Chattering Control Based Algorithm for Nonlinear Optimization of 5-Stands Cold Strip Rolling Process Parameters

Chatter is usually one of the most difficult tasks to overcome in a rolling mill. Chatter is often associated with high rolling speeds and high reductions, conditions that are generally desirable in order to increase mill productivity. This paper presents the combination of various methods to optimize the 5-stands rolling process parameters in order to increase the rolling speed, preventing chatter to occur. This study is done in four stages: Dynamic modeling of rolling process, fractional factorial designing in order to address the response, developing a statistical model to relate response and independent variables, and finally optimizing the process by simulated annealing algorithm. In the first stage, a chatter model of 5-stands rolling process is presented. Every stand has two sub models for rolling process and structure. In the second stage, fractional factorial design is carried out in order to compute the System Equivalent Damping (SED) parameter for various conditions. SED is a parameter that can be used as a criterion to find safety margin to chatter. Selected design factors are friction coefficient and reduction of 5-stands and speed of strip. In this study a 2(11-5) design has been selected. In the third stage, a model has been developed for SED using statistical analysis. In the last stage, an optimization algorithm called simulated annealing is applied to response surface to optimize the rolling process parameters. In the optimization problem the rolling speed was defined as the objective function which should be maximized. Various constraints exist in this problem: bounds that present the minimum and maximum values of each variable, nonlinear constraint according to SED (response surface created by regression) and equality constraint according to total reduction of 50%. By using this combination method, significant improvement achieved in rolling speed.