Particle Swarm Optimization based PID controller for Extrusion Process

Abstract

The process called extrusion was used for making objects of fixed cross-sectional profile by pushing or pulling the material through the die of the preferred cross-section. The extrusion process was a nonlinear process in which the quality of the product relies on various parameters such as temperature, extrusion speed, etc.; hence there should be a controller working in an optimized manner required to maintain these parameters properly suitable for the system, thereby producing the quality product without any defects. A PID controller is designed and tuned for the extrusion process using the transfer function model of the system. The controller parameters such as Kp, Ki, and Kd values are optimized using Particle Swarm Optimization technique. The performance indices such as ISE, IAE, ITAE, and ITSE for the normal PID controller as well as PSO optimized PID controller were compared and the output has been produced using MATLAB/SIMULINK model. Moreover, NARMA-L2 controller is also tuned and the output is produced. Keywords: Controller, model, MATLAB, NARMA-L2, optimization, PID, PSO