IGA-RBF-Based Strip Shape Intelligence Recognition And Control Research

Strip steel is widely applied in various areas and plays an important role everyday life. The development of modern industry imposes a higher requirement on the manufacture of cold strip steel, so the problem of how to increasingly improve the quality of strip shape is indeed necessary and critical. Therefore, shape control for the strip steel is a critical task facing the steel and iron industry. On the other hand, intelligent recognition of strip shape is a precondition and a focal issue for the shape control as well as a key step for closed-loop control system. Shape control system is a class of complicated system with multi-variable, nonlinearity and multi-disturbance, which makes the traditional shape control theory reluctant to deal with it. Over the years, artificial intelligence provides an innovative method for the strip shape intelligence recognition for its powerful capacity in modeling and optimization. Based on the comprehensive analysis for the current development of shape intelligence recognition and control, this paper strives to the shape intelligence recognition and control algorithm to overcome the shortcoming of traditional control methods.Firstly, the problems of traditional pattern recognition such as low anti-disturbance capacity, long learning time and local minimization etc. gains a closed analysis, by using immune genetic algorithm(IGA) to train the RBF parameters, a IGA-RBF shape intelligence recognition model is established. Take the difference of fuzzy disturbance of waiting-recognition sample and reversing basic pattern as IGA-RBF input, to make IGA-RBT input node halved. Consequently, the precision and speed of shape intelligence recognition can be improved.Secondly, in the strip shape control system, the quality of strip shape is affected by many factors , such as rolling force, the condition of feeding strip steel etc., leading to the difficulty of modeling using traditional method. The advantages of predicting control, which has the capacity to establish the model independent on the precise model and the controller design free of system model, is taken in this paper, and a IGA-RBF-based predicting control method is proposed and applied in the shape control system.Finally, Simulation results demonstrate the effectiveness of proposed method.