中文摘要：

针对轧机传动系统扭振控制问题，建立含间隙非线性的轧机系统动力学模型。考虑到轧机扭振模型的非线性和参数不易测量的特点，提出神经网络和模糊PID相结合的控制器设计方法，以模糊PID为主体，通过引入神经网络改变模糊隶属度函数的中心值和宽度，最终得到最佳PID参数。设计神经网络-模糊PID智能控制器，并利用实际轧机参数与经典双闭环控制系统进行对比仿真。仿真结果表明所设计的智能控制系统对轧机传动系统扭振的抑制作用明显优于经典双闭环控制系统。

英文摘要：

A dynamic model for a rolling mill with backlash nonlinearity is established for torsional vibration control analysis of the rolling mill’s drive system. Considering the nonlinearity of the model and the difficulty in parameters measurement, a controller’s design method of combining neural network with fuzzy PID is proposed. In this method, the fuzzy PID is the dominant. By introducing the neural network to adjust the central value and width of the fuzzy membership degree function, the optimal PID parameters are obtained. Then, the intelligent control system of the neural network combined with the fuzzy PID is designed and simulated using the real rolling mall parameters and the parameters of the classic double loop control system. The results show that the designed intelligent control system can suppress the torsional vibration of the rolling mill drive system obviously better than that of the classical double loop control system.