Energy scheduling method for photovoltaic energy storage charging station considering uncertainty under incomplete information

Abstract

The present application belongs to the technical field of energy scheduling of optical storage charging station, and discloses a method for energy scheduling of optical storage charging station considering uncertainty under incomplete information, comprising: constructing a distributed optical storage charging station system under incomplete information; quantifying uncertainty through discretization of predicted scene; establishing an individual optimization problem based on the expected utility of the optical storage charging station, and defining a global optimization problem of the distributed optical storage charging station system based on the utility function of the distribution network; constructing a weighted Bayesian potential game model, and proving the consistency of the individual optimization direction and the global optimization direction; solving the weighted Bayesian potential game model by using a distributed optimization algorithm based on adaptive joint strategy virtual game, and iteratively updating the bidding strategy of the optical storage charging station by using an adaptive strategy updating mechanism based on Mann iteration until converging to an optimal equilibrium state to obtain an optimal bidding strategy. The present application optimizes the total revenue of the optical storage charging station and reduces the impact of power deviation on the power system.

Description

Technical Field

[0001] This invention relates to the field of energy scheduling technology for photovoltaic-storage-charging stations, and in particular to an energy scheduling method for photovoltaic-storage-charging stations that considers uncertainties under incomplete information. Background Technology

[0002] As a major consumer of fossil fuels and emitter of greenhouse gases, the transportation industry's green and low-carbon transformation is a crucial path to long-term decarbonization. As the core hub for energy interaction between electric vehicles and the power grid, photovoltaic-storage charging stations, integrating photovoltaic equipment, energy storage systems, and charging facilities, are gradually becoming key physical infrastructure for the electrification transformation of the transportation sector. With the gradual opening and decentralization of the electricity retail market, the number of independently operated photovoltaic-storage charging stations directly participating in market bidding in non-cooperative competitive scenarios is increasing. In a decentralized competitive market environment, each photovoltaic-storage charging station, as a self-interested and rational decision-making entity, typically does not communicate with each other and cannot obtain global information due to the need to protect commercial secrets, leaving the system in a state of incomplete information. Furthermore, photovoltaic-storage charging stations also face severe environmental uncertainties in actual operation. Affected by changes in weather conditions and users' random charging behavior, both photovoltaic output and electric vehicle charging demand exhibit high uncertainty. Therefore, under the dual influence of incomplete information and a highly uncertain environment, photovoltaic-storage charging stations are prone to making unreasonable bidding decisions when formulating real-time energy dispatch strategies, which not only harms their own revenue but also exacerbates power fluctuations in the power grid. Therefore, there is an urgent need for a distributed optimization strategy that can achieve incentive compatibility in a non-perfect information environment, so as to ensure that individual photovoltaic-storage charging stations can maximize their own interests while promoting the improvement of the overall system efficiency.

[0003] Bayesian game theory is widely used in market bidding and resource allocation scenarios for non-cooperative optimization problems under incomplete information. In traditional incomplete information game solving, each player needs to monitor the actions of all other players individually and calculate the probability distribution of joint actions. However, with the increase in system scale, the exponential growth of the joint state space makes it highly susceptible to the curse of dimensionality in large-scale distributed systems, resulting in excessively high computational and communication complexity, thus failing to meet the requirements of real-time scheduling. Furthermore, distributed solution algorithms often lack suitable adaptive update mechanisms during policy iteration. Inappropriate parameter updates lead to low global search efficiency when the system state deviates from the equilibrium point; conversely, when the system state approaches the equilibrium point, it is prone to violent oscillations or premature convergence, failing to guarantee accurate and stable convergence. Currently, there is a lack of a real-time energy scheduling optimization strategy that can effectively overcome the curse of dimensionality and quickly converge to the equilibrium state under uncertain scenarios with incomplete information. Summary of the Invention

[0004] In view of this, the purpose of this invention is to propose an energy dispatching method for photovoltaic-storage charging stations that considers uncertainties under incomplete information. This method comprehensively considers the dual uncertainties of photovoltaic output and electric vehicle charging load, the expected revenue of photovoltaic-storage charging stations, and the stability of the power distribution network, and provides an optimal energy dispatching strategy for photovoltaic-storage charging stations. This optimizes the total revenue of photovoltaic-storage charging stations and reduces the impact of power deviation on the power system.

[0005] To achieve the above-mentioned technical objectives, the technical solution adopted by this invention is as follows: This invention provides an energy scheduling method for photovoltaic-storage-charging stations considering uncertainties under incomplete information, comprising the following steps: Step 1: Construct a distributed photovoltaic-storage-charging station system under incomplete information, including the power distribution network and multiple photovoltaic-storage-charging stations; Step 2: Under the distributed photovoltaic-storage-charging station, the dual uncertainties of photovoltaic output and electric vehicle charging load are quantified by discretizing the prediction scenario; Step 3: Based on the expected utility of the photovoltaic-storage-charging station, establish an individual optimization problem with the goal of maximizing the expected utility of the independent photovoltaic-storage-charging station, and define a global optimization problem for the distributed photovoltaic-storage-charging station system based on the utility function of the distribution network. Step 4: Construct a weighted Bayesian potential game model based on the individual optimization problem and the global optimization problem, and prove the consistency between the individual optimization direction and the global optimization direction to achieve incentive compatibility; Step 5: The weighted Bayesian potential game model is solved using a distributed optimization algorithm based on adaptive joint strategy virtual game. Each photovoltaic-storage charging station only relies on local prediction data and publicly available information from the distribution network to update its bidding strategy. During the bidding strategy update process, an adaptive strategy update mechanism based on Mann iteration is used to iteratively update the bidding strategy of the photovoltaic-storage charging station until it converges to the optimal equilibrium state, thus obtaining the optimal bidding strategy.

[0006] Furthermore, step 1 specifically includes: Step 11: Construct a distributed photovoltaic-storage-charging station system under incomplete information conditions. The distributed photovoltaic-storage-charging station system consists of a power distribution network and... Each independently operated and non-communication-connected photovoltaic energy storage charging station constitute, ,in, i Index for photovoltaic-storage charging stations. i =1,2…… n ; This indicates the number of photovoltaic and energy storage charging stations; Indicates the first i One photovoltaic energy storage charging station; Step 12: Divide the total time T of a day into Each period, ,in, t This indicates a time period index.

[0007] Furthermore, step 2 specifically includes: Step 21: Based on the empirical probability distribution of historical photovoltaic power output data, construct a set of discrete photovoltaic power output scenarios. ,in, Time-of-use photovoltaic energy storage charging station In photovoltaic power output scenarios The actual output power of the photovoltaic system is denoted as: The probability of occurrence is And satisfy ; Step 22: Based on the empirical probability distribution of historical electric vehicle actual charging load data, construct a set of discrete electric vehicle charging demand scenarios. ,in, Time-of-use photovoltaic energy storage charging station In each electric vehicle charging demand scenario The total charging power required by electric vehicles is denoted as The probability of occurrence is And satisfy ; Step 23: Assume a photovoltaic power output scenario Electric vehicle charging demand scenarios Independent of each other, the set of joint scenarios of the two is defined as For joint scenarios Its probability of occurrence is And satisfy .

[0008] Furthermore, the distribution network, acting as a market seller in real-time dispatch, publishes dynamic electricity prices and uses price signals to guide the electricity purchasing behavior of each photovoltaic-storage charging station; specifically including: (1) The power distribution network in Dynamic electricity prices released during specific time periods The dynamic electricity price is composed of the benchmark time-of-use tariff and the bidding power purchased by each photovoltaic-storage-charging station. The calculation formula is as follows:

[0009] in, for The benchmark time-of-use electricity price for the distribution network during a given period. For the first i One photovoltaic energy storage charging station The elasticity of demand sensitivity coefficient for Before the start of the time period i One photovoltaic energy storage charging station The submitted bid power for electricity purchase; (2) For joint scenarios , Time period i One photovoltaic energy storage charging station Power deviation for:

[0010] in, for Time period i One photovoltaic energy storage charging station In joint scenarios The actual power purchased; (3) When there is a power deviation in the photovoltaic-storage charging station, additional power balancing costs need to be paid to the distribution network. Power deviation penalty The following are examples:

[0011]

[0012] in, for Time period i One photovoltaic energy storage charging station In joint scenarios Power balancing cost under these conditions for Time period i One photovoltaic energy storage charging station In joint scenarios Power deviation penalty For the first i One photovoltaic energy storage charging station Self-defined profit coefficient; j This represents the index of photovoltaic and energy storage charging stations, and i≠j ; s j Indicates the first j One photovoltaic energy storage charging station; for Time period j One photovoltaic energy storage charging station s j In joint scenarios The actual power purchased; The photovoltaic-storage charging station operates as a self-interested game in real-time scheduling, and consists of a photovoltaic system, an energy storage system, a charging pile system, and an energy management system; specifically including: (1) For photovoltaic systems, the energy generated by the photovoltaic system is preferentially used to meet the charging load within the photovoltaic-storage charging station, and any surplus is stored in the energy storage system. If the surplus energy cannot be fully utilized, it will be discarded; specifically including: Time-of-use photovoltaic energy storage charging station In joint scenarios Actual photovoltaic power consumption for:

[0013] in, for Time-of-use photovoltaic energy storage charging station In photovoltaic power output scenarios The actual output power of the photovoltaic system for Time-of-use photovoltaic energy storage charging station In electric vehicle charging scenarios The total charging power required for electric vehicles for Time-of-use photovoltaic energy storage charging station The charging and discharging power of the energy storage system is negative when it indicates charging and positive when it indicates discharging. for Time-of-use photovoltaic energy storage charging station In joint scenarios The actual power purchased; (2) For the energy storage system, the energy storage system provides real-time energy regulation capability for the photovoltaic-energy storage charging station, reducing the operating cost of the photovoltaic-energy storage charging station; specifically including: a, Time-of-use photovoltaic energy storage charging station The charging and discharging power of medium-energy storage system It is a variable that needs to be decided through a scheduling strategy, and is represented as:

[0014] in, for Time-of-use photovoltaic energy storage charging station The energy storage charge-discharge coefficient, Photovoltaic and energy storage charging stations Maximum charging power, Photovoltaic and energy storage charging stations Maximum discharge power; b、 Time-of-use photovoltaic energy storage charging station The SOC of the medium energy storage system is The formula is as follows:

[0015] in, Photovoltaic and energy storage charging stations The upper limit of SOC for medium-sized energy storage systems Photovoltaic and energy storage charging stations The lower limit of medium-sized energy storage systems for Time-of-use photovoltaic energy storage charging station SOC of medium energy storage system Photovoltaic and energy storage charging stations Medium energy storage battery capacity, The length of a time period The energy conversion coefficient of the energy storage system during the charging and discharging process is given by the following formula:

[0016] in, To improve the charging efficiency of energy storage systems. The discharge efficiency of the energy storage system; c. Time-of-use photovoltaic energy storage charging station Energy storage loss cost Represented as:

[0017] in, This is the energy storage loss cost coefficient; (3) For the charging pile system, the charging pile system is the main load source in the photovoltaic-storage charging station, and its charging load is affected by the uncertainty of electric vehicle charging demand; specifically including: a, Time-of-use photovoltaic energy storage charging station In joint scenarios Total charging power of electric vehicles for:

[0018] in, for Time-of-use photovoltaic energy storage charging station In electric vehicle charging scenarios Total charging power demand for electric vehicles in China; b. The electricity purchased by the photovoltaic-storage charging station from the distribution network will generate revenue through sales to electric vehicles. The profit coefficient is hour, Time-of-use photovoltaic energy storage charging station In joint scenarios Comprehensive benefits Represented as:

[0019] in, The estimated revenue from the current power purchase capacity. The marginal revenue of energy within the photovoltaic-storage charging station; (4) For the energy management system, the energy management system makes energy scheduling decisions with the goal of maximizing the revenue of the photovoltaic energy storage charging station itself; Before the start of the scheduling period, the energy management system assesses the uncertainties of the scenario based on local incomplete information and decides on the power capacity to be purchased through bidding. and participate in the game interaction; During the scheduling period, the energy management system performs actual energy scheduling based on the real-world scenario. Time-of-use photovoltaic energy storage charging station In joint scenarios Actual power purchase capacity for: ; Time-of-use photovoltaic energy storage charging station In joint scenarios The actual utility for: .

[0020] Furthermore, step 3 specifically includes: Step 31, Calculation Time-of-use photovoltaic energy storage charging station In the joint scenario set Expected power purchase capacity With expected photovoltaic power absorption capacity The expected power purchase With expected photovoltaic power absorption capacity The calculation formulas are as follows:

[0021]

[0022] Step 32, Calculation Time-of-use photovoltaic energy storage charging station In the joint scenario set Expected utility The formula is as follows:

[0023] ; Step 33, Time-of-use photovoltaic energy storage charging station Individual optimization problem Formal representation:

[0024]

[0025] ; in, for Except for the first period before the start of the time period i The bidding power submitted by all other photovoltaic-storage charging stations besides [the one mentioned] The maximum power purchase capacity for the photovoltaic-storage charging station; Step 34: Convert the global desired potential function As the utility function of the distribution network, the formula is as follows:

[0026] in, Price penalty for power deviation; Step 35: Define the global optimization problem of the distributed photovoltaic energy storage and charging station system. for:

[0027]

[0028] .

[0029] Furthermore, step 4 specifically includes: Step 41: The weighted Bayesian potential game model is a non-cooperative Bayesian game under incomplete information. For each time period... The weighted Bayesian potential game model is expressed as:

[0030] The specific definitions of each element are as follows: (1) Player: For players, that is, a collection of independently operating photovoltaic energy storage and charging stations; (2) Strategy: For each player , its in The action strategy for a given time period is to submit the bid-to-purchase power capacity to the distribution network, i.e. The joint strategy space is In this context, the symbol := indicates that it is defined as , the symbol X represents the Cartesian product, and represents the set of combinations of all player strategy spaces; (3) Type: Player Type Used to indicate the uncertainty it faces. Time Period Player The type is defined as The union type space is That is, the Cartesian product of all player types, representing the set of combinations of all player type spaces; (4) Beliefs: To define in the union type space The joint probability distribution on the tectonic scale represents the common prior beliefs of all players about the system state, derived from historical data; (5) Utility: For players Expected utility; Step 42, assuming in Time period, any player Unilaterally changing the strategy Change to Other players' strategies If it remains unchanged, then the change in the player's expected utility under this policy shift is... Represented as:

[0031] Where · is a placeholder; Meanwhile, the change in the global expected potential function under this strategy offset Represented as:

[0032] Factoring out the common factor yields the change in the player's expected utility under this policy shift. The change in the global expected potential function The relationship between them is:

[0033] And due to the profit coefficient ,so Established; Therefore, the individual optimization direction is consistent with the global optimization direction, and the distributed photovoltaic energy storage charging station system has incentive compatibility characteristics.

[0034] Furthermore, the distributed optimization algorithm based on adaptive joint strategy virtual game in step 5 includes two stages: The first stage is the prediction and initialization stage. Each photovoltaic and energy storage charging station generates a joint scenario based on local prediction data, performs preliminary optimization to minimize local costs, calculates the expected power purchase, generates an initial bidding strategy to minimize costs, and uploads it to the distribution network. The distribution network calculates the initial dynamic electricity price and the total expected power purchase based on all the collected initial bidding strategies. The second stage is the bidding and equilibrium iteration stage. The distribution network broadcasts the current dynamic electricity price and the total expected power purchase to all photovoltaic and energy storage charging stations. Each photovoltaic and energy storage charging station independently calculates its local best response and updates its bidding strategy based on the initial dynamic electricity price and total expected power purchase obtained from the distribution network, and then re-uploads it to the distribution network. The distribution network collects the updated bidding strategies of all photovoltaic and energy storage charging stations, and synchronously updates the dynamic electricity price and the global expected potential function until the convergence condition is met. The weighted Bayesian potential game model reaches the optimal equilibrium state, and the optimal bidding strategy is obtained.

[0035] Furthermore, the first stage specifically includes: Step 511: Initialize the dynamic electricity price of the distribution network. Global expected potential function Each photovoltaic energy storage charging station Expected power purchase capacity Bidding for power purchase Inertia coefficient as well as Time-of-use photovoltaic energy storage charging station Degree of deviation from equilibrium point ; Step 512: For each photovoltaic-storage charging station It employs a distributed parallel computing approach to generate joint scenarios based on local prediction data. ; Step 513: Calculate the expected power purchase capacity according to the formula for expected power purchase capacity. The formula for calculating the expected power purchase is as follows:

[0036] Step 514: Generate the initial bidding power with minimal cost. This serves as the initial bidding strategy and is sent to the distribution network. Step 515: The distribution network collects the initial bidding strategies of all photovoltaic-storage charging stations and calculates the initial dynamic electricity price according to the dynamic electricity price calculation formula. And calculate the total expected power purchase. The formula for calculating the dynamic electricity price is as follows: .

[0037] Furthermore, the second stage specifically includes: Step 521: Enter the iterative loop; the distribution network broadcasts the current dynamic electricity price to all photovoltaic-storage charging stations. Total expected power purchase ; Step 522: For each photovoltaic-storage charging station It employs a distributed parallel computing approach, receiving information broadcast from the distribution network and calculating the optimal response based on the formula. Time-based photovoltaic energy storage charging In the Optimal response during round iteration The formula for calculating the optimal response is as follows:

[0038] Step 523: Calculate according to the formula for calculating the degree of equilibrium deviation. Time-of-use photovoltaic energy storage charging station Degree of deviation from equilibrium point The calculation formula is as follows:

[0039] in, To prevent constants with a denominator of zero; Step 524: Calculate the photovoltaic-storage-charging station based on the inertia coefficient calculation formula. In the Inertia coefficient during round iteration The formula for calculating the inertia coefficient is as follows:

[0040] in, The maximum inertia coefficient, This is the maximum power adjustment range. To balance the judgment threshold; Step 525: Update the formula according to the bidding strategy. Time-based photovoltaic energy storage charging In the Bidding strategy during round iteration The updated bidding strategy is then sent to the distribution network; the bidding strategy update formula is as follows:

[0041] in, for Time-based photovoltaic energy storage charging In the Bidding strategies during round iterations; Step 526: The distribution network collects the updated bidding strategies of all photovoltaic-storage-charging stations and updates the dynamic electricity price according to the dynamic electricity price formula. ; Step 527: Update the global expected potential function according to the calculation formula of the global expected potential function. The formula for calculating the global desired potential function is as follows:

[0042] Step 528: Repeat steps 521 to 527 until the convergence condition is met. The weighted Bayesian potential game model reaches the optimal equilibrium state, and the optimal bidding strategy is obtained.

[0043] Furthermore, the convergence condition is one of the following two conditions: Condition 1: The change in bidding strategies for all photovoltaic-storage-charging stations is less than a preset threshold, i.e. ,in, The preset threshold; Condition 2: Current iteration round Reaching the preset maximum number of iterations ,Right now .

[0044] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art: This invention addresses the self-interested bidding scheduling problem of independent photovoltaic-storage-charging stations under incomplete information and environmental uncertainty. It proposes an energy scheduling strategy for photovoltaic-storage-charging stations based on Adaptive Joint Strategy Fictitious Play (A-JSFP) to achieve incentive compatibility of the system and ensure the synergistic improvement of individual interests and overall benefits.

[0045] First, by discretizing the prediction scenarios to quantify the uncertainties of photovoltaic power output and charging load, and combining dynamic electricity prices and power deviation penalty mechanisms, an individual optimization problem is established with the goal of maximizing the expected utility of independent photovoltaic-storage-charging stations. A global expected potential function reflecting the global benefits of the system is established using the utility function of the distribution network, defining the global optimization problem. Second, a weighted Bayesian potential game model is constructed, and the consistency between the individual optimization and global optimization directions is proven. Finally, to address the curse of dimensionality in solving the weighted Bayesian potential game model under incomplete information, a distributed optimization algorithm based on A-JSFP is proposed, and an adaptive policy update mechanism based on Mann iteration is designed to balance the search efficiency in the early stages of the algorithm with the convergence stability in the later stages. Experimental results show that by solving the incentive-compatible weighted Bayesian potential game model, the curse of dimensionality can be effectively overcome, and the algorithm can converge to an equilibrium state quickly and smoothly using limited information. This strategy can effectively improve the convergence efficiency and reduce power deviation. Furthermore, under different scenarios, this strategy effectively improves the comprehensive operating benefits of photovoltaic-storage-charging stations and the operating efficiency of the distribution network. This invention comprehensively considers the dual uncertainties of photovoltaic power output and electric vehicle charging load, the expected revenue of photovoltaic-storage charging stations, and the stability of the power distribution network, and provides an optimal energy dispatch strategy for photovoltaic-storage charging stations, thereby optimizing the total revenue of photovoltaic-storage charging stations and reducing the impact of power deviation on the power system. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is an execution flowchart of an energy scheduling method for photovoltaic-storage-charging stations that considers uncertainties under incomplete information, provided by an embodiment of the present invention.

[0048] Figure 2 This is a framework diagram of a distributed photovoltaic energy storage and charging station system considering uncertainties, provided by an embodiment of the present invention.

[0049] Figure 3 This is an execution flowchart of the distributed optimization algorithm based on adaptive joint strategy virtual game provided in the embodiments of the present invention. Detailed Implementation

[0050] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be particularly noted that the following embodiments are for illustrative purposes only and do not limit the scope of the invention. Similarly, the following embodiments are only some, not all, embodiments of the present invention, and all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] Please see Figures 1-3 The present invention provides an energy scheduling method for photovoltaic-storage-charging stations considering uncertainties under incomplete information, comprising the following steps: Step 1: Construct a distributed photovoltaic-storage-charging station system under incomplete information, including the power distribution network and multiple photovoltaic-storage-charging stations; In this embodiment, step 1 specifically includes: Step 11: Construct a distributed photovoltaic-storage-charging station system under incomplete information conditions. The distributed photovoltaic-storage-charging station system consists of a power distribution network and... Each independently operated and non-communication-connected photovoltaic energy storage charging station constitute, ,in, i Index for photovoltaic-storage charging stations. i =1,2…… n ; This indicates the number of photovoltaic and energy storage charging stations; Indicates the first i Each photovoltaic-storage charging station conducts real-time independent energy bidding based on incomplete local information and interacts non-cooperatively through the dynamic electricity pricing mechanism of the distribution network.

[0052] Step 12: Divide the total time T of a day into Each period, ,in, t This represents a time period index, assuming that the power of each part of the system is approximately constant within each time period, and ignoring energy loss during transmission.

[0053] Step 2: Under the distributed photovoltaic-storage-charging station, the dual uncertainties of photovoltaic output and electric vehicle charging load are quantified by discretizing the prediction scenario; For real-time scheduling , its in The photovoltaic output and electric vehicle charging demand during a given time period are unknown and uncertain. The uncertainty of the scenario is defined as follows: Step 21, Photovoltaic Output Scenarios: Due to the influence of meteorological conditions, the output power of photovoltaic systems is uncertain. Based on the empirical probability distribution of historical photovoltaic output data, a set of discrete photovoltaic output scenarios is constructed. ,in, Time-of-use photovoltaic energy storage charging station In photovoltaic power output scenarios The actual output power of the photovoltaic system is denoted as: The probability of occurrence is And satisfy ; Step 22, Electric Vehicle Charging Demand Scenarios: The random arrival and uncertain charging behavior of electric vehicles lead to significant uncertainty in charging load. Based on the empirical probability distribution of historical electric vehicle actual charging load data, a discrete set of electric vehicle charging demand scenarios is constructed. ,in, Time-of-use photovoltaic energy storage charging station In each electric vehicle charging demand scenario The total charging power required by electric vehicles is denoted as The probability of occurrence is And satisfy ; Step 23: Assume a photovoltaic power output scenario Electric vehicle charging demand scenarios Independent of each other, the set of joint scenarios of the two is defined as For joint scenarios Its probability of occurrence is And satisfy .

[0054] In this embodiment, the distribution network acts as a market seller in real-time dispatching, publishing dynamic electricity prices and using price signals to guide the electricity purchasing behavior of each photovoltaic-storage charging station; specifically including: (1) The power distribution network in Dynamic electricity prices released during specific time periods The dynamic electricity price is composed of the benchmark time-of-use tariff and the bidding power purchased by each photovoltaic-storage-charging station. The calculation formula is as follows: (1) in, for The benchmark time-of-use electricity price for the distribution network during a given period. For the first i One photovoltaic energy storage charging station The elasticity of demand sensitivity coefficient for Before the start of the time period i One photovoltaic energy storage charging station The submitted bid power for electricity purchase; (2) To mitigate real-time supply and demand fluctuations caused by prediction errors and alleviate the operational pressure on the power system, the distribution network adopts a deviation penalty mechanism based on instantaneous load billing for photovoltaic-storage charging stations. For combined scenarios... , Time period i One photovoltaic energy storage charging station Power deviation for: (2) in, for Time period i One photovoltaic energy storage charging station In joint scenarios The actual power purchased; (3) When there is a power deviation in the photovoltaic-storage charging station, additional power balancing costs need to be paid to the distribution network. Power deviation penalty The following are examples: (3) (4) in, for Time period i One photovoltaic energy storage charging station In joint scenarios Power balancing cost under these conditions for Time period i One photovoltaic energy storage charging station In joint scenarios Power deviation penalty For the first i One photovoltaic energy storage charging station Self-defined profit coefficient; j This represents the index of photovoltaic and energy storage charging stations, and i≠j ; s j Indicates the first j One photovoltaic energy storage charging station; for Time period j One photovoltaic energy storage charging station s j In joint scenarios The actual power purchased; In this embodiment, the photovoltaic-storage charging station is a self-interested player in real-time scheduling, and consists of a photovoltaic system, an energy storage system, a charging pile system, and an energy management system; specifically including: (1) For photovoltaic systems: The photovoltaic system provides free clean energy to the photovoltaic-storage charging station. To maximize the utilization rate of clean energy, the energy generated by the photovoltaic system is first used to meet the charging load within the photovoltaic-storage charging station, and the surplus is stored in the energy storage system. If the surplus energy cannot be fully utilized, it will be discarded; specifically including: Time-of-use photovoltaic energy storage charging station In joint scenarios Actual photovoltaic power consumption for: (5) in, for Time-of-use photovoltaic energy storage charging station In photovoltaic power output scenarios The actual output power of the photovoltaic system for Time-of-use photovoltaic energy storage charging station In electric vehicle charging scenarios The total charging power required for electric vehicles for Time-of-use photovoltaic energy storage charging station The charging and discharging power of the energy storage system is negative when it indicates charging and positive when it indicates discharging. for Time-of-use photovoltaic energy storage charging station In joint scenarios The actual power purchased; (2) For the energy storage system, the energy storage system provides real-time energy regulation capability for the photovoltaic-energy storage charging station, reduces the operating cost of the photovoltaic-energy storage charging station, and mitigates the direct impact of electric vehicle charging load on the power system; specifically including: a, Time-of-use photovoltaic energy storage charging station The charging and discharging power of medium-energy storage system It is a variable that needs to be decided through a scheduling strategy, and is represented as: (6) in, for Time-of-use photovoltaic energy storage charging station The energy storage charge-discharge coefficient, Photovoltaic and energy storage charging stations Maximum charging power, Photovoltaic and energy storage charging stations Maximum discharge power; b、 Time-of-use photovoltaic energy storage charging station The SOC of the medium energy storage system is The formula is as follows: (7) in, Photovoltaic and energy storage charging stations The upper limit of SOC for medium-sized energy storage systems Photovoltaic and energy storage charging stations The lower limit of medium-sized energy storage systems for Time-of-use photovoltaic energy storage charging station SOC of medium energy storage system Photovoltaic and energy storage charging stations Medium energy storage battery capacity, The length of a time period The energy conversion coefficient of the energy storage system during the charging and discharging process is given by the following formula: (8) in, To improve the charging efficiency of energy storage systems. The discharge efficiency of the energy storage system; c. Time-of-use photovoltaic energy storage charging station Energy storage loss cost Represented as: (9) in, This is the energy storage loss cost coefficient; (3) For the charging pile system, the charging pile system is the main load source in the photovoltaic-storage charging station, and its charging load is affected by the uncertainty of electric vehicle charging demand; specifically including: a, Time-of-use photovoltaic energy storage charging station In joint scenarios Total charging power of electric vehicles for: (10) in, for Time-of-use photovoltaic energy storage charging station In electric vehicle charging scenarios Total charging power demand for electric vehicles in China; b. The electricity purchased by the photovoltaic-storage charging station from the distribution network will generate revenue through sales to electric vehicles. The profit coefficient is hour, Time-of-use photovoltaic energy storage charging station In joint scenarios Comprehensive benefits Represented as: (11) in, The estimated revenue from the current power purchase capacity. The marginal revenue of energy within the photovoltaic-storage charging station; (4) For the energy management system, the energy management system makes energy scheduling decisions with the goal of maximizing the revenue of the photovoltaic energy storage charging station itself; Before the start of the scheduling period, the energy management system assesses the uncertainties of the scenario based on local incomplete information and decides on the power capacity to be purchased through bidding. and participate in the game interaction; During the scheduling period, the energy management system performs actual energy scheduling based on the real-world scenario. Time-of-use photovoltaic energy storage charging station In joint scenarios Actual power purchase capacity for: (12) Time-of-use photovoltaic energy storage charging station In joint scenarios The actual utility for: (13) Step 3: Based on the expected utility of the photovoltaic-storage-charging station, establish an individual optimization problem with the goal of maximizing the expected utility of the independent photovoltaic-storage-charging station, and define a global optimization problem for the distributed photovoltaic-storage-charging station system based on the utility function of the distribution network. In this embodiment, step 3 specifically includes: Step 31, Calculation Time-of-use photovoltaic energy storage charging station In the joint scenario set Expected power purchase capacity With expected photovoltaic power absorption capacity The expected power purchase With expected photovoltaic power absorption capacity The calculation formulas are as follows:

[0055] (14) Step 32, Calculation Time-of-use photovoltaic energy storage charging station In the joint scenario set Expected utility The formula is as follows: (15) ; Step 33, Time-of-use photovoltaic energy storage charging station Individual optimization problem Formal representation:

[0057] (16) ; in, for Except for the first period before the start of the time period i The bidding power submitted by all other photovoltaic-storage charging stations besides [the one mentioned] The maximum power purchase capacity for the photovoltaic-storage charging station; Step 34: The distribution network aims to maximize electricity sales revenue and minimize the power deviation before and after the bidding for photovoltaic-storage charging stations. To reflect the impact of the bidding decision on the distribution network, a global expected potential function is introduced. The global expected potential function As the utility function of the distribution network, the formula is as follows: (17) in, Price penalty for power deviation; Step 35: Define the global optimization problem of the distributed photovoltaic energy storage and charging station system. for:

[0058] (18) .

[0059] Step 4: Construct a weighted Bayesian potential game model based on the individual optimization problem and the global optimization problem, and prove the consistency between the individual optimization direction and the global optimization direction to achieve incentive compatibility; In this embodiment, step 4 specifically includes: Step 41: The weighted Bayesian potential game model is a non-cooperative Bayesian game under incomplete information. Due to the high uncertainty of photovoltaic power output and electric vehicle charging demand, and the fact that the system is in a state of incomplete information, this chapter models the real-time energy scheduling problem as a non-cooperative Bayesian game under incomplete information in order to achieve synergistic optimization of distribution network benefits and individual interests of photovoltaic-storage-charging stations. For each time period... The weighted Bayesian potential game model is expressed as:

[0060] The specific definitions of each element are as follows: (1) Player: For players, that is, a collection of independently operating photovoltaic energy storage and charging stations; (2) Strategy: For each player , its in The action strategy for a given time period is to submit the bid-to-purchase power capacity to the distribution network, i.e. The joint strategy space is In this context, the symbol := indicates that it is defined as , the symbol X represents the Cartesian product, and represents the set of combinations of all player strategy spaces; (3) Type: Player Type This is used to represent the uncertainty it faces, since the uncertainty of photovoltaic output and electric vehicle charging demand is mapped to different power purchases in the energy management system. Time Period Player The type is defined as The union type space is That is, the Cartesian product of all player types, representing the set of combinations of all player type spaces; (4) Beliefs: To define in the union type space The joint probability distribution on the tectonic scale represents the common prior beliefs of all players about the system state, derived from historical data; (5) Utility: For players The expected utility is shown in formula (15); In non-cooperative Bayesian games, the stable state of the system is represented by the Bayesian Nash Equilibrium (BNE).

[0061] Define BNE points in Bayesian games In the middle, strategy group The BNE point is constituted if and only if for any player and any feasible strategy It meets the following conditions: (19) When the system is at the BNE point, no player can unilaterally change their strategy to gain a higher expected return. Therefore, the system is in a stable state.

[0062] Define a weighted Bayesian potential game: For a Bayesian game If there exists a global expected potential function and a set of consistently positive weight vectors This makes it possible for any player and any strategy group And another strategy of the player All satisfy: (20) Then the game For weighted Bayesian potential games.

[0063] Theorem 1: Expected Utility Based on Photovoltaic-Storage Charging Stations (Equation (15)) and the global expected potential function The game constituted by (Formula (17)) It is a weighted Bayesian potential game.

[0064] Step 42 (Proof Process): Assume that... Time period, any player Unilaterally changing the strategy Change to Other players' strategies If it remains unchanged, then the change in the player's expected utility under this policy shift is... Represented as: (twenty one) Where · is a placeholder; Meanwhile, the change in the global expected potential function under this strategy offset Represented as: (twenty two) Factoring out the common factor yields the change in the player's expected utility under this policy shift. The change in the global expected potential function The relationship between them is: (twenty three) And due to the profit coefficient ,so Established; therefore, game theory It is a weighted Bayesian potential game, Q.E.D.

[0065] As shown in Theorem 1, the individual optimization direction of a photovoltaic-storage-charging station is consistent with the global optimization direction of the distributed photovoltaic-storage-charging station system, indicating that the distributed photovoltaic-storage-charging station system possesses incentive compatibility characteristics. Due to these incentive compatibility characteristics, focusing solely on maximizing the utility of the photovoltaic-storage-charging station is sufficient to ensure the maximization of the overall system benefits.

[0066] Theorem 2: In weighted Bayesian potential games There exists a unique BNE point in the system.

[0067] Proof: According to the mathematical properties of potential games, the Nash equilibrium point of a weighted Bayesian potential game is equivalent to the maximum point of the global expected potential function. For Regarding Second-order partial derivatives Therefore, the global expected potential function In the joint strategy space The above about It is a strictly concave function. Furthermore, due to the strategy space of each player... All are non-empty and compact convex sets, therefore the joint policy space is... It is also a convex set. According to convex optimization theory, a strictly concave function must have a unique global maximum point on a convex compact set, that is, a unique BNE point. Q.E.D.

[0068] Step 5: The weighted Bayesian potential game model is solved using a distributed optimization algorithm based on adaptive joint strategy virtual game theory. Each photovoltaic-storage-charging station updates its bidding strategy based solely on local prediction data and publicly available information from the distribution network. During the bidding strategy update process, an adaptive strategy update mechanism based on Mann iteration is used to iteratively update the bidding strategy of each photovoltaic-storage-charging station until it converges to the optimal equilibrium state, thus obtaining the optimal bidding strategy. Under the incomplete information and distributed non-cooperative framework, each photovoltaic-storage-charging station cannot know the actions and payoffs of all nodes globally. In traditional virtual games, each player needs to monitor the actions of all other players individually and calculate the probability distribution of joint actions, which can lead to the curse of dimensionality in large-scale distributed systems. To overcome this limitation, a distributed optimization algorithm based on adaptive joint strategy virtual game theory (A-JSFP) is introduced, enabling each photovoltaic-storage-charging station to dynamically update its bidding strategy based solely on local predictions and limited publicly available information from the distribution network. Furthermore, the adaptive control of the inertia coefficient effectively suppresses oscillations, ensuring that the system converges smoothly and efficiently to the optimal equilibrium state.

[0069] (a) Joint Strategy Virtual Game: In an incomplete information and distributed non-cooperative framework, individual photovoltaic-storage-charging stations cannot obtain information about the actions and rewards of all nodes globally. In traditional virtual game theory, each player needs to monitor the actions of all other players individually and calculate the probability distribution of joint actions, which can lead to the curse of dimensionality in large-scale distributed systems. To overcome this limitation, this invention introduces a joint strategy virtual game mechanism. This mechanism allows each photovoltaic-storage-charging station to dynamically update its bidding strategy based solely on local forecast data and limited publicly available information from the distribution network, thereby enabling the system to converge to the optimal equilibrium state.

[0070] During the optimization process, the photovoltaic-storage charging station needs to provide its optimal response based on the current electricity price information. From formula (15), the photovoltaic-storage charging station... Individual optimization problem during bidding It is about The problem of concave quadratic maximization. Also... The optimal response of the photovoltaic-storage charging station is given by formula (24): (twenty four) Therefore, in satisfying physical constraints In the case of the first At the end of the round of iteration, the photovoltaic energy storage charging station Best response As shown in formula (25): (25) (ii) Adaptive policy update mechanism based on Mann iteration: For photovoltaic and energy storage charging stations In the Bidding strategy during round iteration The formula is given by the Mann iteration formula, as shown in formula (26): (26) in, Photovoltaic and energy storage charging stations In the The inertia coefficient during round iteration is calculated as shown in formula (27): (27) in, The maximum inertia coefficient, This is the maximum power adjustment range. To balance the judgment threshold, for Time-of-use photovoltaic energy storage charging station The degree of deviation from the equilibrium point is shown in formula (28): (28) in, To prevent constants with a denominator of zero.

[0071] This adaptive policy update mechanism ensures both the global search efficiency in the early stages and the local convergence stability in the later stages. When the system state is far from the equilibrium point, a larger inertia coefficient is used to update the policy, accelerating its exploration of the solution space. When the system state is close to the equilibrium point, adaptive control of the inertia coefficient suppresses oscillations and prevents premature convergence and stagnation of the algorithm update, ensuring that the system converges accurately and smoothly to the equilibrium point.

[0072] In this embodiment, the distributed optimization algorithm based on adaptive joint strategy virtual game in step 5 includes two stages: The first stage is the prediction and initialization stage. Each photovoltaic-storage charging station generates a joint scenario based on local prediction data, performs preliminary optimization to minimize local costs, calculates the expected power purchase, generates an initial bidding strategy to minimize costs, and uploads it to the distribution network. The distribution network calculates the initial dynamic electricity price and the total expected power purchase based on all collected initial bidding strategies. The first stage specifically includes: Step 511: Initialize the dynamic electricity price of the distribution network. Global expected potential function Each photovoltaic energy storage charging station Expected power purchase capacity Bidding for power purchase Inertia coefficient as well as Time-of-use photovoltaic energy storage charging station Degree of deviation from equilibrium point ; Step 512: For each photovoltaic-storage charging station It employs a distributed parallel computing approach to generate joint scenarios based on local prediction data. ; Step 513: Calculate the expected power purchase capacity according to the formula for expected power purchase capacity. The formula for calculating the expected power purchase is as follows:

[0073] Step 514: Generate the initial bidding power with minimal cost. This serves as the initial bidding strategy and is sent to the distribution network. Step 515: The distribution network collects the initial bidding strategies of all photovoltaic-storage charging stations and calculates the initial dynamic electricity price according to the dynamic electricity price calculation formula. And calculate the total expected power purchase. The formula for calculating the dynamic electricity price is as follows: ; The second stage is the bidding and equilibrium iteration stage. The distribution network broadcasts the current dynamic electricity price and total expected power purchase to all photovoltaic-storage charging stations. Each photovoltaic-storage charging station independently calculates its local best response and updates its bidding strategy based on the initial dynamic electricity price and total expected power purchase obtained from the distribution network, and then re-uploads it to the distribution network. The distribution network collects the updated bidding strategies of all photovoltaic-storage charging stations and synchronously updates the dynamic electricity price and global expected potential function until the convergence condition is met. The weighted Bayesian potential game model reaches the optimal equilibrium state, and the optimal bidding strategy is obtained. The second stage specifically includes: Step 521: Enter the iterative loop; the distribution network broadcasts the current dynamic electricity price to all photovoltaic-storage charging stations. Total expected power purchase ; Step 522: For each photovoltaic-storage charging station It employs a distributed parallel computing approach, receiving information broadcast from the distribution network and calculating the optimal response based on the formula. Time-based photovoltaic energy storage charging In the Optimal response during round iteration The formula for calculating the optimal response is as follows:

[0074] Step 523: Calculate according to the formula for calculating the degree of equilibrium deviation. Time-of-use photovoltaic energy storage charging station Degree of deviation from equilibrium point The calculation formula is as follows:

[0075] in, To prevent constants with a denominator of zero; Step 524: Calculate the photovoltaic-storage-charging station based on the inertia coefficient calculation formula. In the Inertia coefficient during round iteration The formula for calculating the inertia coefficient is as follows:

[0076] in, The maximum inertia coefficient, This is the maximum power adjustment range. To balance the judgment threshold; Step 525: Update the formula according to the bidding strategy. Time-based photovoltaic energy storage charging In the Bidding strategy during round iteration The updated bidding strategy is then sent to the distribution network; the bidding strategy update formula is as follows:

[0077] in, for Time-based photovoltaic energy storage charging In the Bidding strategies during round iterations; Step 526: The distribution network collects the updated bidding strategies of all photovoltaic-storage-charging stations and updates the dynamic electricity price according to the dynamic electricity price formula. ; Step 527: Update the global expected potential function according to the calculation formula of the global expected potential function. The formula for calculating the global desired potential function is as follows:

[0078] Step 528: Repeat steps 521 to 527 until the convergence condition is met. The weighted Bayesian potential game model reaches the optimal equilibrium state, and the optimal bidding strategy is obtained.

[0079] The convergence condition is one of the following two conditions: Condition 1: The change in bidding strategies for all photovoltaic-storage-charging stations is less than a preset threshold, i.e. ,in, The preset threshold; Condition 2: Current iteration round Reaching the preset maximum number of iterations ,Right now .

[0080] The pseudocode for the algorithm is shown in Algorithm 1. In the prediction and initialization phase, the photovoltaic-storage charging station generates an uncertainty scenario based on local prediction data and provides an initial bidding strategy (lines 3-7). The distribution network initializes the dynamic electricity price based on the information shared by the photovoltaic-storage charging stations (line 8). In the bidding and equilibrium iteration phase, the distribution network first broadcasts information to the photovoltaic-storage charging stations (line 11). The photovoltaic-storage charging stations then use this information to calculate and update their bidding strategies and upload them to the distribution network (lines 12-17). Finally, the distribution network calculates and updates its dynamic electricity price and global expected potential function based on the bidding strategies of all photovoltaic-storage charging stations (lines 18-19).

[0081]

[0082] The above description is only a part of the embodiments of the present invention and does not limit the scope of protection of the present invention. Any equivalent device or equivalent process transformation made based on the content of the present invention specification and drawings, or direct or indirect application in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for energy scheduling of photovoltaic-storage-charging stations considering uncertainties under incomplete information, characterized in that, Includes the following steps: Step 1: Construct a distributed photovoltaic-storage-charging station system under incomplete information, including a distribution network and multiple photovoltaic-storage-charging stations; the distribution network acts as a market seller entity to publish dynamic electricity prices in real-time scheduling, and the photovoltaic-storage-charging stations are self-interested game entities in real-time scheduling, consisting of a photovoltaic system, an energy storage system, a charging pile system and an energy management system; Step 2: Under the distributed photovoltaic-storage-charging station, the dual uncertainties of photovoltaic output and electric vehicle charging load are quantified by discretizing the prediction scenario; Step 3: Based on the expected utility of the photovoltaic-storage-charging station, establish an individual optimization problem with the goal of maximizing the expected utility of the independent photovoltaic-storage-charging station, and define a global optimization problem for the distributed photovoltaic-storage-charging station system based on the utility function of the distribution network. Step 4: Construct a weighted Bayesian potential game model based on the individual optimization problem and the global optimization problem, and prove the consistency between the individual optimization direction and the global optimization direction to achieve incentive compatibility; Step 5: The weighted Bayesian potential game model is solved using a distributed optimization algorithm based on adaptive joint strategy virtual game. Each photovoltaic-storage charging station only relies on local prediction data and publicly available information from the distribution network to update its bidding strategy. During the bidding strategy update process, an adaptive strategy update mechanism based on Mann iteration is used to iteratively update the bidding strategy of the photovoltaic-storage charging station until it converges to the optimal equilibrium state, thus obtaining the optimal bidding strategy. The distributed optimization algorithm based on adaptive joint strategy virtual game consists of two stages: The first stage is the prediction and initialization stage. Each photovoltaic and energy storage charging station generates a joint scenario based on local prediction data, performs preliminary optimization to minimize local costs, calculates the expected power purchase, generates an initial bidding strategy to minimize costs, and uploads it to the distribution network. The distribution network calculates the initial dynamic electricity price and the total expected power purchase based on all the collected initial bidding strategies. The second stage is the bidding and equilibrium iteration stage. The distribution network broadcasts the current dynamic electricity price and the total expected power purchase to all photovoltaic and energy storage charging stations. Each photovoltaic and energy storage charging station independently calculates its local best response and updates its bidding strategy based on the initial dynamic electricity price and total expected power purchase obtained from the distribution network, and then re-uploads it to the distribution network. The distribution network collects the updated bidding strategies of all photovoltaic and energy storage charging stations, and synchronously updates the dynamic electricity price and the global expected potential function until the convergence condition is met. The weighted Bayesian potential game model reaches the optimal equilibrium state, and the optimal bidding strategy is obtained.

2. The energy dispatching method for photovoltaic-storage-charging stations considering uncertainties under incomplete information as described in claim 1, characterized in that, Step 1 specifically includes: Step 11: Construct a distributed photovoltaic-storage-charging station system under incomplete information conditions. The distributed photovoltaic-storage-charging station system consists of a power distribution network and... Each independently operated and non-communication-connected photovoltaic energy storage charging station constitute, ,in, i Index for photovoltaic-storage charging stations. i =1,2…… n ; This indicates the number of photovoltaic and energy storage charging stations; Indicates the first i One photovoltaic energy storage charging station; Step 12: Divide the total time T of a day into Each period, ,in, t This indicates a time period index.

3. The energy scheduling method for photovoltaic-storage-charging stations considering uncertainties under incomplete information as described in claim 1, characterized in that, Step 2 specifically includes: Step 21: Based on the empirical probability distribution of historical photovoltaic power output data, construct a set of discrete photovoltaic power output scenarios. ,in, Time-of-use photovoltaic energy storage charging station In photovoltaic power output scenarios The actual output power of the photovoltaic system is denoted as: The probability of occurrence is And satisfy ; Step 22: Based on the empirical probability distribution of historical electric vehicle actual charging load data, construct a set of discrete electric vehicle charging demand scenarios. ,in, Time-of-use photovoltaic energy storage charging station In each electric vehicle charging demand scenario The total charging power required by electric vehicles is denoted as The probability of occurrence is And satisfy ; Step 23: Assume a photovoltaic power output scenario Electric vehicle charging demand scenarios Independent of each other, the set of joint scenarios of the two is defined as For joint scenarios Its probability of occurrence is And satisfy .

4. The energy dispatching method for photovoltaic-storage-charging stations considering uncertainties under incomplete information as described in claim 3, characterized in that, The power distribution network acts as a market seller in real-time dispatch, publishing dynamic electricity prices and using price signals to guide the electricity purchasing behavior of each photovoltaic and energy storage charging station. Specifically, it includes: (1) The power distribution network in Dynamic electricity prices released during specific time periods The dynamic electricity price is composed of the benchmark time-of-use tariff and the bidding power purchased by each photovoltaic-storage-charging station. The calculation formula is as follows: in, for The benchmark time-of-use electricity price for the time-of-use distribution network. For the first i One photovoltaic energy storage charging station The elasticity of demand sensitivity coefficient for Before the start of the time period i One photovoltaic energy storage charging station The submitted bid power for electricity purchase; (2) For joint scenarios , Time period i One photovoltaic energy storage charging station Power deviation for: in, for Time period i One photovoltaic energy storage charging station In joint scenarios The actual power purchased; (3) When there is a power deviation in the photovoltaic-storage charging station, additional power balancing costs need to be paid to the distribution network. Power deviation penalty The following are examples: in, for Time period i One photovoltaic energy storage charging station In joint scenarios Power balancing cost under these conditions for Time period i One photovoltaic energy storage charging station In joint scenarios Power deviation penalty For the first i One photovoltaic energy storage charging station Self-defined profit coefficient; j This represents the index of photovoltaic and energy storage charging stations, and i≠j ; s j Indicates the first j One photovoltaic energy storage charging station; for Time period j One photovoltaic energy storage charging station s j In joint scenarios The actual power purchased; The photovoltaic-storage charging station operates as a self-interested game in real-time scheduling, and consists of a photovoltaic system, an energy storage system, a charging pile system, and an energy management system; specifically including: (1) For photovoltaic systems, the energy generated by the photovoltaic system is preferentially used to meet the charging load within the photovoltaic-storage charging station, and any surplus is stored in the energy storage system. If the surplus energy cannot be fully utilized, it will be discarded; specifically including: Time-of-use photovoltaic energy storage charging station In joint scenarios Actual photovoltaic power consumption for: in, for Time-of-use photovoltaic energy storage charging station In photovoltaic power output scenarios The actual output power of the photovoltaic system for Time-of-use photovoltaic energy storage charging station In electric vehicle charging scenarios The total charging power required for electric vehicles for Time-of-use photovoltaic energy storage charging station The charging and discharging power of the energy storage system is negative when it indicates charging and positive when it indicates discharging. for Time-of-use photovoltaic energy storage charging station In joint scenarios The actual power purchased; (2) For the energy storage system, the energy storage system provides real-time energy regulation capability for the photovoltaic-energy storage charging station, reducing the operating cost of the photovoltaic-energy storage charging station; specifically including: a, Time-of-use photovoltaic energy storage charging station The charging and discharging power of medium-energy storage system It is a variable that needs to be decided through a scheduling strategy, and is represented as: in, for Time-of-use photovoltaic energy storage charging station The energy storage charge-discharge coefficient, Photovoltaic and energy storage charging stations Maximum charging power, Photovoltaic and energy storage charging stations Maximum discharge power; b、 Time-of-use photovoltaic energy storage charging station The SOC of the medium energy storage system is The formula is as follows: in, Photovoltaic and energy storage charging stations The upper limit of SOC for medium-sized energy storage systems Photovoltaic and energy storage charging stations The lower limit of medium-sized energy storage systems for Time-of-use photovoltaic energy storage charging station SOC of medium energy storage system Photovoltaic and energy storage charging stations Medium energy storage battery capacity, The length of a time period The energy conversion coefficient of the energy storage system during the charging and discharging process is given by the following formula: in, To improve the charging efficiency of energy storage systems. The discharge efficiency of the energy storage system; c. Time-of-use photovoltaic energy storage charging station Energy storage loss cost Represented as: in, This is the energy storage loss cost coefficient; (3) For the charging pile system, the charging pile system is the main load source in the photovoltaic-storage charging station, and its charging load is affected by the uncertainty of electric vehicle charging demand; specifically including: a, Time-of-use photovoltaic energy storage charging station In joint scenarios Total charging power of electric vehicles for: in, for Time-of-use photovoltaic energy storage charging station In electric vehicle charging scenarios Total charging power demand for electric vehicles in China; b. The electricity purchased by the photovoltaic-storage charging station from the distribution network will generate revenue through sales to electric vehicles. The profit coefficient is hour, Time-of-use photovoltaic energy storage charging station In joint scenarios Comprehensive benefits Represented as: in, The estimated revenue from the current power purchase capacity. The marginal revenue of energy within the photovoltaic-storage charging station; (4) For the energy management system, the energy management system makes energy scheduling decisions with the goal of maximizing the revenue of the photovoltaic energy storage charging station itself; Before the start of the scheduling period, the energy management system assesses the uncertainties of the scenario based on local incomplete information and decides on the power capacity to be purchased through bidding. and participate in the game interaction; During the scheduling period, the energy management system performs actual energy scheduling based on the real-world scenario. Time-of-use photovoltaic energy storage charging station In joint scenarios Actual power purchase capacity for: ； Time-of-use photovoltaic energy storage charging station In joint scenarios The actual utility for: 。 5. The energy scheduling method for photovoltaic-storage-charging stations considering uncertainties under incomplete information as described in claim 4, characterized in that, Step 3 specifically includes: Step 31, Calculation Time-of-use photovoltaic energy storage charging station In the joint scenario set Expected power purchase capacity With expected photovoltaic power absorption capacity The expected power purchase capacity With expected photovoltaic power absorption capacity The calculation formulas are as follows: Step 32, Calculation Time-of-use photovoltaic energy storage charging station In the joint scenario set Expected utility The formula is as follows: ； Step 33, Time-of-use photovoltaic energy storage charging station Individual optimization problem Formal representation: ； in, for Except for the first period before the start of the time period i The bidding power submitted by all other photovoltaic-storage charging stations besides [the one mentioned] The maximum power purchase capacity for the photovoltaic-storage charging station; Step 34: Convert the global desired potential function As the utility function of the distribution network, the formula is as follows: in, Price penalty for power deviation; Step 35: Define the global optimization problem of the distributed photovoltaic energy storage and charging station system. for: 。 6. The energy scheduling method for photovoltaic-storage-charging stations considering uncertainties under incomplete information as described in claim 5, characterized in that, Step 4 specifically includes: Step 41: The weighted Bayesian potential game model is a non-cooperative Bayesian game under incomplete information. For each time period... The weighted Bayesian potential game model is expressed as: The specific definitions of each element are as follows: (1) Player: For players, that is, a collection of independently operating photovoltaic energy storage and charging stations; (2) Strategy: For each player , its in The action strategy for a given time period is to submit the bid-to-purchase power capacity to the distribution network, i.e. The joint strategy space is In this context, the symbol := indicates that it is defined as , the symbol X represents the Cartesian product, and represents the set of combinations of all player strategy spaces; (3) Type: Player Type Used to indicate the uncertainty it faces. Time Period Player The type is defined as The union type space is That is, the Cartesian product of all player types, representing the set of combinations of all player type spaces; (4) Beliefs: To define in the union type space The joint probability distribution on the tectonic scale represents the common prior beliefs of all players about the system state, derived from historical data; (5) Utility: For players Expected utility; Step 42, assuming in Time period, any player Unilaterally changing the strategy Change to Other players' strategies If it remains unchanged, then the change in the player's expected utility under this policy shift is... Represented as: Where · is a placeholder; Meanwhile, the change in the global expected potential function under this strategy offset Represented as: Factoring out the common factor yields the change in the player's expected utility under this policy shift. The change in the global expected potential function The relationship between them is: And due to the profit coefficient ,so Established; Therefore, the individual optimization direction is consistent with the global optimization direction, and the distributed photovoltaic energy storage charging station system has incentive compatibility characteristics.

7. The energy scheduling method for photovoltaic-storage-charging stations considering uncertainties under incomplete information as described in claim 1, characterized in that, The first stage specifically includes: Step 511: Initialize the dynamic electricity price of the distribution network. Global expected potential function Each photovoltaic energy storage charging station Expected power purchase capacity Bidding for power purchase Inertia coefficient as well as Time-of-use photovoltaic energy storage charging station Degree of deviation from equilibrium point ; Step 512: For each photovoltaic-storage charging station It employs a distributed parallel computing approach to generate joint scenarios based on local prediction data. ; Step 513: Calculate the expected power purchase capacity according to the formula for expected power purchase capacity. The formula for calculating the expected power purchase is as follows: Step 514: Generate the initial bidding power with minimal cost. This serves as the initial bidding strategy and is sent to the distribution network. Step 515: The distribution network collects the initial bidding strategies of all photovoltaic-storage charging stations and calculates the initial dynamic electricity price according to the dynamic electricity price calculation formula. And calculate the total expected power purchase. The formula for calculating the dynamic electricity price is as follows: 。 8. The energy dispatching method for photovoltaic-storage-charging stations considering uncertainties under incomplete information as described in claim 1, characterized in that, The second stage specifically includes: Step 521: Enter the iterative loop. The distribution network broadcasts the current dynamic electricity price to all photovoltaic and energy storage charging stations. Total expected power purchase ; Step 522: For each photovoltaic-storage charging station It employs a distributed parallel computing approach, receiving information broadcast from the distribution network and calculating the optimal response based on the formula. Time-based photovoltaic energy storage charging In the Optimal response during round iteration The formula for calculating the optimal response is as follows: Step 523: Calculate according to the formula for calculating the degree of equilibrium deviation. Time-of-use photovoltaic energy storage charging station Degree of deviation from the equilibrium point The calculation formula is as follows: in, To prevent constants with a denominator of zero; Step 524: Calculate the photovoltaic-storage-charging station based on the inertia coefficient calculation formula. In the Inertia coefficient during round iteration The formula for calculating the inertia coefficient is as follows: in, The maximum inertia coefficient, This is the maximum power adjustment range. To balance the judgment threshold; Step 525: Update the formula according to the bidding strategy. Time-based photovoltaic energy storage charging In the Bidding strategy during round iteration The updated bidding strategy will then be sent to the distribution network; the formula for updating the bidding strategy is as follows: in, for Time-based photovoltaic energy storage charging In the Bidding strategies during round iterations; Step 526: The distribution network collects the updated bidding strategies of all photovoltaic-storage-charging stations and updates the dynamic electricity price according to the dynamic electricity price formula. ; Step 527: Update the global expected potential function according to the calculation formula of the global expected potential function. The formula for calculating the global desired potential function is as follows: Step 528: Repeat steps 521 to 527 until the convergence condition is met. The weighted Bayesian potential game model reaches the optimal equilibrium state, and the optimal bidding strategy is obtained.

9. The energy scheduling method for photovoltaic-storage-charging stations considering uncertainties under incomplete information as described in claim 8, characterized in that, The convergence condition is one of the following two conditions: Condition 1: The change in bidding strategies for all photovoltaic-storage-charging stations is less than a preset threshold, i.e. ,in, The preset threshold; Condition 2: Current iteration round Reaching the preset maximum number of iterations ,Right now .

Images