Integrating DDPG and QPSO for Multi-Objective Optimization in High Proportion Renewable Energy Power Dispatch Systems

Abstract

This study proposes a novel dispatch optimization model that integrates deep deterministic policy gradient (DDPG) and quantum particle swarm optimization (QPSO) to address the challenges posed by high proportions of renewable energy in power systems. The proposed multi-objective optimization framework considers system cost reduction, supply-demand balance, and dynamic adaptability to renewable energy fluctuations. The experimental results on the IEEE 30-bus and 118-bus systems demonstrated significant improvements. This method reduced total system costs by 13.6% and 11.4%, respectively. It also increased supply reliability to 97.1% and achieved an energy utilization rate of 94.85%. Additionally, it minimized frequency deviation to 1.25 Hz. The optimization time was also improved, achieving a reduction of 58.3 seconds in efficiency. The research results have important practical application value in improving power system economy, enhancing system reliability, and dynamic adaptability. It can provide efficient and reliable technical support for power dispatch planning, load management, and real-time control under high percentage renewable energy scenarios.