CN-121981575-A - Optical storage charging station optimization operation method considering cooperation of ladder-type carbon transaction and multiple bodies

CN121981575ACN 121981575 ACN121981575 ACN 121981575ACN-121981575-A

Abstract

The invention relates to the technical field of optical storage charging stations and discloses an optical storage charging station optimizing operation method considering cooperation of ladder-type carbon transaction and multiple bodies, which comprises the steps of collecting carbon emission data of all bodies in a charging station, and establishing a ladder-type carbon transaction mechanism considering interaction between the charging station and a power grid; the method comprises the steps of comprehensively considering interest requirements and operation safety of each main body, establishing a comprehensive optimization objective function with the aim of maximizing comprehensive benefits of the charging station, abstracting schedulable resources in the charging station into intelligent bodies, and constructing a multi-intelligent body reinforcement learning framework to solve the comprehensive optimization objective function in cooperation with a carbon transaction mechanism to obtain an optimal operation scheduling strategy in cooperation of multiple main bodies. According to the invention, by establishing the nonlinear ladder-type carbon transaction mechanism close to the real carbon transaction market, the optimal scheduling strategy obtained based on the carbon transaction mechanism is prospective, the benefits of a power grid and an optical storage station can be balanced, and the overall market competitiveness and risk resistance of the power system are effectively improved.

Inventors

CHENG LANFANG
JI XIAOLU
CHEN XIAODONG
WEI XULIANG
WEN TAO
DENG FENGLIN
ZHANG KEJIAN
DU CHENGBIN
XU PENGFEI
HU CHENGJIE

Assignees

国网安徽省电力有限公司宣城供电公司

Dates

Publication Date: 20260505
Application Date: 20260126

Claims (10)

1. The optimal operation method of the optical storage charging station considering the cooperation of the stepped carbon transaction and the multiple bodies is characterized by comprising the following steps: Collecting carbon emission data of each main body in the charging station, and establishing a ladder-type carbon transaction mechanism by considering interaction between the charging station and a power grid; Comprehensively considering interest requirements and operation safety of each main body, and establishing a comprehensive optimization objective function with the aim of maximizing comprehensive benefits of the charging station; Abstracting schedulable resources in the charging station into intelligent agents, and constructing a multi-intelligent-agent reinforcement learning framework to solve the comprehensive optimization objective function in cooperation with the carbon transaction mechanism to obtain an optimal operation scheduling strategy of multi-main-body cooperation.
2. The method of optimizing operation of a photovoltaic charging station in view of ladder-type carbon transaction in conjunction with multiple bodies of claim 1, wherein each body within the charging station comprises a photovoltaic power generation system, an energy storage system, and an electric vehicle.
3. The method for optimizing operation of an optical storage charging station in view of ladder-type carbon transaction and multi-subject collaboration of claim 1, wherein the carbon emission data of each subject is obtained by constructing each subject operation model in the charging station, comprising: Taking the influence of illumination intensity into consideration, establishing a photovoltaic output model of a photovoltaic power generation system, and predicting photovoltaic output; An energy storage system operation model is built, and a state of charge constraint and a charge and discharge power constraint are built based on a state of charge equation of the energy storage system operation; and (3) constructing an electric vehicle charging and discharging behavior model, and constructing electric vehicle charging and discharging power constraint and state of charge boundary and user satisfaction constraint based on a state of charge equation in the electric vehicle charging and discharging process.
4. The method for optimizing operation of a light storage charging station in view of ladder-type carbon transactions in conjunction with multiple bodies of claim 1, the method is characterized in that the step-type carbon transaction mechanism is established, and comprises the following steps: Calculating the actual carbon emission of the optical storage station to obtain the residual carbon emission quota by considering the interaction of the optical storage station and the power grid; Based on the residual carbon emission quota, a segmented ladder pricing mechanism is adopted to establish a ladder-type carbon transaction mechanism for stimulating low-carbon behaviors; the step-type carbon transaction mechanism is the step-type carbon transaction cost which is set from carbon superrank punishment to carbon low-rank rewarding layer by layer.
5. The optimal operation method for the optical storage charging station considering the cooperation of the ladder-type carbon transaction and the multi-main body according to claim 1, wherein the comprehensive benefit of the charging station is the difference between the real-time operation benefit of the optical storage charging station and the equipment operation and maintenance cost, the total carbon transaction cost and the idle punishment of the electric automobile total charging pile; The real-time operation income is the difference between electricity selling income, electricity purchasing cost and carbon transaction cost of the optical storage station, the equipment operation and maintenance cost comprises operation and maintenance costs of a photovoltaic power generation system and an energy storage system, and the total carbon transaction cost is obtained based on the stepped carbon transaction cost.
6. The method of claim 1, wherein constructing a multi-agent reinforcement learning framework to solve the comprehensive optimization objective function comprises: Abstracting schedulable resources into agents respectively, and defining a state space containing key environment information and an action space containing an agent action set respectively, wherein the key environment information is environment information influencing resource scheduling decisions; Based on the comprehensive optimization objective function, layering and designing a ladder type reward function; Based on the intelligent agent, the state space and the action space, according to the rewarding function, adopting MATD algorithm to solve and obtain an optimal operation scheduling strategy; The abstracting the schedulable resources into the agents respectively includes: abstracting the energy storage system with schedulable resources as an energy storage system intelligent agent; Abstracting the electric automobile charging pile cluster with schedulable resources into a charging pile cluster intelligent body.
7. The method of optimizing operation of a light storage charging station in view of ladder carbon transaction and multi-principal collaboration of claim 6, wherein the optimal operation scheduling strategy comprises: Based on the global state space and the action space, the optimal global regulation strategy obtained by centralized training and And acquiring the acquired respective scattered executed regulation actions in the respective action space of the intelligent agent according to the optimal global regulation strategy.
8. The optimal operation method for the optical storage charging station considering the cooperation of the ladder-type carbon transaction and the multi-main body according to claim 6, wherein the reward function comprises a global basic reward, an agent special reward and a constraint violation penalty which are designed in sequence; and calculating the final rewards of the obtained intelligent agents according to the weight based on the global basic rewards, the special rewards of the intelligent agents and the constraint violation punishment.
9. The method for optimizing operation of a light storage charging station in consideration of ladder-type carbon transaction and multi-body synergy as claimed in any one of claims 6-8, wherein said adopting MATD algorithm solution to obtain optimal operation scheduling policy comprises: configuring a strategy network, a double commentator network and a target network for each agent, and establishing an experience playback pool for storing state transition samples; Randomly sampling in batches from the experience playback pool, generating a smooth target action with noise based on the strategy network, and calculating a target Q value of each intelligent agent; updating the network parameters of double critics by taking the error between the minimized predicted Q value and the target Q value as a target, and updating the network parameters of the strategy by taking the maximized target Q value as a target delay; the target network tracks parameters of a corresponding strategy network and a double criticism network through soft update, and iterative training is carried out until convergence, so that an optimal global regulation strategy is obtained; based on the optimal global regulation strategy, collecting the local observation state of each intelligent agent, and respectively inputting the respective trained strategy network forward propagation to obtain the regulation action of each intelligent agent in a scattered way.
10. The method for optimizing operation of an optical storage station taking into account ladder-type carbon transaction and multi-body cooperation as claimed in claim 9, wherein the state transition samples are generated by interaction of each agent with the environment respectively, and the delay updating of strategy network parameters is based on gradient delay updating of double criticism network output.

Description

Optical storage charging station optimization operation method considering cooperation of ladder-type carbon transaction and multiple bodies Technical Field The invention relates to the technical field of optical storage charging stations, in particular to an optical storage charging station optimizing operation method considering ladder-type carbon transaction and multi-main-body cooperation. Background The light storage charging station is used as a typical comprehensive energy unit for integrating a distributed photovoltaic energy storage system and an electric vehicle charging load, and the construction and the operation of the light storage charging station are increasingly wide, so that the light storage charging station becomes a key node for constructing a novel electric power system. However, the current operational optimization of light storage charging stations still has the following drawbacks: (1) The double uncertainty superposition exacerbates the running risk that the photovoltaic output is affected by weather and has intermittence and volatility, and the charging behavior of the electric automobile user has high randomness, so that the strong uncertainty of the source-charge double sides brings serious challenges to the power balance, economic dispatch, safety and stability of the optical storage charging station; (2) The multi-main-body conflict of interests is difficult to coordinate, namely, an operator main body of an optical storage station in the optical storage station pursues maximization of operation income and minimization of low-carbon cost, an electric automobile user pursues minimum-cost quick charging and highest travel satisfaction, and a power grid needs to ensure stability and high efficiency, so that natural conflict exists in targets of all main bodies, and the traditional centralized optimization method is difficult to effectively describe game and interaction relations between the targets; (3) The market incentive mechanism is not sound, the existing research is mostly focused on economic dispatch in the electric power market environment, or only carbon emission is converted into fixed cost, the deep integration of the optical storage station into the carbon trade market is not realized, and the economic incentive effect of the carbon trade mechanism cannot be fully utilized to guide low-carbon operation; (4) The optimization method is conservative or idealized, aiming at double uncertainties, the traditional random optimization severely depends on an accurate probability distribution model, the probability distribution model is difficult to obtain in practice, and the robust optimization can guarantee the feasibility in the worst scene, but usually at the cost of economy, the strategy is too conservative. In view of the above shortcomings, there is a need for an intelligent optimal operation method that can deeply fuse carbon transaction mechanisms, coordinate multi-body interests, and effectively cope with source load uncertainty to promote high quality development of optical storage stations. Disclosure of Invention In order to overcome the defects in the prior art, the embodiment of the invention provides an optimal operation method of the optical storage charging station considering the cooperation of stepped carbon transaction and multiple bodies, and solves the problems of how to realize the cooperation optimization of multiple agents and realize the economic, low-carbon and efficient operation of the optical storage charging station under the conflict of double uncertainty and multi-body interests. In order to achieve the above purpose, the embodiment of the invention provides an optical storage charging station optimizing operation method considering the cooperation of ladder-type carbon transaction and multiple bodies, which comprises the steps of collecting carbon emission data of all bodies in the charging station, and establishing a ladder-type carbon transaction mechanism considering the interaction of the charging station and a power grid; Comprehensively considering interest requirements and operation safety of each main body, and establishing a comprehensive optimization objective function with the aim of maximizing comprehensive benefits of the charging station; Abstracting schedulable resources in the charging station into intelligent agents, and constructing a multi-intelligent-agent reinforcement learning framework to solve the comprehensive optimization objective function in cooperation with the carbon transaction mechanism to obtain an optimal operation scheduling strategy of multi-main-body cooperation. In a preferred embodiment, each body within the charging station includes a photovoltaic power generation system, an energy storage system, and an electric vehicle. In a preferred embodiment, the carbon emission data of each subject is obtained by constructing each subject operation model in the charging station, comprising: Taking the influence o