新能源系统配电网络自适应优化方法  

Adaptive optimization method of distribution network in new energy system

作　　者：周博滔 刘政 饶尧 ZHOU Bo-tao;LIU Zheng;RAO Yao(State Grid Electric Power Research Institute Wuhan Efficiency Evaluation Company Limited,Wuhan 430074,China)

机构地区：[1]国网电力科学研究院武汉能效测评有限公司,武汉430074

出　　处：《信息技术》2024年第3期140-145,共6页Information Technology

摘　　要：以解决新能源系统配电网络网损过高的问题为目标,研究基于深度强化学习的新能源系统配电网络自适应优化方法。设置线路电流与节点电压约束、分布式电源出力约束、潮流约束三项约束条件,建立新能源系统配电网络自适应优化的目标函数。利用马尔可夫决策模型,简化新能源系统配电网络自适应优化目标函数求解过程,选取深度确定性策略梯度算法求解完成转化后的自适应优化目标函数,完成新能源系统配电网络自适应优化。实验结果表明,该方法可以降低新能源系统配电网络的设备动作成本,降低运行网损。In order to solve the problem of high network loss of distribution network in new energy system,an adaptive optimization method of distribution network in new energy system based on deep reinforcement learning is studied.Three constraints,including line current and node voltage constraints,distributed generation output constraints and power flow constraints are set,and the objective function of adaptive optimization of distribution network of new energy system is established.The Markov decision model is used to simplify the solution process of the adaptive optimization objective function of the distribution network of the new energy system,and the depth deterministic strategy gradient algorithm is selected to solve the transformed adaptive optimization objective function to complete the adaptive optimization of the distribution network of the new energy system.The experiment results show that this method can reduce the equipment operation cost and operation network loss of the distribution network of new energy system.

关 键 词：深度强化学习 新能源系统 配电网络 自适应 目标函数 

分 类 号：TM712[电气工程—电力系统及自动化]