Autonomous Decentralized Control of Distribution Network Voltage using Reinforcement Learning

Recently, power flow patterns in distribution network has considerably modified due to a large amount of PV installation and increase of consumers behavior patterns. Hence, a system operator has become difficult to control voltage by the conventional method. As a countermeasure of this problem, the reactive power control method with the PV inverters has been reported. In this paper, we proposed a new voltage control method with reactive control of PV inverters using reinforcement learning. In particular, the purpose of proposed method is to bring out the cooperative reactive power control without the information between PV inverters. We verify the effectiveness of proposed method through computational simulation using distribution network model. (C) 2018, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.