基于改进BP神经网络的轧机扭振智能控制

针对轧机传动系统扭振控制问题,建立考虑负载转矩的轧机传动系统动力学模型。考虑到扭振模型比较复杂和参数不易测量的特点,提出基于神经网络的状态观测器,并对标准BP网络进行优化处理。设计基于改进BP神经网络状态观测器的智能控制系统,并利用SIMULINK对轧机实例进行仿真。结果表明设计的智能控制系统对轧机传动系统的扭振有良好的控制效果。

Intelligent Control of Torsional Vibration of Rolling Mills Based on Improved BP Neural Network

For the torsional vibration control problem of rolling mill’s drive system, the dynamic model of the drive system of the rolling mill is established considering load torque. Considering the characteristics of torsional vibration model is complicated and not easy to measure the parameters. The state observer is proposed based on neural network, and the standard BP network is optimized. The design of intelligent improved BP neural network observer based control system, and the simulation of the mill examples using simulink. The results show that the design of the smart control system of torsional vibration of the rolling mill drive system has good control effect.