Optimal integration of photo voltaic sources in unbalanced distribution system using Reinforcement Learning

The thrive for environment friendly sources of energy to meet the growing energy demand has driven the integration of more Distributed Generation(DG) sources into the Distribution network. In order to achieve the benefits of DG integration in terms of improvement in voltage profile, minimization of losses etc, optimal placement of DG sources is of much significance. This is to be done by using robust optimization techniques that can handle the uncertainty associated with the DG sources. Therefore the optimal placement of DG in unbalanced distribution network is a challenging issue. The paper presents the application of Reinforcement Learning (RL) for optimally allocating the Photo Voltaic (PV) units in an unbalanced distribution network. The proposed algorithm is validated for the unbalanced IEEE 13 bus distribution network.