Intelligent Voltage Control Method in Active Distribution Networks Based on Averaged Weighted Double Deep Q-network Algorithm

High penetration of distributed renewable energy sources and electric vehicles (EVs) makes future active distribution network (ADN) highly variable. These characteristics put great challenges to traditional voltage control methods. Voltage based on the deep Q-network (DQN) algorithm offers a potential solution this problem because it possesses human-level performance. However, DQN methods may produce overestimation action reward values, resulting in degradation obtained solutions. In paper, an intelligent method averaged weighted double (AWDDQN) is proposed overcome shortcomings values underestimation (DDQN) algorithm. Using method, objective incorporated into designed normalized form Markov decision process (MDP) model which solved by AWDDQN The AWDDQN-based agent trained offline used as online dynamic regulator for ADN. validated using IEEE 33-bus 123-bus systems containing EVs, compared with DDQN algorithms methods, mixed-integer nonlinear program simulation results show that has better convergence less volatility than other ones.