Research on Multivariable Fuzzy Dynamic Matrix Control Algorithm

Abstract

To improve the control quality of distillation tower systems that suffer from issues such as coupling, delay, and nonlinearity, traditional control methods have been ineffective. Consequently, this paper proposes using a variable-domain fuzzy dynamic matrix control algorithm for control system optimization. The traditional dynamic matrix control algorithm is augmented with a fuzzy controller, which is utilized to compensate for predicted output. The concept of variable domain is introduced to address the problems of poor adaptive ability and low control accuracy in the Fuzzy-DMC algorithm. Scaling factors are designed to adjust the fuzzy domain, enabling the fuzzy controller to compensate for predicted output even in the event of a significant model mismatch, resulting in improved control performance. The experimental results indicate that the algorithm proposed in this paper is superior to both traditional DMC and Fuzzy-DMC algorithms in terms of rise time, overshoot, and steady-state rate. In addition, this algorithm possesses a simple structure and does not modify the conventional DMC algorithm structure, making it theoretically applicable in practice.