Real time investigation of nonlinear controllers for a conical tank process

Abstract

In chemical process industries, many demanding control problems arise due to nonlinear dynamics, time varying process parameters, annoyance of constraints on the manipulated variable, interaction between the process, dead time, and unmeasured disturbance. Efficient operation of industrial process would require control of enormous process variables within their specified bounds. The Proportional Integral Derivative (PID) Controller has been widely used in the process industries for several decades, due to its simple structure and ease of deployment. Standard tuning methods include Zeigler Nichol s (ZN) closed loop method and Cohen Coon s (CC) open loop method. However control of process variables in a nonlinear process using PID Controller needs fine tuning until satisfactory performance is obtained. A conical bottom tank process is widely used in many of the process industries and level control in such process is a challenging task due to its varying cross section and nonlinear dynamics.The conical tank process has rigorous nonlinearity due to varying process parameters. In this work, empirical modeling of variable tank process has been carried out for design of Direct Synthesis (DS) based Proportional Integral (PI) controller and demonstrated in real time experimental set up. The drawbacks of DS based PI controller is that, the controller design solely depends on process model and it cannot adapt to process behavior when there is a change in process dynamics. Finally, the limitations of the DS based PI controller for the selected process and need for nonlinear control schemes are brought out.Adaptive control based on reference model, is called as MRAS is designed and demonstrated in the conical tank process in the real time experimental set up. Its behavior is specified by the model and the parameters of the controller are adjusted based on error. newline newline newline