MSc. Thesis Defense:Arda Ağababaoğlu14-06-2019
ON-LINE BAYESIAN REINFORCEMENT LEARNING WITH MCMC TO MAXIMIZE ENERGY OUTPUT OF VERTICAL AXIS WIND TURBINE
Mechatronics Engineering, MSc. Thesis, 2019
Assoc. Prof Ahmet Onat(Thesis Advisor), Prof. Dr Serhat Yeşilyurt,
Prof. Dr. Müjde Güzelkaya
Date & Time: 20th, June 2019 – 14 PM
Place: FENS G025
Keywords: Reinforcement Learning, Markov Chain Monte Carlo, Radial Basis Function Neural Network, Wind Energy Conversation Systems, Vertical Axis Wind Turbines
Optimization of energy output of small scale wind turbines requires a controller which keeps the wind speed to rotor tip speed ratio at the optimum value. An analytic solution can be obtained if the dynamic model of the complete system is known and wind speed can be anticipated.
However, aging and errors in modeling and wind speed prediction prevent a straightforward solution.
This thesis proposes to apply a reinforcement learning approach designed to optimize dynamic systems with continuous state and action spaces, to the energy output optimization of Vertical Axis Wind Turbines (WAVT). The dynamic modeling and load control of the wind turbine are
accomplished in the same process. The proposed algorithm is an on-line model-free Bayesian Reinforcement Learning using Markov Chain Monte Carlo method (MCMC) to obtain the parameters of an optimal policy which is implemented as a Radial Basis Function Neural Network (RBFNN).
The proposed method learns wind speed profiles, and therefore is able to utilize all system states and observed wind speed profiles to calculate an optimal control signal. The proposed method is validated by performing simulation studies for a permanent magnet synchronous generator-based VAWT Simulink model for comparing with classical maximum power point tracking algorithm. The results show significant improvement over the classical method, especially during the wind speed transients, promising a superior energy output in turbulent settings; which coincide
with the expected application areas of WAVTs.