Comparative analysis of machine learning methods for active flow control

Machine learning frameworks such as Genetic Programming (GP) and Reinforcement Learning (RL) are gaining popularity in flow control. This work presents a comparative analysis of the two, bench-marking some their most representative algorithms against global optimization techniques Bayesian Optimization (BO) Lipschitz (LIPO). First, we review general framework model-free control problem, bringing together all methods black-box problems. Then, test on three cases. These (1) stabilization nonlinear dynamical system featuring frequency cross-talk, (2) wave cancellation from Burgers' (3) drag reduction cylinder wake flow. We present comprehensive comparison to illustrate differences exploration versus exploitation balance between `model capacity' law definition `required complexity'. believe that paves way toward hybridization various methods, offer perspective future development literature