Conference Proceedings
The International Conference on Emerging Technologies in Electronics, Computing and Communication 2022
(ICETECC`22)
Control and Coordination of Multiple PV Inverters in Power Distribution Network using Multi Agent Deep Reinforcement Learning
Anis ur Rehman1*; Muhammad Ali1; Sheeraz Iqbal1; Syed Danish Ali1; Aqib Shafiq1; Raja Tahir Iqbal1;| 1Department of Electrical Engineering, University of Azad Jammu and Kashmir, Muzaffarabad 13100, AJK, Pakistan |
ABSTRACT
The growing demand of power can be realized with the increased penetration of PVs in the power distribution network (PDN). Moreover, low-cost energy with less emission of polluted gases can be achieved. Along with these advantages, it has some disadvantages as well. The integration of high number of PVs in PDN causes voltage deviation, which is undesirable. Real-time control and coordination among the agents (PVs) are required to minimize the voltage deviation and to maintain the voltage in a certain range. This real-time control and coordination are achieved through a multi-agent deep reinforcement learning algorithm. Each PV inverter is considered an agent and its action can be divided into actor and critic network. Actor-network of each PV-inverter produces or absorbs reactive power according to the requirement. The critic network evaluates the performance of the actor-network and produces a Q-value according to the action. Each agent tries to maximize its Q-value. Moreover, all the agents are arranged in distributed and decentralized scheme to achieve real-time coordination among them. The proposed framework is tested on the PV-integrated IEEE-33 bus system. Reactive power control of all the PVs collectively maintains the voltage in a certain range of ±5%.