demand, thus obtaining a late departure response plan. The system integrates the adaptive adjustment rule, the early departure response plan, and the late departure response plan to form a robust scheduling execution scheme, ensuring that charging stations can flexibly adjust charging strategies to improve charging efficiency while meeting users' charging needs when faced with uncertainties in user departure times.

[0038] In one specific embodiment, the process of performing V2G bidirectional energy flow control may specifically include the following steps: (1) Perform V2G function identification processing on electric vehicles, set V2G capability indication function to control the conditions for vehicles to participate in bidirectional energy exchange, add V2G discharge power decision variables and set V2G power constraints in the scheduling optimization model to obtain the V2G function integration model. (2) Discharge scheduling is performed on vehicles with V2G capability during the demand response period. The V2G discharge power is constrained by the maximum discharge power of the charging pile and the V2G capability indication function of the vehicle. This supports the grid peak shaving demand and obtains discharge benefits, thus obtaining the V2G discharge scheduling scheme. (3) Establish battery capacity safety constraints during V2G discharge to ensure that the battery charge does not fall below the safety limit during vehicle discharge. Establish a V2G benefit calculation model and incorporate discharge benefit into the total benefit maximization objective function to obtain the V2G safety benefit model. (4) The V2G functional integration model, V2G discharge scheduling scheme and V2G safety benefit model are integrated and processed to obtain the V2G bidirectional energy flow control scheme.

[0039] Specifically, the process of implementing V2G bidirectional energy flow control involves identifying the V2G function of electric vehicles, setting a V2G capability indicator function, and controlling whether a vehicle meets the conditions for participating in bidirectional energy exchange. For example, if a vehicle's battery capacity meets a certain standard and supports V2G discharge, it is considered to have V2G capability. In the scheduling optimization model, the system adds a decision variable for V2G discharge power and sets V2G power constraints. For example, the discharge power is limited to the maximum discharge power of each charging station, such as 10kW, and constrained in conjunction with the vehicle's own maximum discharge capacity. In this way, the system obtains a V2G function integrated model. During demand response, the system performs discharge scheduling processing on vehicles with V2G capability, ensuring that the V2G discharge power does not exceed the maximum discharge power of the charging station and the vehicle's V2G capability limit. For example, if the maximum discharge power of the charging station is 10kW and the maximum V2G capability of the vehicle is 5kW, then the vehicle's V2G discharge power will be subject to both limitations. Through such scheduling, the system can both support the grid's peak-shaving demand and obtain discharge benefits, thus forming a V2G discharge scheduling scheme. The system also establishes battery capacity safety constraints during V2G discharge, ensuring that the battery level does not fall below a safe minimum during vehicle discharge; for example, the battery level must not fall below 20%. Simultaneously, the system incorporates the discharge revenue calculations into a goal function that maximizes total revenue. For example, if the discharge revenue is $0.1 per kilowatt-hour, then a vehicle discharging 5kW will generate $0.5 in revenue over one hour. The system integrates the V2G functional integration model, the V2G discharge scheduling scheme, and the V2G safety and revenue model to form a complete V2G bidirectional energy flow control scheme. This ensures that charging stations can be efficiently scheduled during demand response periods, achieving economic benefits while meeting grid demand through V2G technology.

[0040] In one specific embodiment, the process of establishing a first-stage decommissioning decision optimization model based on the charging pile availability classification results may specifically include the following steps: (1) Based on the charging pile availability classification results, cost assessment processing is performed on the charging piles that can be deactivated, and the opportunity cost of AC piles, DC piles and supercharging piles are calculated. The deactivation priority weight is dynamically updated according to the historical revenue of the charging piles to obtain deactivation cost assessment data. (2) Establish the decommissioning decision objective function, and perform a weighted summation of the product of the decommissioning priority weight and the binary variable of the decommissioning status of each charging pile and the product of the opportunity cost coefficient of the time period and the variable of the decommissioning time period to obtain the objective function of minimizing the decommissioning cost; (3) Set power shutdown constraints, requiring that the total rated power of the selected shutdown charging piles among all available shutdown charging piles be greater than or equal to the grid peak-shaving demand. Set shutdown status constraints to ensure that the shutdown decision variables are binary variables, and obtain the shutdown decision constraint system. (4) The greedy algorithm is used to solve the objective function of minimizing the decommissioning cost under the decommissioning decision constraint system. The charging piles with low opportunity cost are selected for decommissioning first, and the decommissioning strategy of the charging piles is obtained.

[0041] Specifically, in the first phase of the deactivation decision-making, the system analyzes the charging pile status at 11:30 AM. It is estimated that 16 vehicles are charging, with 8 using AC charging piles, 4 using DC charging piles, and 4 using supercharging piles. Since the charging positions occupied by these vehicles cannot be deactivated, the system only considers idle charging pile resources. The candidate resources for deactivation include 4 idle AC charging piles, 6 idle DC charging piles, and 4 idle supercharging piles. Next, the system calculates the opportunity cost for each type of charging pile, finding that the opportunity cost for AC charging piles is approximately 11.2 yuan / hour, for DC charging piles it is 67 yuan / hour, and for supercharging piles it is 134 yuan / hour. Based on these opportunity costs, the system uses a greedy algorithm to prioritize deactivating charging piles with lower opportunity costs. Through this strategy, the system decides to prioritize deactivating 3 AC charging piles, saving 21kW of power, while simultaneously adjusting the charging power of one DC charging pile from 60kW to 40kW, further saving 20kW. By fine-tuning the power of other occupied charging piles, a total reduction target of 200kW is achieved.

[0042] In the second phase of rescheduling optimization, the system reallocated remaining resources. Since newly arriving vehicles had completed charging before 11:30, the occupied supercharging stations could be released to other users. Simultaneously, the system activated the V2G (Vehicle-to-Grid) function, assigning one V2G-capable vehicle to discharge 20kW into the grid. This not only supported the grid's peak-shaving needs but also generated a revenue of 0.8 yuan / kWh through discharge, further optimizing the charging station's economic efficiency and peak-shaving capacity. The entire process, through flexible adjustments to charging station usage and the rational application of V2G discharge functionality, ensured efficient operation of the charging station during demand response periods, maximizing the satisfaction of both the grid and user needs.

[0043] refer to Figure 3The system visually demonstrates the power changes throughout the process: Under normal circumstances, the charging station maintains a power output of 720-780kW between 11:30 and 12:30. After implementing demand response, the power output drops to 480-580kW, successfully reducing the load by approximately 200kW, a reduction rate of 26%. The charging pile usage status is also adjusted accordingly: the number of AC charging piles decreases from 10 to 7, DC charging piles from 5 to 4, and the number of supercharging piles remains unchanged at 3. To verify robustness, the system tests scenarios of users leaving early / late. When a user leaves early at 11:40 (20 minutes early), the system's affine decision rules automatically adjust: increasing the charging power for the vehicle between 10:30 and 11:30 from 90kW to 110kW, ensuring sufficient charge (82% SOC) is achieved before 11:40. When a user leaves late at 12:20, the system has ample time to complete charging. Even in the extreme case of a 12:50 delay, the 85% SOC target is still met due to the reserved time buffer. During the response period, the charging station received demand response compensation of 1.2 yuan / kWh × 200kW × 1h = 240 yuan. After deducting the outage loss of approximately 60 yuan, the net profit increased by 180 yuan. Simultaneously, service continuity for all charging vehicles was ensured, and user satisfaction surveys showed a rating of 4.7 / 5.0. This embodiment verifies the effectiveness of the system in a real-world operating environment: the two-stage decoupled architecture enables rapid response, robust optimization ensures user experience, V2G integration improves economic efficiency, and the charging pile deactivation strategy meets the needs of significantly reduced demand, demonstrating good engineering practical value.

[0044] In one specific embodiment, the process of establishing the second-stage rescheduling optimization model based on the charging pile deactivation strategy may specifically include the following steps: (1) Based on the charging pile deactivation strategy, the remaining available charging pile resources are reconfigured, and a rescheduling objective function including charging revenue, V2G discharge revenue, electricity purchase cost and waiting penalty is established. The vehicle allocation and power scheduling decision variables within the range of available charging piles are set to obtain the rescheduling optimization framework. (2) Establish robust power constraints to ensure that charging demand is met under the condition of uncertainty of user departure time. Establish physical constraints for deactivated charging piles, set the charging power and V2G discharge power of deactivated charging piles to 0, and obtain demand response-specific constraints. (3) Establish protection constraints for already charged vehicles to ensure that the minimum power demand of already charged vehicles is not interrupted and the vehicle-charging pile matching relationship remains unchanged. Establish remaining resource allocation constraints to ensure that the capacity of available charging piles is not exceeded during the reallocation process, thus obtaining the vehicle protection constraint system. (4) Input the rescheduling optimization framework, demand response-specific constraints and vehicle protection constraints into the mixed integer linear programming solver for solution processing to obtain the rescheduling scheme.

[0045] Specifically, during the second phase of rescheduling optimization, the system reconfigures the remaining available charging pile resources according to the charging pile deactivation policy. For example, assuming that during grid demand response, the available resources at the charging station include 4 idle AC charging piles, 2 idle DC charging piles, and 1 idle supercharging pile, the system allocates these idle charging piles to arriving vehicles according to demand. For instance, if a newly arrived electric vehicle needs 20kWh of charging, the system prioritizes allocating it to an idle AC charging pile and sets the charging power according to the pile's power limit. The system calculates the charging revenue and V2G discharge revenue for different charging piles. For example, the charging revenue of an idle DC charging pile is 0.6 yuan per kilowatt-hour, while the V2G discharge revenue is 1 yuan per kilowatt-hour; the system will prioritize providing discharge services to vehicles with V2G capabilities.

[0046] The system incorporates robust power constraints to ensure charging demand is met even with uncertain user departure times. For example, a vehicle with a charging demand of 15kWh is expected to leave at 14:30, but the actual departure time may be earlier or later. To ensure charging demand is met, the system may increase charging power at 14:00 to ensure the vehicle finishes charging before its expected departure. If the charging station's power is 10kW, the system will increase it to 15kW to ensure the remaining charging is completed before 14:30. The system also establishes protection constraints for already charged vehicles. For instance, a vehicle has already charged 12kWh and is expected to need an additional 8kWh. To avoid interruptions during charging, the system will ensure the vehicle continues using its original charging station without adjusting its charging power, thus maintaining unaffected charging progress. If other vehicles need to be allocated charging stations, the system will rationally schedule charging based on available stations, prioritizing the charging needs of already charged vehicles. The system uses a mixed-integer linear programming solver to obtain the rescheduling scheme.

[0047] In one specific embodiment, the process of inputting a robust scheduling execution scheme into a mixed-integer linear programming solver for optimization may specifically include the following steps: (1) Input the binary allocation variables and continuous power variables in the robust scheduling execution scheme into the mixed integer linear programming solver, and use the branch and bound algorithm to enumerate the binary variables to obtain the feasible solution space; (2) For each branch in the feasible solution space, the simplex method is used to solve the continuous variables using linear programming. The objective function values ​​of each branch are calculated and compared and selected to obtain the current optimal solution. (3) When the problem is large in scale, the column and constraint generation algorithm is used to decompose the original problem into the main problem and sub-problems for iterative solution. The Lagrange relaxation technique is used to relax the complex constraints to obtain an approximate optimal solution. (4) Establish a real-time monitoring and dynamic re-optimization mechanism. When a sudden event such as a user leaving early is detected, local re-optimization is triggered immediately. By dynamically adjusting parameters to balance robustness and economy, the optimal control command for the charging station is obtained.

[0048] Specifically, when performing scheduling optimization based on charging station load reduction targets, normal scheduling optimization is initiated based on the grid demand response signal and the current charging status. Assume the system has a 15-minute time slice and, after preliminary calculations, a charging plan for a vehicle is determined: from 10:00 to 10:30, the vehicle charges at 50kW using a DC charging station, obtaining 25kWh of electricity; from 10:30 to 11:30, the vehicle switches to a supercharging station, continuing to charge at 90kW, obtaining 30kWh of electricity, and is expected to complete charging before 11:30, with a waiting time of 30 minutes. However, at 11:15, the grid issues a demand response command, requiring a load reduction of 200kW during the period from 11:30 to 12:30. At this time, the total load of the charging station is 780kW, requiring a reduction of approximately 26% in electricity consumption. To address this demand, the system immediately initiates a two-stage optimization process.

[0049] In the first phase of the shutdown decision-making process, the charging status at 11:30 AM was analyzed, determining that an estimated 16 vehicles were charging at that time: 8 using AC charging stations, 4 using DC charging stations, and 4 using supercharging stations. Since the charging stations occupied by these vehicles could not be shut down, the system only considered available, idle charging station resources, including 4 idle AC charging stations, 6 idle DC charging stations, and 4 idle supercharging stations. The system calculated the opportunity cost of these charging stations, finding it to be approximately 11.2 yuan / hour for AC charging stations, 67 yuan / hour for DC charging stations, and 134 yuan / hour for supercharging stations. The system applied a greedy algorithm to prioritize shutting down charging stations with lower opportunity costs, deciding to shut down 3 AC charging stations (saving 21kW of power) and simultaneously reducing the charging power of one DC charging station from 60kW to 40kW, saving 20kW. Through fine-tuning of other occupied charging stations, the system successfully achieved a total load reduction target of 200kW.

[0050] In the second phase of rescheduling optimization, the system reallocated remaining resources. Since newly arrived vehicles had completed charging before 11:30, freeing up occupied supercharging station resources, these stations could be reallocated to other users. Simultaneously, the system activated V2G functionality, assigning a V2G-capable vehicle to discharge to the grid at 20kW. This not only supported the grid's peak-shaving needs but also generated a revenue of 0.8 yuan / kWh through V2G discharge. Through this bidirectional scheduling of charging and discharging, the system effectively reduced the load on charging stations, ensured grid demand response requirements, and balanced the economic benefits of charging stations with users' charging needs.

[0051] Each step in this process is optimized using a mixed-integer linear programming solver, ensuring that during peak grid shaving periods, charging stations can dynamically adjust the use of charging piles to minimize costs and maximize benefits. Through this two-stage optimization process, the system successfully addresses the grid's demand response requirements, ensuring efficient operation of charging stations while effectively balancing economy and robustness. (Reference) Figure 4 The diagram illustrates the two-stage optimized scheduling process for charging stations.

[0052] The charging station control method considering orderly charging and demand response in the embodiments of this application has been described above. The charging station control system considering orderly charging and demand response in the embodiments of this application is described below. Please refer to [link / reference]. Figure 5 One embodiment of the charging station control system considering orderly charging and demand response in this application includes: The demand forecasting module is used to perform feature identification and pattern analysis on the raw charging station operation data. It uses a non-homogeneous Poisson process to establish a vehicle arrival demand forecasting model and a truncated normal distribution to establish a user departure time uncertainty model, thus obtaining demand forecasting data that includes time uncertainty processing capabilities. The scheme formulation module is used to build a scheduling optimization model for new user arrival based on demand forecast data. With the goal of maximizing charging station revenue, it comprehensively considers charging revenue, V2G discharge revenue, electricity purchase cost and user waiting penalty to obtain a normal operation mode scheduling scheme. The response adjustment module is used to perform two-stage decoupled optimization processing on the normal operation mode scheduling scheme when a grid demand response signal is received. The first stage calculates the optimal charging pile deactivation strategy under the premise of protecting the already charged vehicles. The second stage re-optimizes the vehicle scheduling scheme under the deactivation constraint to obtain the demand response mode scheduling scheme. The constraint processing module is used to perform immutable constraint processing on the vehicle-charging station matching relationship of charged vehicles, and to adaptively adjust the time uncertainty through affine decision rules to obtain a robust scheduling execution scheme. The instruction generation module is used to input the robust scheduling execution scheme into the mixed integer linear programming solver for optimization, supporting V2G bidirectional energy flow control and obtaining the optimal control instructions for the charging station.

[0053] Through the collaborative efforts of the aforementioned components, this system achieves precise optimization and dynamic adjustment of charging station scheduling. The demand forecasting module provides predicted data on vehicle arrival and user departure times based on non-homogeneous Poisson processes and truncated normal distributions, ensuring the scheduling model fully considers time uncertainties. The scheme formulation module generates optimized scheduling schemes by maximizing charging station revenue and considering factors such as charging and V2G discharge revenue. The response adjustment module performs multi-stage decoupling optimization under the grid demand response signal, ensuring a balance between grid demand and charging station operation. The constraint handling module ensures the immutability of charging pile matching relationships and flexibly addresses uncertainties through affine decision rules. The instruction generation module transforms the optimized scheme into executable control instructions, achieving optimal operation of charging stations and bidirectional energy flow control, thereby improving overall operational efficiency and revenue.

[0054] above Figure 5 The charging station control system considering orderly charging and demand response in the embodiments of the present invention will be described in detail from the perspective of modular functional entities. The charging station control device considering orderly charging and demand response in the embodiments of the present invention will be described in detail from the perspective of hardware processing.

[0055] Reference Figure 6 This invention also provides a charging station control device that considers orderly charging and demand response. This device can be a server, and its internal structure can be as follows: Figure 6 As shown. The charging station control device considering orderly charging and demand response includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor, designed as a computer, provides computing and control capabilities. The memory of the charging station control device considering orderly charging and demand response includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the charging station control device considering orderly charging and demand response is used to store the data corresponding to this embodiment. The network interface of the charging station control device considering orderly charging and demand response is used for communication with external terminals via network connection. When the computer program is executed by the processor, it implements the above-described method.

[0056] Those skilled in the art will understand that Figure 6The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the charging station control equipment considering orderly charging and demand response applied thereto.

[0057] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the charging station control method that takes into account orderly charging and demand response.

[0058] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0059] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a charging station control device (which may be a personal computer, server, or network device, etc.) that considers orderly charging and demand response to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0060] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A charging station control method considering orderly charging and demand response, characterized in that, include: Feature identification and pattern analysis were performed on the original charging station operation data. A vehicle arrival demand prediction model was established using a non-homogeneous Poisson process, and a user departure time uncertainty model was established using a truncated normal distribution, resulting in demand prediction data that includes time uncertainty handling capabilities. Based on the demand forecast data, a new user arrival scheduling optimization model is constructed. Taking the maximization of charging station revenue as the objective function, and comprehensively considering charging revenue, V2G discharge revenue, electricity purchase cost and user waiting penalty, a normal operation mode scheduling scheme is obtained. When a grid demand response signal is received, the normal operation mode scheduling scheme is decoupled and optimized in two stages. In the first stage, the optimal charging pile shutdown strategy is calculated under the premise of protecting the already charged vehicles. In the second stage, the vehicle scheduling scheme is re-optimized under the shutdown constraint to obtain the demand response mode scheduling scheme. The vehicle-charging station matching relationship of already charged vehicles is subject to immutable constraints, and time uncertainty is adaptively adjusted through affine decision rules to obtain a robust scheduling and execution scheme. The robust scheduling execution scheme is input into a mixed-integer linear programming solver for optimization, supporting V2G bidirectional energy flow control and obtaining the optimal control command for the charging station.

2. The charging station control method considering orderly charging and demand response according to claim 1, characterized in that, The process involves feature identification and pattern analysis of the original charging station operation data, establishing a vehicle arrival demand prediction model using a non-homogeneous Poisson process, and establishing a user departure time uncertainty model using a truncated normal distribution, including: The raw charging station operation data is cleaned and standardized to extract vehicle arrival time series, user expected departure time series and actual departure time series to obtain a standardized historical dataset. Based on the standardized historical dataset, time pattern mining is performed to establish a non-homogeneous Poisson process model that includes basic arrival rate parameters, daily periodicity coefficient parameters, weekday correction factors, and holiday correction factors. The parameters are fitted using the maximum likelihood estimation algorithm, and the prediction bias is corrected by combining the ARIMA time series model to obtain the corrected vehicle arrival prediction model. The deviation statistical analysis was performed on the user's expected departure time and actual departure time in the standardized historical dataset. A truncated normal distribution model was established and a 3-times standard deviation truncation boundary was set to obtain the user departure time uncertainty model. The corrected vehicle arrival prediction model and the user departure time uncertainty model are integrated to obtain demand prediction data that includes time uncertainty processing capabilities.

3. The charging station control method considering orderly charging and demand response according to claim 1, characterized in that, The step of constructing a new user arrival scheduling optimization model based on the demand forecast data includes: Based on the demand forecast data, a multi-objective revenue function is designed, which includes charging service revenue, V2G discharge revenue, electricity purchase cost, and user waiting time penalty. Binary decision variables for vehicle charging pile allocation, continuous decision variables for charging power, continuous decision variables for V2G discharge power, and continuous decision variables for vehicle waiting time are set to obtain the optimization objective and decision variable system. Based on the physical limitations of charging piles and the service needs of vehicles, a constraint system is established, including charging pile capacity constraints to ensure that each charging pile is allocated to a maximum of 1 vehicle at any given time, vehicle allocation constraints to ensure that each vehicle is allocated to a maximum of 1 charging pile at any given time, and power limit constraints to limit the charging and discharging power within the rated range of the equipment, thus obtaining the basic constraint set. Based on user charging needs, robust charging demand constraints are established. Combined with the user departure time uncertainty model, the vehicle completes the target charging amount before the actual departure time. Battery capacity constraints are established to ensure that the battery charge remains within a safe range during charging and discharging, resulting in a set of safety constraints. The optimization objective and decision variable system, the basic constraint set, and the safety constraint set are integrated and processed to obtain the normal operation mode scheduling scheme.

4. The charging station control method considering orderly charging and demand response according to claim 1, characterized in that, The two-stage decoupling optimization process for the normal operation mode scheduling scheme includes: When a grid demand response signal is received, the charging pile occupancy status at the start of the demand response is analyzed and processed. The set of vehicles that have started charging but have not finished charging is identified and the corresponding charging piles are set to an unusable state. Idle charging piles and charging piles that are expected to finish charging before the start of the demand response are set to an unusable state, and the charging pile availability classification results are obtained. Based on the charging pile availability classification results, a first-stage shutdown decision optimization model is established. The objective function is to minimize the shutdown cost weighted by the shutdown priority weight and the time period opportunity cost coefficient. The total rated power of the shutdownable charging piles is constrained to be greater than or equal to the grid peak-shaving demand. The solution is obtained by a greedy algorithm. Based on the charging pile deactivation strategy, a second-stage rescheduling optimization model is established. A power robust constraint is set to ensure that the user's charging demand is met, a deactivated pile physical constraint is set to the power of the deactivated charging pile to 0, and a charged vehicle constraint is set to keep the vehicle-pile matching relationship unchanged. The rescheduling scheme is obtained by solving the problem through mixed integer linear programming. The charging pile deactivation strategy and the rescheduling scheme are integrated to obtain the demand response mode scheduling scheme.

5. The charging station control method considering orderly charging and demand response according to claim 1, characterized in that, The adaptive adjustment of time uncertainty using affine decision rules includes: An affine decision rule model is established, in which the charging power is represented as a linear combination of the base charging power, the affine decision coefficients, and the deviation of the user's actual departure time. The affine decision coefficient matrix is ​​obtained through training with historical data, and the adaptive adjustment rule is derived. When a user is detected leaving early, the charging power is increased according to the adaptive adjustment rules to ensure that the target charging amount is completed before the actual departure time. At the same time, a local re-optimization process is triggered to release charging pile resources, thus obtaining a solution for dealing with early departure. When a user is detected to be leaving late, the original charging power is maintained or reduced according to the adaptive adjustment rules, and the charging demand is completed using the extra time, thus obtaining a solution for dealing with the late departure. The adaptive adjustment rule, the early departure response scheme, and the delayed departure response scheme are integrated to obtain a robust scheduling execution scheme.

6. The charging station control method considering orderly charging and demand response according to claim 1, characterized in that, The support for V2G bidirectional energy flow control includes: V2G function identification processing is performed on electric vehicles, and a V2G capability indication function is set to control the conditions for vehicles to participate in bidirectional energy exchange. A V2G discharge power decision variable is added to the scheduling optimization model and a V2G power constraint is set to obtain a V2G function integration model. During the demand response period, vehicles with V2G capability are scheduled to discharge. The V2G discharge power is constrained by the maximum discharge power of the charging pile and the vehicle's V2G capability indication function. This supports the grid's peak shaving demand and also generates discharge benefits, resulting in a V2G discharge scheduling scheme. Establish a safety constraint on battery capacity during V2G discharge to ensure that the battery charge does not fall below the safety lower limit during vehicle discharge. Establish a V2G revenue calculation model and incorporate the discharge revenue into the total revenue maximization objective function to obtain the V2G safety revenue model. The V2G functional integration model, the V2G discharge scheduling scheme, and the V2G safety and benefit model are integrated to obtain the V2G bidirectional energy flow control scheme.

7. The charging station control method considering orderly charging and demand response according to claim 4, characterized in that, The step of establishing a first-stage decommissioning decision optimization model based on the charging pile availability classification results includes: Based on the charging pile availability classification results, cost assessment processing is performed on the charging piles that can be deactivated, calculating the opportunity cost of AC piles, DC piles and supercharging piles, and dynamically updating the deactivation priority weight according to the historical revenue of the charging piles to obtain deactivation cost assessment data. Establish a decommissioning decision objective function by weighting and summing the product of the decommissioning priority weight and the binary variable of the decommissioning status of each charging pile, and the product of the opportunity cost coefficient of the time period and the variable of the decommissioning time period, to obtain the objective function of minimizing the decommissioning cost. Set a power shutdown constraint condition, requiring that the total rated power of the selected shutdown charging piles among all available shutdown charging piles be greater than or equal to the grid peak-shaving demand. Set a shutdown state constraint to ensure that the shutdown decision variable is a binary variable, thus obtaining the shutdown decision constraint system. A greedy algorithm is used to solve the objective function of minimizing the deactivation cost under the deactivation decision constraint system, and the deactivation of charging piles with low opportunity cost is prioritized to obtain the charging pile deactivation strategy.

8. The charging station control method considering orderly charging and demand response according to claim 4, characterized in that, The second-stage rescheduling optimization model based on the charging pile deactivation strategy includes: Based on the charging pile deactivation strategy, the remaining available charging pile resources are reconfigured. A rescheduling objective function is established, which includes charging revenue, V2G discharge revenue, electricity purchase cost, and waiting penalty. Vehicle allocation and power scheduling decision variables are set within the range of available charging piles to obtain the rescheduling optimization framework. Establish a robust power constraint to ensure that charging demand is met under the condition of uncertainty of user departure time, and establish a physical constraint for deactivated charging piles to set the charging power and V2G discharge power of deactivated charging piles to 0, thereby obtaining demand response-specific constraints. Establish protection constraints for already charged vehicles to ensure that vehicles that have started charging maintain their minimum power requirements without interruption and that the vehicle-charging pile matching relationship remains unchanged. Establish remaining resource allocation constraints to ensure that the reallocation process does not exceed the available charging pile capacity, thus obtaining the vehicle protection constraint system. The rescheduling optimization framework, the demand response-specific constraints, and the vehicle protection constraint system are input into a mixed-integer linear programming solver for solution processing to obtain the rescheduling scheme.

9. The charging station control method considering orderly charging and demand response according to claim 1, characterized in that, The step of inputting the robust scheduling execution scheme into a mixed-integer linear programming solver for optimization includes: The binary allocation variables and continuous power variables in the robust scheduling execution scheme are input into the mixed integer linear programming solver, and the branch and bound algorithm is used to enumerate the binary variables to obtain the feasible solution space. For each branch in the feasible solution space, the simplex method is used to perform linear programming on the continuous variables. The objective function values ​​of each branch are calculated and compared and selected to obtain the current optimal solution. When the problem is large in scale, the column and constraint generation algorithm is used to decompose the original problem into a main problem and sub-problems for iterative solution. The Lagrange relaxation technique is used to relax the complex constraints to obtain an approximate optimal solution. Establish a real-time monitoring and dynamic re-optimization mechanism. When an emergency event such as a user leaving early is detected, local re-optimization is triggered immediately. By dynamically adjusting parameters to balance robustness and economy, the optimal control command for the charging station is obtained.

10. A charging station control system considering orderly charging and demand response, characterized in that, A charging station control method considering orderly charging and demand response as described in any one of claims 1-9, wherein the charging station control system considering orderly charging and demand response comprises: The demand forecasting module is used to perform feature identification and pattern analysis on the raw charging station operation data. It uses a non-homogeneous Poisson process to establish a vehicle arrival demand forecasting model and a truncated normal distribution to establish a user departure time uncertainty model, thus obtaining demand forecasting data that includes time uncertainty processing capabilities. The scheme formulation module is used to construct a new user arrival scheduling optimization model based on the demand forecast data. With the maximization of charging station revenue as the objective function, it comprehensively considers charging revenue, V2G discharge revenue, electricity purchase cost and user waiting penalty to obtain a normal operation mode scheduling scheme. The response adjustment module is used to perform a two-stage decoupling optimization process on the normal operation mode scheduling scheme when a grid demand response signal is received. The first stage calculates the optimal charging pile shutdown strategy under the premise of protecting the already charged vehicles. The second stage re-optimizes the vehicle scheduling scheme under the shutdown constraint to obtain the demand response mode scheduling scheme. The constraint processing module is used to perform immutable constraint processing on the vehicle-charging station matching relationship of charged vehicles, and to adaptively adjust the time uncertainty through affine decision rules to obtain a robust scheduling execution scheme. The instruction generation module is used to input the robust scheduling execution scheme into a mixed integer linear programming solver for optimization, supporting V2G bidirectional energy flow control, and obtaining the optimal control instructions for the charging station.

Images