A multi-virtual power plant-oriented real-time electric energy sharing and robust aggregated scheduling method for electric vehicle charging station groups

Abstract

This invention discloses a real-time power sharing and robust aggregation scheduling method for electric vehicle charging station clusters with multiple virtual power plants. First, a hierarchical collaborative architecture is established, comprising virtual power plant operators, multiple virtual power plants, charging station clusters, and electric vehicle users. Charging and discharging behaviors are determined based on user revenue and evolving responses, and inter-station power sharing clearing is completed in conjunction with in-station generation, load, and shared network constraints. The net sharing results of each charging station are aggregated according to their affiliation to form the electric vehicle aggregation power boundary for the corresponding virtual power plant. Uncertainty intervals for electric vehicle access, arrival / departure, charging / discharging power, and user response are constructed, and robust verification of the aggregation power boundary and virtual power plant power balance is performed. Through hierarchical iterative coordination of shared electricity prices, transaction prices, purchased and sold power, and internal resource output, the accuracy, stability, and executability of aggregation scheduling are improved.

Description

Technical Field

[0001] This invention relates to the field of charging, and specifically to a method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters with multiple virtual power plants. Background Technology

[0002] With the large-scale integration of distributed renewable energy, electric vehicles, energy storage devices, and flexible loads, the power system is gradually exhibiting characteristics of resource decentralization, operational randomization, and multi-entity dispatch. Virtual power plants, through communication, aggregation, and optimized control technologies, can integrate distributed resources such as wind power, photovoltaics, energy storage, interruptible loads, and electric vehicles into a unified dispatchable resource entity, participating in electricity market transactions and grid regulation. Due to differences in resource structure, load characteristics, and renewable energy output timing, multiple virtual power plants possess certain potential for power complementarity and energy sharing.

[0003] Existing technologies have studied the joint dispatching of multiple virtual power plants. CN112308334A discloses a joint optimization dispatching method for multiple virtual power plants based on master-slave cooperative game theory. This method establishes optimization dispatching models for different types of virtual power plants and a master-slave game model between the distribution network operator and multiple virtual power plants, determining the trading strategy through iteration. Based on this, the Shapley value method is used to redistribute cooperative profits to promote energy sharing among multiple virtual power plants. This technology focuses on solving the problems of transaction pricing, cooperation among multiple virtual power plants, and revenue distribution between the distribution network operator and multiple virtual power plants.

[0004] However, CN112308334A primarily models distributed resources at the overall level of virtual power plants. For electric vehicle resources, it doesn't further distinguish the specific charging stations that electric vehicles actually connect to, nor does it specifically depict user charging and discharging behavior under the influence of factors such as shared electricity prices, arrival and departure times, additional driving costs, and battery degradation costs. Therefore, this method cannot directly reflect the actual formation process of electric vehicle resources, from single-vehicle access, charging station aggregation, to virtual power plant aggregation. When different charging stations experience power surpluses and deficits, simply performing total scheduling at the virtual power plant level may not accurately obtain the actual executable electric vehicle charging and discharging power boundaries for each virtual power plant within the corresponding time period.

[0005] For the coordinated scheduling of charging station clusters, existing technologies have also proposed corresponding solutions. CN117424268A discloses a method for scheduling electric vehicle charging stations to achieve regional energy supply and demand balance. This method clusters multiple charging stations based on their location and load correlation to form charging station clusters and regional agents. The power dispatch center issues load dispatch instructions to the regional agents, which iteratively adjust the charging and discharging power of each charging station to enable the charging station cluster to respond to regional load demand and coordinate regional wind power, photovoltaic, and charging station resources to participate in supply and demand balance.

[0006] However, the key technical focus of CN117424268A is to group charging stations according to their correlation and to achieve coordinated charging and discharging control of the charging station group based on the upper-level load instructions. This method does not further establish an energy sharing and clearing mechanism among charging stations based on the real-time power surplus or deficit of each station, nor does it consider the dynamic selection and bounded rationality response of electric vehicle users among multiple charging stations. Its charging station group mainly serves as an overall adjustable load or energy storage resource under a regional agent, and it has not yet converted the station-level scheduling results into the aggregated power boundary of electric vehicles that can be called by each of the multiple virtual power plants according to the affiliation relationship between the charging station and different virtual power plants, nor has it further linked the transaction pricing of virtual power plant operators and the power purchase and sale decisions of multiple virtual power plants.

[0007] Furthermore, the number of electric vehicles connected, arrival and departure times, state of charge, charging and discharging power, and user participation all fluctuate, and the distributed renewable energy output within charging stations is also uncertain. If the adjustable power or shared results calculated at the charging station level are directly used as deterministic values ​​for virtual power plant scheduling, it may lead to inconsistencies between the upper-level scheduling plan and the actual executable power at the lower level, potentially resulting in power deficits, additional power purchases, or readjustment of equipment output. Although existing technologies involve joint scheduling of multiple virtual power plants and coordinated control of charging station clusters, a cross-layer power transfer and deviation verification mechanism for electric vehicle resources is still lacking between the two.

[0008] Therefore, existing technologies still need to address the following issues: First, there is a lack of continuous data transmission between electric vehicle user response, charging station power sharing, and multi-virtual power plant aggregation scheduling; second, the real-time sharing results between charging stations are difficult to accurately convert into the electric vehicle aggregation power boundary that each virtual power plant can actually call upon; third, user behavior fluctuations and charging / discharging deviations have not yet been uniformly verified with the power balance constraints of multi-virtual power plants; and fourth, there is a lack of a closed-loop update mechanism between charging station sharing prices, user response, virtual power plant transaction prices, and internal resource output.

[0009] Based on this, it is necessary to propose a real-time power sharing and robust aggregation scheduling method for electric vehicle charging station clusters with multiple virtual power plants. This method incorporates electric vehicle user response, power sharing among charging stations, multi-virtual power plant transaction scheduling, and uncertainty verification into a unified hierarchical coordination framework to improve the accuracy of electric vehicle aggregation power boundaries, the feasibility of scheduling schemes, and the power balance capability under uncertain operating scenarios. Summary of the Invention

[0010] To address the aforementioned issues, this invention proposes a real-time power sharing and robust aggregation scheduling method for electric vehicle charging station clusters with multiple virtual power plants. This method establishes a hierarchical collaborative structure among virtual power plant operators, multiple virtual power plants, charging station clusters, and electric vehicle users. It generates the charging and discharging power of each charging station based on the dynamic responses of electric vehicle users, and completes power sharing clearing among charging stations by combining in-station generation, load, and shared network constraints. The shared station-level net power is converted into the electric vehicle aggregation power boundary of each virtual power plant according to the affiliation relationship between the charging station and the virtual power plant. Furthermore, considering the uncertainties in electric vehicle response and charging / discharging power, robust verification is performed on the aggregation power boundary and virtual power plant power balance constraints. Through hierarchical iteration, closed-loop coordination between user response, inter-station sharing, aggregation scheduling, and transaction pricing is achieved, thereby improving the consistency between the aggregation power boundary and the actual executable power.

[0011] The technical solution of the method of the present invention is: Step S1: Establish a multi-layer collaborative architecture and acquire operational data; acquire operational data of virtual power plant operators, multiple virtual power plants, charging station clusters and electric vehicle users within the scheduling cycle, and establish a multi-layer collaborative scheduling architecture of virtual power plant operators, multiple virtual power plants, charging station clusters and electric vehicle users.

[0012] Step S2: Establish an electric vehicle user revenue model and a charging station real-time power sharing model; determine the user's revenue from participating in sharing based on the charging and discharging utility, sharing cost, time cost, additional energy consumption cost, and battery degradation cost of electric vehicle users; determine the power supply and demand relationship between charging stations based on the adjustable resource status within each charging station and user response results, and perform real-time power sharing clearing.

[0013] Step S3: Establish a master-slave game model between the virtual power plant operator and multiple virtual power plants; with the goal of maximizing the net profit of the upper-level virtual power plant operator, determine the transaction purchase price and transaction sales price of electricity for each time period; with the goal of minimizing the operating cost of each lower-level virtual power plant, determine the power purchase and sales capacity and internal resource output of each virtual power plant, and summarize the charging station sharing results into the electric vehicle aggregated power input of the corresponding virtual power plant.

[0014] Step S4: Establish an uncertain set of electric vehicle aggregated resources and perform robust reinforcement; for the random fluctuations of electric vehicle access status, charging and discharging power and user response results, construct an uncertain interval and robust coefficient control mechanism, and embed it into the virtual power plant power balance constraint to robustly reinforce the scheduling model.

[0015] Step S5: Jointly solve and output the real-time sharing and robust aggregation scheduling scheme; adopt a hierarchical iterative approach to jointly solve the electric vehicle user evolution response model, the charging station sharing clearing model, the virtual power plant lower-level scheduling model, and the virtual power plant operator upper-level pricing model, and output the charging station sharing price, sharing power, electric vehicle user selection results, power purchased and sold by each virtual power plant, internal resource output, and robust scheduling scheme.

[0016] In step S1, to achieve the hierarchical aggregation of electric vehicle resources from the user side to the virtual power plant side, a multi-layered entity structure with a clear decision-making order needs to be established first. The upper-layer entities are virtual power plant operators and multiple virtual power plants, mainly responsible for transaction pricing and aggregation scheduling; the lower-layer entities are charging station clusters, shared clearing institutions, and electric vehicle users, mainly responsible for shared response and real-time power mutual assistance. The input data for virtual power plant operators includes the on-grid electricity price, the grid purchase price, the electricity purchase and sale demand reported by each virtual power plant, and the results of electricity market transactions; the input data for each virtual power plant includes the upper limit of micro gas turbine output, the state of charge of energy storage, the available output of wind turbines, the interruptible load capacity, and the state of aggregated electric vehicle resources; the input data for the charging station layer includes the station's fixed generating power, adjustable generating power, fixed power consumption, callable power, and shared network constraints; the input data for the electric vehicle user layer includes the state of charge, the upper limit of charging power, the upper limit of discharging power, the estimated departure time, the estimated arrival time, the path distance, the time cost coefficient, and the battery degradation parameters. Through the above data collection and hierarchical division, subsequent electricity price signals, bottom-up sharing results, and robust scheduling results can be transmitted in a closed loop.

[0017] In step S2, a revenue model for electric vehicle users participating in sharing is first established. Electric vehicle users during specific time periods... Choose a charging station The total benefits gained from participating in sharing can be expressed as: ; in, Indicates the first Electric vehicle users during the time period Choose a charging station Total revenue shared by participants; Indicates the cumulative utility of charging; Indicates the cumulative cost of discharge; Indicates the revenue generated from charging and discharging costs; Indicates heading to a charging station Time cost; This indicates the additional energy consumption cost of driving; This represents the cost of battery degradation. This revenue model comprehensively characterizes the impact of shared electricity prices, user travel behavior, and battery lifespan deterioration on users' willingness to participate in shared electricity services.

[0018] In step S2, the charging and discharging fee revenue of electric vehicle users during the sharing process can be expressed as: ; in, Indicates time period charging station Shared electricity pricing; Indicates the first Electric vehicle users during the time period The discharge power; Indicates the first Electric vehicle users during the time period The charging power; This indicates the length of the sharing period. When the discharge power is greater than the charging power, the user can obtain a net benefit through sharing; when the charging power is greater than the discharge power, the user pays the corresponding electricity fee to the charging station.

[0019] In step S2, the electric vehicle user's state of charge is dynamically updated according to the charging and discharging behavior, and the relationship between these changes can be expressed as follows: ; in, Indicates the first Electric vehicle users during the time period Battery charge; Indicates the charge level in the previous time period; Indicates charging efficiency; This represents the discharge efficiency. This relationship is used to ensure that users can still meet their subsequent travel needs and minimum state of charge constraints after participating in sharing.

[0020] In step S2, considering that electric vehicle users do not have complete information and find it difficult to instantly obtain the optimal strategy during actual sharing, an evolutionary game theory approach is used to describe the changes in the probability of users choosing different charging stations. The discretized replicator dynamics can be represented as... ; in, Indicates the first During the next iteration, the user During the period Choose a charging station The probability of; This indicates the probability of selection after the next iteration; Indicates the iteration step size; Indicates user During the period The average expected return. When the return of a certain charging station is higher than the average return, the probability of choosing that station increases; conversely, it decreases, until an evolutionary steady state is reached.

[0021] In step S2, the charging and discharging power of a single electric vehicle user is aggregated at the charging station level to form a shared response aggregation item. This aggregation item can be represented as... ; in, Indicates charging station During the period Electric vehicles share response power; Indicates selection of charging station The set of electric vehicle users. This aggregation item is used to characterize the net demand or net supply of the external shared network by the total number of users within the site. If A positive value indicates that the charging station has a high demand for sharing during that time period; if... A negative value indicates that the charging station has the ability to provide shared power to other charging stations.

[0022] In step S2, after obtaining the shared response aggregation item, the charging station establishes a shared utility model by combining the station's fixed power generation, adjustable power generation, fixed power consumption, available power consumption, and shared electricity price, and reports the shared power demand to the shared clearing mechanism. Under the premise of satisfying the shared network transmission constraints and line transmission cost constraints, the shared clearing mechanism performs clearing and matching of shared electricity between stations, thereby obtaining the net shared power result for each charging station in each time period. This result reflects both the dynamic response on the user side and the operational boundary of the inter-station shared network, and can serve as the input boundary for subsequent electric vehicle aggregation resources at the virtual power plant layer.

[0023] In step S3, a two-layer master-slave game structure is formed between the virtual power plant operator and multiple virtual power plants. The net profit of the virtual power plant operator can be expressed as... ; in, This indicates the net revenue of virtual power plant operators; This indicates the on-grid electricity price in the electricity market; This indicates the electricity purchase price in the electricity market; This represents the amount of electricity sold by virtual power plant operators to the electricity market; This represents the amount of electricity purchased by the virtual power plant operator from the electricity market; This indicates the electricity price offered to virtual power plants. This indicates the electricity purchase price for virtual power plants; Indicates the first The amount of electricity purchased by a virtual power plant from an operator; Indicates the first The amount of electricity sold by a virtual power plant to an operator; Indicates the number of virtual power plants; This indicates the total number of time periods within the scheduling cycle.

[0024] In step S3, to ensure that multiple virtual power plants have the willingness to transact with the virtual power plant operator, the transaction price satisfies... ; in, This indicates the on-grid electricity price in the electricity market. This indicates the electricity purchase price for virtual power plants. This indicates the electricity price sold to virtual power plants. This represents the electricity purchase price in the electricity market. This constraint ensures that the prices set by virtual power plant operators are within the range of electricity market purchase and sale prices, making it economically reasonable for virtual power plants to trade through operators.

[0025] In step S3, each virtual power plant, under a given electricity trading price, coordinates its internal conventional distributed resources with electric vehicle aggregated resources with the goal of minimizing operating costs. The operating cost of a virtual power plant can be expressed as: ; in, Indicates the first Operating costs of a virtual power plant; Indicates the first A collection of resources within a virtual power plant, including micro gas turbines, energy storage systems, and interruptible loads; Indicates belonging to the first A collection of electric vehicles in a virtual power plant; Indicates the cost of a micro gas turbine; Indicates the cost of energy storage dispatch; This indicates the cost of interruptible load compensation; This indicates the cost of charging and discharging electric vehicles.

[0026] In step S3, the cost of the micro gas turbine can be expressed as: ; in, , , The cost coefficient for micro gas turbines is given; the energy storage dispatch cost can be expressed as... ; in, This is the energy storage dispatch cost coefficient. Indicates the time period of the energy storage system The charging and discharging power; the interruptible load compensation cost can be expressed as ; in, To compensate for the interruption of electricity price, This indicates that interruptible loads can reduce power output.

[0027] In step S3, the power balance relationship of the virtual power plant can be expressed as follows: ; in, Indicates the first A virtual power plant during the time period Net trading power with virtual power plant operators; Indicates the output of a micro gas turbine; Indicates the energy storage charging and discharging power; This indicates that interruptible loads can reduce power consumption. Indicates wind power output; This indicates the number of accesses or the daily load corresponding to the aggregated resources for electric vehicles; Indicates the discharge power of the electric vehicle; Indicates the power of a normal load; This represents the charging power of electric vehicles. By aggregating the results of charging station sharing into the aggregation boundary of the virtual power plant layer, user-side behavior and station group sharing results can be directly incorporated into the virtual power plant scheduling process.

[0028] In step S4, to address the fluctuations caused by electric vehicle access time, departure time, charging power, discharging power, and user response behavior, an uncertain range for aggregated electric vehicle resources is constructed. The uncertain variables can be represented as... ; in, Indicates the first Uncertain factors in the first The actual values ​​taken on each aggregate object; This indicates the lower bound of the uncertainty factor; This represents its upper bound. By using this interval constraint, the charging and discharging deviation of electric vehicles can be limited to a calculable fluctuation range.

[0029] In step S4, a robustness coefficient is introduced to avoid overly conservative scheduling results. Adjustments are made for the uncertainty interval. Robustness coefficient. Used to adjust the degree of worst-case scenario consideration, when When the value is small, the model is more biased towards economy; when When the value is large, the model is more stability-oriented. By adjusting the robustness coefficient, the scheduling model can strike a trade-off between operating cost and resistance to fluctuations.

[0030] In step S4, after considering the fluctuations in aggregated resources of electric vehicles, the power balance constraints of the virtual power plant are robustly strengthened. This ensures that the system still meets the power supply and demand balance, energy storage power boundary, wind power utilization boundary, interruptible load boundary, and trading constraints even under unfavorable charging and discharging scenarios. This avoids power gaps, excessive power purchases, or resource output mismatches caused by electric vehicle user response deviations. The robustly strengthened scheduling results also serve as the verification basis for upper-layer transaction pricing and lower-layer shared clearing, thereby improving the executability and stability of the overall scheduling scheme.

[0031] In step S5, a hierarchical iterative approach is used for joint solution. First, given an initial shared electricity price, the electric vehicle user evolution response model is solved to obtain the user shared response power of each charging station. Second, based on the shared response power and resource status of each charging station, the shared clearing model is solved to obtain the inter-station shared power allocation result. Subsequently, the sharing results of multiple charging stations are aggregated according to their affiliation to form the electric vehicle aggregated power boundary of the corresponding virtual power plant, and the lower-level scheduling model of each virtual power plant is solved under a given transaction price. Finally, the virtual power plant operator updates the transaction purchase price and transaction sales price of electricity based on the response results of multiple virtual power plants, and continues to iterate until the price, shared power, and resource output simultaneously meet the convergence criterion.

[0032] In step S5, the convergence criteria are set as follows: the change in the shared electricity price of each charging station in two adjacent iterations is less than a preset threshold, the change in the purchased and sold power of each virtual power plant is less than a preset threshold, and the change in the overall objective function is less than a preset threshold. When the above conditions are met, the system is considered to have reached a stable equilibrium state, and the iteration stops.

[0033] In step S5, the output results include at least: the transaction purchase price and transaction sales price of virtual power plant operators for each time period, the purchase and sales power of each virtual power plant, the output of micro gas turbines, the charging and discharging power of energy storage, the interruptible load reduction power, the wind power utilization power, the electric vehicle aggregated charging and discharging plan, the shared electricity price and shared power of each charging station, and the probability of electric vehicle users choosing each charging station. Based on the above output results, a real-time power sharing and robust aggregation scheduling scheme applicable to multi-virtual power plant scenarios can be formed.

[0034] Compared with the prior art, the beneficial effects of the present invention include: (1) By establishing a hierarchical power transfer structure of “electric vehicle users - charging station groups - multiple virtual power plants - virtual power plant operators”, the electric vehicle resources no longer directly enter the virtual power plant scheduling in a preset total amount, but first form the actual response power at the station level according to the user's willingness to participate, the site selection, the state of charge and the arrival and departure constraints, thereby improving the consistency between the modeling of electric vehicle aggregated resources and the actual access status.

[0035] (2) Based on the user response power, station power generation power, load power and shared network operation constraints of each charging station, the present invention shares and clears the power surplus and power deficit among charging stations, so that charging stations with surplus power supply capacity can provide power to charging stations with insufficient power, improve the utilization efficiency of new energy and electric vehicle energy storage resources within the charging station group, and reduce the additional grid power purchase demand caused by local power imbalance.

[0036] (3) The present invention converts the net power after the charging station is shared into the electric vehicle aggregate power boundary of the corresponding virtual power plant according to the affiliation relationship between the charging station and the virtual power plant, so that the station-level user response and sharing results can directly enter the power balance and internal resource scheduling process of the virtual power plant, thereby reducing the deviation between the upper-level planned scheduling power and the lower-level actual executable power.

[0037] (4) In view of the fluctuations in the number of electric vehicles connected, arrival and departure times, charging and discharging power and user response results, this invention constructs an uncertain set of electric vehicle aggregated resources and robustly strengthens the aggregated power boundary and virtual power plant power balance constraints, so that the scheduling scheme can still meet the equipment operation boundary, power supply and demand balance and transaction constraints under adverse response scenarios, thereby improving the stability and actual executability of the scheduling scheme.

[0038] (5) This invention uses a hierarchical iteration of user response, charging station sharing clearing, multi-virtual power plant internal scheduling and virtual power plant operator transaction pricing, so that the actual sharing results of the lower layer can correct the upper layer aggregation boundary, and the upper layer transaction price and robust verification results can in turn adjust user response, shared power and internal resource output, forming a two-way closed-loop correction mechanism to avoid power mismatch caused by independent solution of each layer model.

[0039] (6) This invention unifies electric vehicle user behavior, charging station physical power sharing, multi-virtual power plant resource scheduling and operator transaction coordination into the same scheduling framework, outputs the shared power of each charging station, the electric vehicle aggregate charging and discharging plan of each virtual power plant and the internal equipment output plan, which can be used for the collaborative operation of the charging station control platform and the virtual power plant energy management system, improving the scheduling flexibility of electric vehicle resources and the system operation reliability in the multi-virtual power plant scenario. Attached Figure Description

[0040] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0041] Figure 1 This is a diagram illustrating the overall architecture of the method of the present invention.

[0042] Figure 2 This is a diagram of the master-slave game structure of the method of the present invention.

[0043] Figure 3 This is a comparison chart of electric vehicle user response results under different shared time period parameters in an embodiment of the present invention.

[0044] Figure 4 This is a comparison chart of the aggregated capacity of electric vehicles before and after robust scheduling in an embodiment of the present invention. Detailed Implementation

[0045] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. Equivalent substitutions or modifications made by those skilled in the art based on the content of this specification without departing from the technical concept of the present invention should all fall within the scope of protection of the present invention.

[0046] Example 1 This implementation first acquires basic operational data within the target scheduling period. This basic operational data includes at least: the on-grid electricity price and the purchase price of electricity in the power market; the electricity purchase and sale demand of each virtual power plant, the upper limit of output of micro gas turbines, the state of charge of the energy storage system, the available output of wind turbines, the interruptible load capacity, and the aggregated resource status of electric vehicles; the fixed generating power, adjustable generating power, fixed power consumption, available power consumption, and shared network transmission constraints of each charging station; and the state of charge, upper limit of charging power, upper limit of discharging power, estimated arrival and departure time, path distance, time cost coefficient, and battery degradation parameters of each electric vehicle user. This basic operational data serves as the unified input for the present invention to perform user response modeling, shared clearing, multi-virtual power plant scheduling, and robustness verification.

[0047] like Figure 1As shown, the electricity market in this invention is located at the top layer, used to provide external electricity purchase and sale price signals to virtual power plant operators; the virtual power plant operators are located at the lower layer, acting as the unified coordinating entity for multiple virtual power plants, determining the transaction purchase price and transaction sale price for each virtual power plant based on the electricity market price signals and the transaction response results of multiple virtual power plants; multiple virtual power plants are located at the middle layer, performing optimized scheduling based on the transaction prices and internal resource status; the charging station cluster and shared clearing mechanism are located at the lower layer, used to complete the inter-station shared power declaration and shared power clearing; electric vehicle users are located at the bottom layer, making dynamic decisions on whether to participate in sharing and which charging station to choose based on factors such as shared electricity price, time cost, driving energy consumption cost, battery degradation cost, and charging and discharging benefits.

[0048] In this embodiment, multiple charging stations can belong to different virtual power plants, or they can be assigned to the same virtual power plant according to region or aggregation relationship. Each charging station corresponds to at least one shared node, and multiple shared nodes are matched for sharing among stations through a shared clearing mechanism. The net sharing results of each charging station are aggregated to the corresponding virtual power plant according to their affiliation, forming the electric vehicle aggregation power boundary of the virtual power plant. Through this processing method, electric vehicle resources do not directly enter the virtual power plant scheduling as an abstract total amount, but first go through a step-by-step aggregation process of "user response - charging station sharing - result aggregation", making the electric vehicle aggregation power boundary entering the virtual power plant layer more consistent with the actual operating state.

[0049] In this embodiment, electric vehicle users during time periods Choose a charging station The total benefit from participating in sharing can be expressed as: ; in, Indicates the first Electric vehicle users during the time period Choose a charging station Total revenue shared by participants; Indicates the cumulative utility of charging; Indicates the cumulative cost of discharge; Indicates the revenue generated from charging and discharging costs; Indicates heading to a charging station Time cost; This indicates the additional energy consumption cost of driving; This represents the cost of battery degradation. This revenue model comprehensively describes the combined impact of shared electricity prices, user travel constraints, and battery lifespan deterioration on users' willingness to share.

[0050] The charging and discharging cost revenue can be expressed as: ; in, Indicates time period charging station Shared electricity pricing; Indicates the first Electric vehicle users during the time period The discharge power; Indicates the first Electric vehicle users during the time period The charging power; This indicates the length of the sharing period. When the discharge power is greater than the charging power, the user can obtain a net benefit through sharing; when the charging power is greater than the discharge power, the user pays the corresponding electricity fee to the charging station.

[0051] The state of charge (SBC) of electric vehicle users changes dynamically during the charging and discharging process, and its update relationship can be expressed as: ; in, Indicates the first Electric vehicle users during the time period Battery charge; Indicates the charge level in the previous time period; Indicates charging efficiency; This represents the discharge efficiency. This relationship is used to ensure that users can still meet their subsequent travel needs and minimum state of charge constraints after participating in sharing.

[0052] Considering that electric vehicle users lack complete information during actual sharing and cannot immediately obtain the optimal selection strategy within a single time period, this implementation uses evolutionary game theory to describe the changes in the probability of users choosing different charging stations. The discretized replicator dynamics can be represented as: ; in, Indicates the first During the next iteration, the user During the period Choose a charging station The probability of; This indicates the probability of selection after the next iteration; Indicates the iteration step size; Indicates user During the period The average expected return. When the return of a certain charging station is higher than the average return, the probability of choosing that station increases; conversely, it decreases, until an evolutionary steady state is reached.

[0053] Once the user base has stabilized, a charging station will be selected. The shared response aggregation term, which aggregates the charging and discharging power of all users, can be represented as follows: ; in, Indicates charging station During the period Electric vehicles share response power; Indicates selection of charging station The set of electric vehicle users. This aggregation item is used to characterize the net demand or net supply of the external shared network by the total number of users within the site. If A positive value indicates that the charging station has a high demand for sharing during that time period; if... A negative value indicates that the charging station has the ability to provide shared power to other charging stations.

[0054] Based on this, each charging station, according to its fixed power generation capacity, adjustable power generation capacity, fixed power consumption capacity, available power consumption capacity, and the aforementioned shared response aggregation item, generates a shared power demand or supply declaration and submits it to the shared clearing mechanism. The shared clearing mechanism, based on the shared network structure, line transmission capacity, and power balance conditions among the charging stations, performs real-time matching of shared power, enabling charging stations with surplus power to provide shared power to charging stations with insufficient power, thereby obtaining the net sharing result for each charging station in each time period. This net sharing result is used to form the aggregation sharing boundary for each charging station in that time period and serves as the input condition for subsequent virtual power plant-level electric vehicle aggregation scheduling.

[0055] like Figure 2 As shown, a two-layer master-slave game structure exists between the virtual power plant operator and multiple virtual power plants. The virtual power plant operator, as the leader, determines the transaction purchase price and transaction sales price of electricity based on the electricity market price and the purchase and sale response results of each virtual power plant; the multiple virtual power plants, as followers, perform internal resource optimization scheduling under the given transaction price conditions and provide feedback on their purchase and sale responses to the virtual power plant operator.

[0056] In this embodiment, the net revenue of the virtual power plant operator can be expressed as: ; in, This indicates the net revenue of virtual power plant operators; This indicates the on-grid electricity price in the electricity market; This indicates the electricity purchase price in the electricity market; This represents the amount of electricity sold by virtual power plant operators to the electricity market; This represents the amount of electricity purchased by the virtual power plant operator from the electricity market; This indicates the electricity price offered to virtual power plants. This indicates the electricity purchase price for virtual power plants; Indicates the first The amount of electricity purchased by a virtual power plant from an operator; Indicates the first The amount of electricity sold by a virtual power plant to an operator; Indicates the number of virtual power plants; This indicates the total number of time periods within the scheduling cycle.

[0057] To ensure the rationality of transactions between multiple virtual power plants and virtual power plant operators, the transaction price must satisfy: ; This price constraint ensures that the transaction prices set by virtual power plant operators are within the range of electricity purchase and sale prices in the electricity market, thereby guaranteeing both the profitability of virtual power plant operators and the incentive for each virtual power plant to respond to transaction price signals.

[0058] For the The operating cost of a virtual power plant, under a given transaction price, can be expressed as: ; in, Indicates the first Operating costs of a virtual power plant; Indicates the first A collection of resources within a virtual power plant, including micro gas turbines, energy storage systems, and interruptible loads; Indicates belonging to the first A collection of electric vehicles in a virtual power plant; Indicates the cost of a micro gas turbine; Indicates the cost of energy storage dispatch; This indicates the cost of interruptible load compensation; This indicates the cost of charging and discharging electric vehicles.

[0059] The operating cost of a micro gas turbine can be expressed as: ; The energy storage dispatch cost can be expressed as: ; The cost of interruptible load compensation can be expressed as: ; Using the aforementioned cost function, the optimal resource output combination for each virtual power plant can be determined under a given transaction price.

[0060] No. A virtual power plant during the time period The power balance relationship can be expressed as: ; in, Indicates the first A virtual power plant during the time period Net trading power with virtual power plant operators; Indicates the output of a micro gas turbine; Indicates the energy storage charging and discharging power; This indicates that interruptible loads can reduce power consumption. Indicates wind power output; This indicates the number of accesses or the daily load corresponding to the aggregated resources for electric vehicles; Indicates the discharge power of the electric vehicle; Indicates the power of a normal load; This represents the charging power of electric vehicles. By aggregating the results of charging station sharing into the aggregation boundary of the virtual power plant layer, user-side behavior and station group sharing results can be directly incorporated into the virtual power plant scheduling process.

[0061] In this embodiment, the net sharing result after the charging station sharing clearing is not directly used as the final scheduling output. Instead, it is first aggregated into the electric vehicle aggregated power boundary of the corresponding virtual power plant according to the affiliation relationship, and then incorporated into the virtual power plant power balance relationship. Through this processing method, user-side behavior and station group sharing results can directly affect the scheduling decision of the multi-virtual power plant layer.

[0062] In this embodiment, the final output results include at least: the transaction purchase price and transaction sales price of virtual power plant operators for each time period, the purchase and sales power of each virtual power plant, the output of micro gas turbines, the charging and discharging power of energy storage, the interruptible load reduction power, the wind power utilization power, the electric vehicle aggregated charging and discharging plan, the shared electricity price and shared power of each charging station, and the probability of electric vehicle users choosing each charging station. Based on the above output results, a real-time power sharing and robust aggregation scheduling scheme applicable to multi-virtual power plant scenarios can be formed.

[0063] Example 2 like Figure 3 As shown, the response results of electric vehicle user clusters differ significantly under different shared time period parameters. With the change in shared time period parameters... Changes in the time span of user participation in sharing, the achievable charging and discharging process, and the degree to which sharing revenue constraints are satisfied alter, leading to variations in the response power and the number of user cluster responses at different charging stations. This result indicates that the sharing time period parameter is one of the important factors influencing the sharing behavior of electric vehicle users.

[0064] In this embodiment, when the sharing time period parameter is small, the time available for electric vehicle users to complete the charging and discharging process is short. Some users do not participate in sharing due to insufficient sharing benefits, inadequate time constraints, or incomplete charging and discharging processes, resulting in a low overall number of responses. As the sharing time period parameter gradually increases, the operational space for users to participate in sharing increases, and the number of sharing responses at some charging stations significantly increases. When the sharing time period parameter continues to increase, the response distribution among different charging stations will be readjusted. This indicates that in this invention, the sharing time period parameter should not be set in isolation, but needs to be considered in conjunction with the sharing electricity price, user travel time window, and battery state of charge constraints to obtain a more reasonable user response result.

[0065] Example 3 like Figure 4 As shown, there are significant differences in the aggregated capacity curves of electric vehicles before and after the introduction of robust optimization. Without robust optimization, the aggregated capacity of electric vehicles fluctuates greatly over time, indicating that the system is quite sensitive to the access status of electric vehicles, charging and discharging power, and user response deviations. After introducing robust optimization, the aggregated capacity curve is smoother, indicating that the scheduling scheme has better stability after considering uncertainties.

[0066] In this embodiment, considering the fluctuations caused by electric vehicle access time, departure time, charging power, discharging power, and user response behavior, an uncertainty interval for electric vehicle aggregated resources is constructed. The uncertain variables can be represented as... ,in, Indicates the first Uncertain factors in the first The actual values ​​taken on each aggregate object; This indicates the lower bound of the uncertainty factor; This indicates its upper bound. The uncertainties include at least fluctuations in the number of electric vehicles connected, fluctuations in charging power, fluctuations in discharging power, and deviations in user sharing response.

[0067] To avoid overly conservative scheduling results, a robustness coefficient is introduced. The uncertainty interval is controlled. Robustness coefficient. Used to adjust the degree of consideration for the worst-case scenario, when When the value is small, the scheduling model is more biased towards economy; when When the value is large, the scheduling model is more stability-oriented. By adjusting the robustness coefficient, a trade-off can be struck between operating costs and resistance to fluctuations.

[0068] During the scheduling solution process, the uncertain interval is embedded in the virtual power plant power balance constraints and the electric vehicle aggregated power boundary to robustly verify the candidate solutions. If the power balance, energy storage power boundary, wind power utilization boundary, interruptible load boundary, and trading constraints are still satisfied under adverse operating scenarios, the scheduling scheme is considered feasible; otherwise, the shared electricity price, trading price, or internal resource output is readjusted and the iterative solution continues. Figure 4 The results show that after introducing robust optimization, the fluctuation of the electric vehicle aggregation capacity curve is reduced, indicating that the robust processing mechanism adopted in this invention can effectively smooth the aggregation capacity change and improve the stability and engineering feasibility of the scheduling scheme.

[0069] In summary, the real-time power sharing and robust aggregation scheduling method for electric vehicle charging station clusters with multiple virtual power plants proposed in this invention can organically combine electric vehicle user behavior, charging station cluster sharing mechanism, and multi-virtual power plant transaction scheduling. It achieves shared response on the user side, real-time power mutual assistance on the charging station side, aggregation optimization on the virtual power plant side, and transaction price coordination on the virtual power plant operator side. Furthermore, it improves the stability and executability of the system under adverse operating scenarios through robust optimization, thereby constructing a closed-loop collaborative scheduling method suitable for multi-virtual power plant scenarios.

[0070] The embodiments described above are merely illustrative of implementation methods of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters with multiple virtual power plants, characterized in that, Includes the following steps: S1. Obtain the operational data of virtual power plant operators, multiple virtual power plants, charging station clusters, and electric vehicle users within the scheduling period, and establish a multi-layer collaborative scheduling architecture of virtual power plant operators, multiple virtual power plants, charging station clusters, and electric vehicle users. S2. Based on the charging and discharging utility, sharing cost, time cost, additional driving energy consumption cost, battery degradation cost, state of charge and arrival and departure time of electric vehicle users, determine the probability of users choosing each charging station and the charging and discharging power; summarize the charging and discharging power of users who choose the same charging station, and determine the sharing supply and demand between stations in combination with the station's power generation, load and sharing network constraints, and perform real-time clearing to obtain the net sharing result of each charging station; S3. Establish a master-slave game model between virtual power plant operators and multiple virtual power plants. The virtual power plant operators determine the transaction purchase price and transaction sales price of electricity, while each virtual power plant determines the power purchase and sales capacity and internal resource output. The net sharing results of each charging station are summarized according to the relationship between the charging station and the virtual power plant to form the electric vehicle aggregated power boundary of the corresponding virtual power plant, and incorporated into the power balance and internal resource scheduling of the virtual power plant. S4. To address the fluctuations in electric vehicle access status, arrival and departure times, charging and discharging power, and user response results, an uncertainty interval is constructed and a robustness coefficient is set. The uncertainty interval is then embedded into the electric vehicle aggregated power boundary and the virtual power plant power balance constraint to robustly verify the candidate scheduling scheme. S5. Iterate in layers according to the order of user evolution response, charging station sharing clearing, aggregated power boundary formation, virtual power plant lower-level scheduling and virtual power plant operator upper-level pricing; if the robustness check is not met, adjust the shared electricity price, transaction price or internal resource output and continue iterating until the shared electricity price, purchased and sold power and the total objective function meet the preset convergence conditions, and output the real-time power sharing and robust aggregation scheduling scheme.

2. The method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters for multiple virtual power plants according to claim 1, characterized in that, The operational data includes: the on-grid electricity price and purchase price in the electricity market corresponding to the virtual power plant operator; the electricity purchase and sale demand reported by each virtual power plant and the results of electricity market transactions; the upper limit of output of micro gas turbines, the state of charge of energy storage systems, the available output of wind turbines, the interruptible load capacity, and the aggregated resource status of electric vehicles corresponding to each virtual power plant; the fixed power generation, adjustable power generation, fixed power consumption, available power consumption, and shared network constraints corresponding to each charging station; and the state of charge, upper limit of charging power, upper limit of discharging power, estimated arrival time, estimated departure time, path distance, time cost coefficient, and battery degradation parameters corresponding to each electric vehicle user.

3. The method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters for multiple virtual power plants according to claim 1, characterized in that: The total revenue from electric vehicle users participating in sharing is expressed as follows: ； in, Indicates the first Electric vehicle users during the time period Choose a charging station Total revenue shared by participants; Indicates the cumulative utility of charging; Indicates the cumulative cost of discharge; Indicates the revenue generated from charging and discharging costs; Indicates heading to a charging station Time cost; This indicates the additional energy consumption cost of driving; This represents the cost of battery degradation. This revenue model comprehensively characterizes the impact of shared electricity prices, user travel behavior, and battery life deterioration on users' willingness to participate in sharing.

4. The method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters for multiple virtual power plants according to claim 1, characterized in that: In step S2, evolutionary game theory is used to describe the changes in the probability of electric vehicle users choosing different charging stations, and the selection probability is dynamically updated through a replicator. When the benefit obtained by an electric vehicle user from choosing a charging station is higher than the average expected benefit, the probability of choosing that charging station is increased; when the benefit obtained from choosing a charging station is lower than the average expected benefit, the probability of choosing that charging station is decreased, until the selection probability of each charging station reaches an evolutionary stable state.

5. The method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters for multiple virtual power plants according to claim 4, characterized in that: After the selection probability reaches an evolutionary steady state, the charging power and discharging power of electric vehicle users who choose the same charging station are summarized, as expressed by: ； in, Indicates charging station During the period Electric vehicles share response power; Indicates selection of charging station The set of electric vehicle users; When the shared response power of the electric vehicle is positive, it is determined that the corresponding charging station has a shared power demand in the corresponding time period; when the shared response power of the electric vehicle is negative, it is determined that the corresponding charging station has the ability to provide shared power to other charging stations in the corresponding time period.

6. The method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters for multiple virtual power plants according to claim 5, characterized in that: Each charging station, based on its fixed power generation capacity, adjustable power generation capacity, fixed power consumption capacity, available power consumption capacity, shared electricity price, and the shared response power of the electric vehicles, formulates a shared power demand or shared power supply, and submits it to the shared clearing agency. The shared clearing mechanism, under the conditions of satisfying the shared network transmission constraints, line transmission cost constraints, and power balance, clears and matches the shared power demand and shared power supply of each charging station, so that charging stations with power surplus can provide shared power to charging stations with power deficit, thereby obtaining the net sharing result of each charging station in each time period.

7. The method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters for multiple virtual power plants according to claim 1, characterized in that: In step S3, multiple charging stations belong to different virtual power plants, or to the same virtual power plant according to regional or aggregation relationships. Each charging station corresponds to at least one shared node, and multiple shared nodes are matched for inter-station sharing through the shared clearing mechanism. The net sharing result of each charging station is not directly used as the final scheduling output, but is aggregated to the corresponding virtual power plant according to the affiliation between the charging station and the virtual power plant, forming the electric vehicle aggregation power boundary of the corresponding virtual power plant. The electric vehicle aggregation power boundary reflects the dynamic response results of electric vehicle users, the resource status within the charging station, and the operating boundary of the inter-station sharing network, and serves as the electric vehicle aggregation resource input of the corresponding virtual power plant, entering the power balance and internal resource scheduling process of the corresponding virtual power plant.

8. The method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters for multiple virtual power plants according to claim 1, characterized in that: In step S3, the virtual power plant operator, acting as the leader of the master-slave game model, aims to maximize the net profit of the virtual power plant operator, as expressed in the following expression: ； in, This indicates the net revenue of virtual power plant operators; This indicates the on-grid electricity price in the electricity market; This indicates the electricity purchase price in the electricity market; This represents the amount of electricity sold by virtual power plant operators to the electricity market; This represents the amount of electricity purchased by the virtual power plant operator from the electricity market; This indicates the electricity price offered to virtual power plants; This indicates the electricity purchase price for virtual power plants; Indicates the first The amount of electricity purchased by a virtual power plant from an operator; Indicates the first The amount of electricity sold by a virtual power plant to an operator; Indicates the number of virtual power plants; Indicates the total number of time periods within the scheduling period; Based on the electricity market purchase and sale prices and the purchase and sale responses from multiple virtual power plants, the transaction purchase price and transaction sale price for the multiple virtual power plants are determined; the transaction purchase price and transaction sale price are located between the electricity market on-grid price and the electricity market purchase price. Each virtual power plant, acting as a follower in the master-slave game model, aims to minimize its operating costs by coordinating internal conventional distributed resources and electric vehicle aggregated resources, given the transaction purchase price and transaction sales price of electricity, and determining the power output of the purchased and sold electricity and the output of internal resources.

9. The method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters for multiple virtual power plants according to claim 1, characterized in that: The uncertainties corresponding to the uncertain interval include fluctuations in the number of electric vehicles accessing the station, arrival time, departure time, charging power, discharge power, and user sharing response deviation. The uncertain interval is defined by the lower and upper bounds of the corresponding uncertainties. The robustness coefficient is used to adjust the degree of consideration for adverse operating scenarios. When the robustness coefficient decreases, the scheduling model is biased towards operating economy. When the robustness coefficient increases, the scheduling model is biased towards operating stability. The uncertain interval is embedded into the electric vehicle aggregate power boundary and the virtual power plant power balance constraint to perform robust verification of the candidate scheduling scheme; If a candidate scheduling scheme satisfies the power supply and demand balance, energy storage power boundary, wind power utilization boundary, interruptible load boundary, and transaction constraints under unfavorable operating scenarios, the candidate scheduling scheme is determined to be feasible; otherwise, the shared electricity price, transaction price, or internal resource output is adjusted and the solution is iterated again.

10. The method for real-time power sharing and robust aggregation scheduling of electric vehicle charging station clusters for multiple virtual power plants according to claim 1, characterized in that, The hierarchical iteration of step S5 includes: Given an initial shared electricity price, solve the electric vehicle user evolution response model to obtain the shared electric vehicle response power of each charging station; Based on the electric vehicle sharing response power and the resource status within each charging station, the charging station sharing clearing model is solved to obtain the net sharing result of each charging station. The net sharing results are aggregated into the electric vehicle aggregate power boundary of the corresponding virtual power plant according to the affiliation relationship between the charging station and the virtual power plant; Given the purchase price and sale price of electricity, solve the lower-level scheduling model of each virtual power plant to obtain the purchase and sale power and internal resource output of each virtual power plant. The virtual power plant operator updates the transaction purchase price and transaction sales price based on the power purchase and sale responses from multiple virtual power plants, and continues to iterate. The iteration stops when the change in the shared electricity price of each charging station is less than the preset threshold, the change in the purchased and sold electricity power of each virtual power plant is less than the preset threshold, and the change in the total objective function is less than the preset threshold in two consecutive iterations. Output the transaction purchase price and transaction sales price of electricity for each time period, the purchase and sales power of each virtual power plant, the output of internal resources, the electric vehicle aggregated charging and discharging plan, the shared electricity price and shared power of each charging station, and the probability of electric vehicle users choosing each charging station.

Images