Research On Flatness Recognition Based On Fuzzy Chaotic And Control Technique

With the expanding demands of plate/sheet output and quality, the technique of flatness measurement and control has been the research key of modern high precise plate/sheet rolling mill. It is important investigation aspects for Signal recognition, advanced control method and strategies of flatness. In this study, the system of flatness measurement and control has been researched based on fuzzy technique, chaotic optimization and Generalized Predictive Control (GPC). Main works are as follows:For the advance and absence of traditional recognition method, fuzzy recognition method and neural network recognition method for flatness signal pattern, a fuzzy chaotic recognition method is presented firstly. There are two stages in Signal recognition pattern of flatness. Firstly, fuzzy recognition as preliminary one to decrease the solution dimension and reduce search band is adopted. Subsequently, chaotic optimization is used to further recognize. The recognition method has many merits, such as simplification, fast and stability, high accuracy.In two stages optimization recognite method, Based on the analysis and investigation of the chaotic optimization algorithm, a span of mapping optimization vriable has been built-up to divide chaotic searching range; For the chaotic optimization has low ability in the aspect of searching local range, a new way, combining the prior knowledge with the local search, has been developed to improve the searching ability with the aid of gradient decrease idea.A fuzzy prediction model of rolling force has been established based on the fuzzy recognition theory. In this report, the triangle subjection function and the clustering subjection function are employed to recognize precondition parameters of the fuzzy model. And weighting recursion least-squares algorithm (WRLSA) is employed to recognize inclusion parameters. Simulation experiment show that this model has the characteristics of high prediction accuracy and high operation speed.Due to non-linearity, time-variation and non-determinacy of hydraulic roll-bending control, as well the anti-interference request, a hydraulic roll-bending control scheme has been put forward based on GPC after analysis of system mathematical models. Widening the using of the advanced control theory apply to plate/sheet rolling mill control.GPC has been sturdied, the compensated GPC model with predicting output error has been set-up. The time series AR model with the high real-time property can predict and compensate the system future time error; A new simplified recursion optimization algorithm has been developed to solve the GPC real-time problems caused by the amount calculation; The GPC algorithm is improved to decrease overshoot, namely, the output parameter replaced with weighting average variable both predicted current and future controlled variable. Meanawhile, the output parameter is also one of rolling optimization objects to decrease system overshoot effectively. The system dynamic response speed and steady control quality has been realized.