Human-simulated Intelligent Control Based On Parameters Setting By Fuzzy Neural Network Of Hot-rolling Strip Thickness

The strip thickness has been suffered from size, shape, type of raw material, as well as the accuracy of heating temperature, chemical composition and rolling force, which makes it is difficult to establish its accurate mathematical model. The strip thickness control has the characteristics of uncertainties、time-varying parameters and strongly nonlinear, the traditional control method is difficult to meet the requirements of customers for high precision strip steel thickness. Therefore, a good property of control algorithm is very necessary.Human-simulated intelligent control is a kind of human control experience, way of intuitive reasoning as the foundation, to avoid solving complicated object model method, so it shows the unique advantages in the process control, and provides a good solution to solve the complex industry control, but the human-simulated intelligent control is also insufficient. Therefore, this paper in view of the human-simulated intelligent control in the controller parameter correction of the problems existing in the practical application, designed human-simulated intelligent control strategy based on fuzzy neural network parameters setting of the hot rolled strip steel thickness. The neural network has high self learning capacity and adaptive capacity and strong generalization ability, but the traditional neural network learning speed is slow and easy to fall into local minimum, and fuzzy control is able to take advantage of a priori knowledge to approximate reasoning, can reduce the learning time, but it lacks of self-learning and adaptive ability, fuzzy neural network is a control strategy which fully absorbs the advantages of neural network and fuzzy. This paper uses parameters setting by fuzzy neural network for human-simulated intelligent controller, because it simulates some of the features of the brain, so it is higher intelligent.Supported by Natural Science Foundation of Shanxi Province (No:2010011022-3), process of thickness control is used as a research object in the process of strip production, in view of the exiting problems in the thickness control, this paper design a control strategy of human-simulation intelligent control of fuzzy neural network of hot rolled strip thickness.The main research content of this paper has:(1) The description is given of the background and significance of the research topics, summarized the development of hot rolled strip production technology and research status of hot rolled strip thickness automatic control technology, discussed research situation of human-simulated intelligent control.(2) The production process of hot-rolling strip was studied after extensive research in a hot rolling steel production, analysis of the impact of the main performance indexes of products in hot strip quality and the influencing factors of the strip thickness, according to the existing problems in the field of strip thickness control of hot rolled strip thickness control system, study on hot rolled strip thickness control system.(3) Based on the study of humanoid intelligent control theory and its basic principles, focusing on design procedure and method of human-simulated intelligent control, especially the key technology in the process. Hot-rolled strip thickness control process is used as a research object, a lot of simulations are done in the MATLAB simulation platform. Through the analysis of simulation results, if human intelligent control parameters are different, the output of strip thickness will be different.(4) Intensive study of the structure of fuzzy neural network and its basic principle, the deviation change of thickness are used as input getting human-simulated intelligent control parameters setting by fuzzy neural network, so as to further enhance the ability to adapt to the imitation of human intelligent control.(5) According to the hot rolling strip steel thickness control accuracy requirements, the strategy of human-simulated intelligent control based on parameters setting by fuzzy neural network of hot-rolling strip thickness is designed. In the MATLAB simulation platform, the simulation of human-simulated intelligent control based on parameters setting by fuzzy neural network of hot-rolling strip thickness under different expectations and in the presence of interference is studied, research results show that human-simulated intelligent control based on parameters setting by fuzzy neural network of hot-rolling strip thickness has a stronger adaptability and robustness.