Reinforcement Learning based MPPT Control for Induction Motor Fed PV Array

Authors

  • Ashutosh Yadav, Archana

Abstract

Solar irradiation is a renewable energy source with promising power generation capacity and inexpensive. It becomes an important source of energy in isolated areas where the main grid supply is not accessible. This scenario is very evident in developing countries like India. In this article, we studied the PV array connected with a load of water pump operated with a 3-phase induction motor. The efficiency of the PV array is articulated through an efficient tracking of maximum power points. The most challenging part is to design a model that can track the maximum point irrespective of variations in environmental conditions and its parametric variations. The application of Deep Q-learning makes the model parametric free, and once the model is trained can be implanted in a different scenario and run effectively. The designing of action, state, and reward space verbalizes the success of the RL algorithm. The trained agent has achieved the maximum power and has shown improved performance up to 73% under changing environmental conditions.

Keywords-Solar energy; MPPT control; Reinforcement learning; Deep Q-Learning.

Published

2021-04-12

Issue

Section

Articles