Multi-stage formulation of Optimal Distributed Generation Placement using Reinforcement Learning

The advancements in Distributed Generation Technology and the environmental concerns has given a way for adopting more and more Distributed Generation(DG) sources into the power system. The benefits from the Distributed Generation can be guaranteed only if they are integrated at appropriate locations with optimal capacity. This was one of the research interest in the past decade and numerous studies were done on the same. The uncertainty associated with the DG sources is a keen aspect to be considered for the integration. This necessitates the application an optimization algorithm that can handle uncertainties. The paper presents the formulation of Optimal DG placement problem as a multi-stage decision making problem which is solved using Reinforcement Learning that can be applied for handling uncertainties. The proposed algorithm is validated in IEEE-33 Bus Radial distribution System.