Prediction Model-Assisted Optimization Scheduling Strategy for Renewable Energy in the Microgrid

Abstract

As the global reliance on renewable energy sources grows, wind and photovoltaic power, as pivotal components, pose significant challenges to power system dispatch due to their volatility and uncertainty. To effectively address this challenge, this paper proposes a renewable energy optimization dispatch strategy based on a prediction model. First, this paper constructs a prediction model combining functional data analysis and recurrent neural networks (RNNs) to achieve an accurate prediction of renewable energy output. On this basis, considering the economic and environmental benefits of system operation, an optimal multi-objective dispatch model for renewable energy is established, and the multi-objective optimization problem is transformed into a single-objective optimization problem using weighting methods to reduce the complexity of the solution. Finally, a typical microgrid test system is used to verify the effectiveness and feasibility of the proposed method. The results of the numerical example show that the proposed model can achieve an accurate prediction of renewable energy sources, reduce the conservatism of traditional dispatch decisions, and balance economic and environmental benefits.