Mill Deformation Resistance And Self-learning Mathematical Model Study

In steel production, the rolling process is finished or semi-finished products a key part, In this process how to improve product quality and overall efficiency is the key. Rolling the development of modern technology can be divided into two phases. One was to address issues such as rolling pressure, torque, power, spread, before sliding rolling process parameters, such as approximate calculation problem. Kalman equation and put forward Orowan equation, and rolling pressure formula, spread formula, formed a law as the core of the traditional rolling system. After the 1970s, mainly towards improving product quality, reduce consumption and optimize the rolling direction. Especially the modern electronic computer and start rolling for process control, and promote the stability of the rolling speed and improve the accuracy of the products improved significantly. In the 1990s, the first in Japan, Germany, and then set off in the world in a rolling process in the application model adaptive, self-learning, artificial intelligence craze. This article is with the second Lian rolled steel plant project with the board control system optimization project, a self-learning on how to improve the mill control methods were studied precision, and its main contents are:(1) 3500 mm single-cycle machine-hot technology and control system. Level 1 is the first layer sensor, control equipment at the end of the control network; level 2 is the mill APC (Area Process Controller) and coiling / laminar flow APC; level 3 is the main model and rolling Tabulation calculation.(2) rolling mill control system mathematical model study. Mathematical model is the realization of the production process control of the foundation, and its role is to forecast parameters, and parameters describing the changes. We combine several 3500 mm , a mathematical model of the classification, the establishment of the assessment methods, research and development of its origin and its main mathematical model of rolling force, the temperature drop model for detailed analysis.(3)The Deformation resistance 'parameter model of self-learning and research. As Hot Roll existence in the course of elastic deformation, roll wear strip temperature changes, and so on many complex factors, many from the experimental summed up in the rolling model for rolling changes in the environment can not meet the original parameters have been high Rolling the needs of accuracy, thus in practice in accordance with the actual production process, using self-learning method, the parameters of the amendment, to further enhance the control accuracy, and ensure control of the mill requirements, improve product quality...