论文部分内容阅读
针对板形控制和板厚控制之间存在强烈的耦合作用这一特点,提出一种板形板厚综合控制系统。对于这种纯迟延、时变、强耦合的多输入多输出系统,首先采用模糊RBF神经网络解耦方法对系统进行解耦,然后对解耦后的已近似成为两个独立的单输入单输出系统分别设计模糊神经网络PID控制器,从而建立了一种板形板厚模糊RBF神经网络分散解耦PID控制系统。通过MATLAB对此系统进行仿真,结果表明,该方法解耦控制效果良好,控制精度高,抗干扰性强,且具有良好的静态性、动态性及鲁棒性。
Aiming at the strong coupling between plate shape control and plate thickness control, a comprehensive control system of plate shape plate thickness is proposed. For this kind of pure delay, time-varying and strong-coupling multiple-input multiple-output system, firstly the system is decoupled by fuzzy RBF neural network decoupling method and then the decoupled approximation becomes two independent single-input single-output The system designs the fuzzy neural network PID controller respectively, so as to establish a decentralized decoupling PID control system of the plate thickness fuzzy RBF neural network. The simulation of the system by MATLAB shows that the method has good decoupling control effect, high control precision, strong anti-interference, and has good static, dynamic and robustness.