Research On Identification Technology Of Asymmetric Characteristic Parameters Of Rolling Mill

Social development has put forward the requirements of higher strength and thinner strips.Moreover,the service time of rolling equipment has increased,and mills are more likely to exceed the design load threshold during operation.The original load balance state is destroyed,and asymmetric rolling occurs.The characteristic information of mill stiffness,stiffness difference,and roll system cross angle are vital parameters to evaluate the asymmetric state of the mill,which is essential for the intelligent operation and maintenance of the mill equipment.However,the existing measurements of asymmetric characteristic parameters are performed under static or quasi-static conditions and do not reflect the actual operating conditions of the mill system.In the face of the demand for the steel industry intelligent transformation,this paper analyzes the theoretical techniques of asymmetric parameter identification under dynamic conditions of the mill system.The details are as follows:Firstly,taking the F2 and F3 stands of a steel mill finishing unit as an example,the vibration behavior of the mill is analyzed based on the measured vibration response,strip parameters,and rolling process data.The focus is on the asymmetric phenomenon of the stiffness on both sides of the mill and the cross-deflection phenomenon between the rolls to elucidate the asymmetric problem under the dynamic conditions of the mill.Secondly,the mill stiffness on both sides,stiffness difference,and roll system crossover angle are introduced as the asymmetric characteristic parameters of the mill.The dynamics model of the vertical and horizontal coupling of the mill and the axial dynamics model of the mill considering the roll system crossover are established.Further,the parameter identification model is built by combining the particle swarm algorithm and the state space model method.The dynamic identification method of asymmetric characteristic parameters based on a kinetic model combined with rolling process data is proposed.The model validation and error evaluation are carried out,and good identification results are achieved.Finally,the rolling processes data,such as the mill’s vibration response and process parameters,are intensely mined to identify the variation of the asymmetric characteristic parameters of the mill under different rolling conditions.This paper also designed and developed the data analysis software for hot rolling mills based on Visual Studio with mixed programming in C# and MATLAB.The research results enrich the theoretical study of rolling mill parameter identification,which is of great significance to the intelligent operation of rolling mill equipment and the robust operation of the rolling process.