Study On Mill Drive System Based On Torsional Vibration Intelligentcontrol Method

With the increasing high demand for product quality in modern steel rolling industry,it is also required for the high accuracy and high dynamic performance of the rolling mill drive system.Because of the elastic connection between the motor and the rolls,when the load disturbance occurs,in particular, the mill bite the steel,the torsional vibration phenomena often occur in the transmission system. Torsional vibration mill not only reduce the dynamic persformance of the mill drive system, but also reduce the fatigue life of the mill drive system, even devastating damage to rolling machinery, affecting transmission of safety in production.For the actual mill system, this paper will driveline mechanical part and the electrical part of research as a whole. By analyzing the mechanical components of the transmission system,the mill system will be equivalent to the “mass-spriong system”which includes number of inertia parts and elastic parts and. Then the mass-spring is simplified into two quality models, and establish the mathematical model by using kinetic into two quality. This aricle will consider the rolling torsional vibration control system as the system design problems contatining nonlinear gap and link with external disturbances.From the perspective of the automatic control system, mill driving system appear in the presence of disturbances deceleration and dynamic mechanical torsional vibration phenomenon can be attributed to the automatic control of external disturbance regulation problem. Therefore, this paper studied the basic dual-loop control system,then use one of the two tone PI structure. Form the simulation resulrs, the contol performance of the control system is not very good, but it is the basis for other more advanced control systems.In order to be able to load the intermediate shaft torque and speed which in reality are difficult to directly measure the feedforward compensation and feedback control, we design the state observer to observer the load torque and speed. Taking into account the aspects of the system with nonlinear and uncertainty of model parametes, the state observer is designed for neural network state observer, using a neural mapping function and online self-learning adaptive, online observing the speed and load torque.For the slow rate of convergence in BP algorithm and the caught local minima shortcomings, this paper presents an improved BP algorithm. The simulation results show that for the neural network observer and torque feedforward compensation mill drive double closed loop control system, dynamic performance and torsional vibration suppression effects are better achieved good control effect,which demonsrrates that the designed control system for suppressing mill drives torsional vibration phenomenon is valid.To achieve better control effect, this paper fuzzy control technology into the mill control torsional vibration field, and make it combined with neural network, combing fuzzy-neural PIDcontrol system. Fuzzy control is a combination of rule-based expert system, fuzzy control theories and theoretical results of the set control theory. Through simulation results,it can be seen that the control effect is better that the traditional dual-loop control system.