Design of Decoupling Model Predictive Control and PID Controllers

• Main Authors: Yun-Ju, Chang, 張筠茹
• Other Authors: 鄭智成
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/9j7f3s

碩士 === 國立臺北科技大學 === 化學工程研究所 === 104 === Most industrial processes are multivariable systems which often present complicated interactions between controlled and manipulated variables. A multivariable system with strong interactions between the channels can be much more difficult to control, and the decoupling control by designing a decoupler to minimize the process interactions is advised. However, the ideal decouplers have complicated structures and are not always physically realizable. Model predictive control (MPC) has become the method of choice in several process industries to solve difficult multivariable control problems that include inequality constraints. However, the conventional MPC design does not explicitly address the process interactions so that satisfactory decoupling performance cannot be attained. This study proposes a novel design approach that incorporates the concept of decoupling control into MPC controller using the dynamic matrix model and extended non-minimal state space model. This decoupling MPC design achieves the condition of decoupling control without adding any decouplers. Furthermore, this study applies the decoupling MPC design to the on-line tuning of multivariable PID controllers. In this way, decoupling control can be achieved by implementing simple multivariable PID controllers without adding decouplers. Simulation results demonstrate the effectiveness of the proposed design of decoupling model predictive control.