Reinforcement Learning Based PID Control of Wind Energy Conversion Systems

In this paper an adaptive PID controller for Wind Energy Conversion Systems (WECS) has been developed. The adaptation technique applied to this controller is based on Reinforcement Learning (RL) theory. Nonlinear characteristics of wind variations as plant input, wind turbine structure and generator operational behavior demand for high quality adaptive controller to ensure both robust stability and safe performance. Thus, a reinforcement learning algorithm is used for online tuning of PID coefficients in order to enhance closed loop system performance. In this study, at start the proposed controller is applied to two pure mathematical plants, and then the closed loop WECS behavior is discussed in the presence of a major disturbance.