Sub-optimal tracking in switched systems with fixed final time and fixed mode sequence using reinforcement learning
Approximate dynamic programming is used to solve optimal tracking problems in switched systems with controlled subsystems and fixed mode sequence. Two feedback control solutions are generated such that the system tracks a desired reference signal, and the optimal switching instants are sought. Simulation results are provided to illustrate the effectiveness of the solutions.
Sardarmehni, Tohid, and Xingyong Song. "Sub-optimal tracking in switched systems with fixed final time and fixed mode sequence using reinforcement learning." Neurocomputing 420 (2021): 197-209. https://doi.org/10.1016/j.neucom.2020.09.011