Study And Optimization On Flatness Control System Function In Cold Strip Rolling

The flatness control technology of cold rolling is one of the core technologies of strip processing in cold rolling.In recent years,with the development stage of China's steel industry from extensive development stage to industrial structure transformation stage,an increasingly important position has been occupied by flatness quality of cold rolling in the enterprise competition.In this paper,the research background is a flatness control system upgrade project in a 1450mm 5 tandem cold rolling mill.Under the premise of analyzing and studying the basic model of flatness control,the core model in the flatness control system has been optimized and improved.At the same time,the research results have been applied to cold rolling production line and the good control effect has been achieved.The main research contents are as follows:(1)An iterative matrix of roll pressure is established based on the roll deformation equation.The effects of roll elastic deformation and roll flattening on strip flatness are analyzed.The asymmetrical temperature compensation,coiling compensation and edge thinning compensation in flatness curve are studied.The radial force,wrap angle and area coverage factor in the processing of flatness are calculated.The optimization model of flatness actuator adjustment in rolling production line is studied.Based on the allowable direction method and simplex method,a calculation method of flatness actuator quantity is designed.This method can meet the requirements of the accuracy in on-line flatness control.(2)A collaborative optimization strategy for flatness control system is proposed.A collaborative optimization algorithm based on neural network and Topkis-Veinott is designed.Through the combination of search thinking mode and learning thinking mode,the determination manner of search direction is improved and the possibility of iterative trajectory in a similar route is reduced.The adjust direction consistency between work roll bending and intermediate roll bending is ensured.The offset phenomenon in adjust effect is effectively avoided.And the roll wear degree is greatly reduced.(3)A model of flatness adjustment strategy library is proposed.Through the establishment of flatness analysis model,the actual flatness discriminant factor and the flatness actuator adjustment are respectively solved by artificial neural network and genetic algorithm.The optimal adjustment mechanism combination is selected according to the reasonable range of discriminant factor.The reasonable flatness control means can be accurately chosen.On the basis of giving full play to the flatness adjustment ability in rolling mill,the control accuracy of strip flatness is improved.(4)A dynamic adjustment model of flatness curve is designed.A hybrid algorithm based on GENOCOP is proposed.By analyzing the relationship between the saturation state of the adjustment mechanism and the setting of the flatness curve,the intervention amount in the flatness target curve for eliminating the flatness deviation is solved in the saturated state of the adjustment mechanism.The problem of excessive adjustment or inadequate regulation in conventional intervention can be solved.The shortcoming of adjusting a single flatness factor in conventional intervention can be overcome.The global control function of all flatness coefficients can be realized.(5)This paper development the design scheme of transformation of flatness control system in a 1450mm 5 stand cold tandem rolling mill,including hardware configuration,the main interface functions,adjust the process interface functions,adjust the parameter interface functions.The gain test in the flatness control system and the calculation of the speed of the actuator are analyzed.For thin gauge strip,conventional gauge strip,thick gauge strip,the effect of flatness control before optimization and the effect of flatness control after optimization are analyzed.The application results show that the optimized flatness control system is stable and the flatness control precision is improved obviously.