Optimal Actuation of Controller using Predictive PI for Nonlinear Level Process

Abstract

Background/Objective: Energy consumption of the actuator illustrates the performance of controller on the process. Objective of this paper is to prove how predictive control approach minimizes the energy utilization of actuator in nonlinear process (conical tank level process). Method/Statistical Analysis: Conical tank is divided into three operating regions and model identification using two-point method has been carried out at oneselected operating region. Efficiency of controller is validated by subjecting it with the identified model at all operating regions. Prediction filter which acts as dead time compensation was introduced in cascade with controller. Findings: Simulation and experimental results validates the approach. The effectiveness of the design has been compared with the advanced control schemes. Application and Improvements: This methodology can be implemented in the nonlinear processes with dead time. It can be applicable for the processes where energy consumption of actuator is considered as key objective. Keywords: FOPDT, Nonlinear Process, Predictive PI, SIMC