Model-free learning control for processes with constrained incremental control

This paper proposes a technique to design controllers for systems with constrained incremental control and input-output constraints called Model-Free Learning Control (MFLC). MFLC, which is based on Reinforcement Learning algorithms, is a simple approach without needing precise detailed information of the system. MFLC is proposed for process control, which in practical problems exhibit constraints. As a simple example, the controller is designed and tested for a two-tank system. Simulation results show that the MFLC controller learns to adequately control the process.