The Proportional, Integral, and Derivative (PID) is a linear control technique that constitutes more than 90% of all industrial control systems. In reality, industry faces nonlinear systems that possess multiple equilibrium points with various dynamical properties. As a result, nonlinear dynamics have to be linearised near an operating region and a PID controller must be distinctly tuned to regulate the system near that region. However, linearised dynamics work only for small operating regions that if escaped, can cause deteriorated performance or instability. A common nonlinear system seen in industry is the Nonlinear Continuous Stirred Tank Reactor (NCSTR) which is often controlled using PID at linearised operating points. In this paper, a novel Nonlinear PID (NLPID) controller is designed and simulated for a saturated NCSTR system in MATLAB/Simulink. A Particle Swarm Optimisation (PSO) algorithm is used to tune the proposed NLPID controller. In addition, a parametric uncertainty study is conducted to show the robustness of the proposed NLPID controller. The proposed NLPID controller shows robust regulation of the NCSTR nominal system with reasonable overshoot, fast rise time and fast settling time for various operating regions. The proposed NLPID controller shows fast disturbance rejection and is able to eliminate steady-state errors and maintain stability. The results also show that the proposed NLPID controller has a large operating region for the NCSTR system. Finally, the proposed NLPID controller shows fast performance and is robust against \(\pm 5\%\) parametric uncertainty with minimal impact on transient speed.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

A PSO Tuned Robust Novel NLPID Controller with Application to Nonlinear Systems

  • Stefanos Charkoutsis,
  • Mohamed Kara-Mohamed

摘要

The Proportional, Integral, and Derivative (PID) is a linear control technique that constitutes more than 90% of all industrial control systems. In reality, industry faces nonlinear systems that possess multiple equilibrium points with various dynamical properties. As a result, nonlinear dynamics have to be linearised near an operating region and a PID controller must be distinctly tuned to regulate the system near that region. However, linearised dynamics work only for small operating regions that if escaped, can cause deteriorated performance or instability. A common nonlinear system seen in industry is the Nonlinear Continuous Stirred Tank Reactor (NCSTR) which is often controlled using PID at linearised operating points. In this paper, a novel Nonlinear PID (NLPID) controller is designed and simulated for a saturated NCSTR system in MATLAB/Simulink. A Particle Swarm Optimisation (PSO) algorithm is used to tune the proposed NLPID controller. In addition, a parametric uncertainty study is conducted to show the robustness of the proposed NLPID controller. The proposed NLPID controller shows robust regulation of the NCSTR nominal system with reasonable overshoot, fast rise time and fast settling time for various operating regions. The proposed NLPID controller shows fast disturbance rejection and is able to eliminate steady-state errors and maintain stability. The results also show that the proposed NLPID controller has a large operating region for the NCSTR system. Finally, the proposed NLPID controller shows fast performance and is robust against \(\pm 5\%\) parametric uncertainty with minimal impact on transient speed.