Research On Modeling And Optimization Of Automatic Gauge Control System In Hot Rolling Mill

Hot strip can be divided into ordinary carbon steel, low alloy steel and alloy steel on the basis of its material and performance. According to the different applications it also can be divided into cold forming steel, structural steel, automotive structural steel, corrosion-resistant structural steel, mechanical structural steel, welding gas cylinder and pressure vessel steel, pipeline steel, etc. As a raw material for a variety of products, the hot-rolled strip steel is used widely. With the development of society, a higher demand on the yield and quality of hot rolled strip has been put forward.The thickness accuracy is one of the two quality indicators of the strip and its control method is also one of the two key technologies in the strip rolling fields. Therefore depth study of the rolling mill automatic gauge control system is not only necessary but also has important theoretical significance and practical value.In this paper,we complete a study aimed at modeling and optimization of the hot rolling automatic gauge control system, the main contents and results are as follows:1) Based on referring to an extensive literature, an overview of the development of automatic gauge control, domestic and foreign status as well as the basic theory of automatic gauge control is reviewed;2) Start from the control algorithm, using the simulation method to study the mathematical control model of the Baoshan Iron and Steel1580hot rolling mill pressure automatic gauge control system, come to a conclusion that it exists the issue of slower control speed since its "multi-step regulation". After deriving the mathematical significance of the model rigorously and modifying the coefficient of pressure efficiency, we have realized the "one-step regulation" and improved the response speed of the pressure automatic gauge control system;3) Aimed at solving the problem of inevitable time lag in the monitoring automatic gauge control system, Smith predictor strategy is used to reduce the influence of the time lag on system dynamic performance to minimum. With the actual measured rolling force and rolling gap, using the spring equation to calculate the strip thickness and thus eliminate the drawbacks resulted from the inaccurate model of the Smith predictor control system;4) Strip tension changes with the thickness variation, which leads to the frequent jitter of the looper angle. To solve this problem, using flow compensation to compensate the flow balance timely and adjusting the real-time speed matching with the relationship to maintain the dynamic equilibrium of the strip flow, and thus can reduce the looper action and improve the thickness control accuracy;5) According to the conditions that when the thickness control system can be put into use, dynamic set method is used to control the accuracy of the thickness of the strip head and tail. Through improving the input timing of the dynamic set to optimize its effect;6) Due to the lack of physical instrument to measure strip thickness, the feed forward control is difficult to be applied in the strip hot-rolling field. To solve this problem, using the predict method rather than increasing thickness gauge to realize the feedforward control. According to the rolling force measured by the last mill,using the spring equation to calculate the thickness that will be rolled in the next mill, then tracking the calculated strip thickness properly, and ultimately realize the feedforward control to improve the strip thickness control precision effectively;7) Conduct multi-objective optimization settings for the finished rolling schedule. Make us of the characteristics of the finished rolling process and its mathematical model to establish a multi-objective optimization model for the finished rolling schedule, and use the non-dominated sorting Genetic Algorithm II to optimize the finished rolling schedule to obtain better results than the original rolling schedule;8) Use spectral analysis and other means to do the strip thickness deviation data pretreatment and analyze its reasons, then make a preliminary diagnosis of the fault and cause of the large thickness deviation. Develop the thickness deviation fault diagnosis system which can identify and classify typical fault type based on the Windows platform.