Day-Ahead Optimizing Dispatch of the Renewable Energy Systems Based on A Deep Q- Network Approach and Demand Response Strategies

With the growing integration of renewable energy sources, power systems are encountering substantial and rapid variations in power and voltage profiles on a daily basis. An effective and precise dispatch strategy is essential for maintaining system stability and optimizing economic performance. Based on this, an advanced deep reinforcement learning algorithms are proposed for making demand response decisions. The demand response can enhance dispatch decisions by optimizing load and price control. Additionally, the approach can significantly aid the power grid in making rapid and effective decisions based on day-ahead dispatch. Furthermore, the Security-Constrained Units Commitment (SCUC) and the Deep Q-Network (DQN) technique are considered in the proposed model. The method was verified on a modified IEEE 24-bus system. The findings illustrate the approach's effectiveness, emphasizing its potential to aid power grid operators in making informed dispatch decisions.