Research on Optimization Algorithm and Decision System for Autonomous Operation of New Power System

This paper focuses on the autonomous operation optimization algorithm and decision-making system of new power system (NPS), and proposes an optimization algorithm framework based on reinforcement learning (RL) to improve the operation efficiency and security of power system. In this paper, the framework of autonomous operation optimization algorithm based on RL is constructed, the power system is abstracted as markov decision processes (MDP), and an agent is constructed through deep neural network (DNN) to learn the optimal strategy. The algorithm framework can sense the state of power system in real time, realize the optimal control decision through the learning of agents, and comprehensively consider economy, safety and new energy consumption ability. In the simulation experiment, this paper uses MATLAB/Simulink platform to build NPS model, which verifies the remarkable advantages of RL algorithm in reducing system operation cost and improving decision efficiency. Compared with traditional linear programming and dynamic programming algorithms, RL algorithm performs better in cost reduction rate and decision-making time, with the cost reduction rate as high as 15.24% and the distribution of decision-making time more centralized and stable. In addition, this paper also discusses the evaluation and analysis method of key technologies based on patent indicators, identifies the "stuck-neck" technology in the power industry, and designs a decision support system (DSS) architecture for NPS. DSS architecture includes four core modules: data acquisition and processing, RL algorithm, decision analysis and user interaction. Through modular design, information can be transmitted and shared efficiently, which provides a strong guarantee for intelligent optimization and decision support of power system. The research results of this paper provide new ideas and technical support for the autonomous operation optimization of NPS, which is of great significance to promoting energy transformation and achieving the goals of peak carbon dioxide emissions and carbon neutrality.