Adaptive Control Based on Pattern Recognition and Its Applications in Typical Industrial Processes

To improve the dynamic response, regulation precision and robustness of the closed-loop system, a novel adaptive Pattern Recognition based PID (PR-PID) control policy was proposed for industrial processes by combining the adaptive control, pattern recognition and PID control technique. The proposed method can achieve better performance by identifying the output of the system on line with different control strategies. The control strategies set mainly consists of PID controllers, which guarantees the basic requirement on stability and product quality. The effectiveness of PR-PID control is illustrated by its applications in some typical industrial processes. Simulation results show that the proposed method is efficient to design a robust control system for typical industrial processes with good dynamic character and high regulation. Since the proposed method need not identify the uncertain parameters of the plant, it has a very good real-time performance, and is easy to be implemented on-line.