Multi Objective Optimization of Multi Component Isothermal Liquid-Phase Kinetic Sequence using Multivariable PI Control

Samira Zamiri,Hossein Baloochian,Saeed Balochian

doi:10.11591/eei.v3i4.293

Abstract

In this paper, an optimal tuned saturated PI type controller with anti-windup structure is used for process control. In first step, a single objective genetic algorithm is used to find the optimal values of controller parameters. To show the difference between optimal and non-optimal control, we use this controller to track the square pulse. The results show that by choosing the control parameters randomly the output cannot track the reference signal but by optimizing the control parameters, the error, and settling time decreases significantly and efficiency of control increases but it needs more control effort. To find the optimal control parameters with lower control input, a multi objective genetic algorithm is used in next step and three points in Pareto front are analysed. It is shown that this method increases the control efficiency and needs lower control input than obtained by single objective genetic algorithm.

Highlights

In the recent years, Control of chemical reaction procedure attract many researchers and they used different control methods to control the nonlinear manner of chemical reactions such as including feed-back linearization, sliding mode control, adaptive/neural control and nonlinear model predictive control [1], [2], [3], [4], [5]
The results showed that optimizing the control parameters improve the control efficiency and reduces the error settling time in both cases
We use single objective genetic algorithm to optimize the control parameters and Error mean square is used as objective function

Summary

Introduction

Control of chemical reaction procedure attract many researchers and they used different control methods to control the nonlinear manner of chemical reactions such as including feed-back linearization, sliding mode control, adaptive/neural control and nonlinear model predictive control [1], [2], [3], [4], [5]. A genetic algorithm (GA) is an optimization technique that looks for the solution of the optimization problem, imitating species evolutionary mechanism [8], [9], [10], [11], [12]. In this type of algorithms, a set of individuals (so-called population) changes generation by generation (evolution) adapting better to the environment. The results showed that optimizing the control parameters improve the control efficiency and reduces the error settling time in both cases. This paper is organized as follow: in section 2 the mathematical modeling of multi component isothermal liquid-phase kinetic sequence carried out in a continuous stirredtank reactor (CSTR) is presented. One might view as tuning parameters and reduce it gradually until the transient response under partial state feedback is close enough to the ideal SMC. k0 , k1 and k 2 are control parameters and in this paper we will optimize to increase controlling efficiency

Optimization and Simulation

Conclusion