Design of unconstrained and constrained model predictive control for pneumatic actuator system: Set-point tracking

Abstract

This paper presents the performance analysis of Model Predictive Controller (MPC) for the controlling of position of cylinder's stroke for pneumatic actuator system. The mathematical model of the system was obtained using System Identification (SI) technique and the linear approach using AutoRegressive with Exogenous Input (ARX) was chosen as a model structure to describe the dynamic behavior and characteristics of the system. The input constraints were applied to the MPC algorithm to make the controller operates closer to the limits. In this paper, the performance of MPC in controlling the position of the pneumatic actuator system as well as tracking the set-point was compared with Proportional-Integral (PI) controller. The performance of the controllers were assessed by taking into account the percentage overshoot (%OS), settling time (ts), rise time (tr), and steady-state error (SSE) of the responses. Simulation results show the effectiveness of the MPC with constrained input in producing better transient response and minimizing SSE.