Online Self Optimizing Posicast Controller Framework for Smart Process Control

Posicast control is a strategy employed for overshoot compensation of lightly damped systems. Its classical approach is mainly used in a feedforward setup to compensate for the system's initial oscillation. However, if the system has a parametric uncertainty and high oscillation, the Posicast control effectiveness is reduced. This paper presents the design of a Self Optimizing Posicast controller with an online optimizing agent to effectively mitigate the oscillatory behaviour and overshoot on highly oscillatory systems. The controller comprises of a closed-loop setup regulator, a posicast filter, and an online optimizer based on the Globalized Constrained N elder- Mead optimization algorithm to perform real-time estimation of the controller and posicast filter parameters. The proposed framework improves the system's robustness under parametric uncertainty and is validated for a flyback converter on voltage regulation tasks.