脉冲滴丸机压力控制系统是具有非线性、参数时变性和耦合性特征的复杂系统,系统响应要求快速、准确、鲁棒性强,但传统的控制方法如常规的同定参数的PID控制器很难使系统同时达到这种最优控制。因此,提出了将自适应控制、模式识别和PID控制相结合的思想,设计了一种基于模式识别的脉冲压力自适应控制策略,通过在线辩识系统的动态输出,根据系统响应所处的状态,采取相应的控制算法,使系统得到了较好的控制。MATLAB仿真和现场实验结果均表明该方法具有较高的动态性能和控制精度,以及较强的鲁棒性。该控制系统已成功投入现场运行。
The pulse pressure control system in pulse capsule forming machine is a nonlinear, time-variant and coupling system. The system requires good speed, accuracy and robustness performances, but they are hardly obtained simultaneously using traditional methods, such as preset parameter PID controller. To handle this problem, a new adaptive control policy based on pattern recognition is proposed by combining adaptive control, pattern recognition and PID control technique, which can achieve better results and identify the dynamic output of the system on line using different control strategies. MATLAB simulation and physical experiment results show that the proposed method is efficient in designing a robust pulse pressure control system for pulse capsule forming machine and has good dynamic characteristics, high accuracy and strong robustness. The proposed control system has already used in field application successfully.