End-to-end electric energy transaction market user decision-making method based on reinforcement learning

Abstract

The invention discloses an end-to-end electric energy transaction market user decision-making method based on reinforcement learning, and the method comprises the steps: building an end-to-end electric energy transaction market model from the perspectives of a transaction main body model, an equipment model, a transaction electricity price model and a physical constraint model; establishing a Markov decision process model of an end-to-end electric energy transaction process; and adopting a reinforcement learning algorithm based on energy storage action uniform discrete processing to analyze and solve the user participation end-to-end electric energy transaction market decision problem. According to the end-to-end electric energy transaction market user decision-making method based on reinforcement learning provided by the invention, the user can autonomously learn and update the own optimal action, the transaction can also be carried out under the condition that the market mechanism isimperfect and only the local operation information of the user is known, and complex optimization calculation does not need to be carried out, therefore, smooth development of end-to-end electric energy transaction is promoted.

Description

technical field [0001] The invention relates to the field of user decision-making in an electric power market, in particular to a method for user decision-making in an end-to-end electric energy trading market based on reinforcement learning. Background technique [0002] In recent years, with the widespread use of distributed power sources such as photovoltaic equipment and energy storage on the user side, the power network is undergoing a fundamental transformation, and traditional passive energy consumption users are transforming into "production and consumption users". At present, these user-side resources are usually purchased by grid companies at fixed electricity prices or premiums in accordance with the new energy priority acquisition policy stipulated by law. However, at present, the purchase cost of distributed power and energy storage is still high. In the context of the gradual cancellation of new energy subsidy measures, in order to promote the development of ne...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art, and provide a user decision-making method based on reinforcement learning in the end-to-end electric energy trading market, which is used to improve the decision-making mechanism for users to participate in the distributed photovoltaic end-to-end trading market, and to solve the problems on the user side. End-to-end energy trading problem in electricity market

Images