| With the continuous development of iron and steel industry, in order to increase production by increasing rolling speed, the technology of flying gauge change (FGC) has been improved constantly in tandem cold mills (TCM). Flying gauge change is a very concernful dynamic state of TCM. The quality of the strip's head and tail (the disqualification lenth of the strip's head and tail) will be seriously affected by the quality of FGC's control.The accuracy of thickness'preset is the main factor affecting the accuracy of FGC. So how to improve the accuracy of thickness'preset has been the key issue in the study of FGC.By taking the five stands mill of tandem cold mill in Shanghai Baosteel as the research object, this thesis analyzes the process of FGC and the technology of dynamic setting, The main contents include the following three parts:The first part mainly analyzes the existing technology of FGC in TCM. Starting from the formation of wedged transitional zone and its parameters calculation strategy, the time choosing of transform is studied. Then the basic control strategy and control law are analyzed. Then a study of the model of tension calculation is carried out of FGC. The model of tension calculation of wedged transitional zone is constructed based on the model of routine tension calculation at first. As the transition zone is not an ideal wedge, the model of tension calculation of arbitrary shape transitional zone is constructed. Then the dynamic discrete model of tension calculation is found.The dynamic setting model of rolling parameter of 1800 TCM is constructed in the second part. The dynamic setting models of before first stand and after first stand are constructed. Then gap dynamic setting up and roll-speed dynamic setting up are constructed when the welding line passes through stands.A research of the optimization of the rolling parameter of FGC is carried out in the third part. When the mathematical model of rolling force is constructed, in order to facilitate the calculation many measures are simplified, so even using the adaptive learning methods to modify the deformation resistance coefficient and friction coefficient, rolling force's accuracy is still not particularly high. In order to further improve the rolling force of the preset, based on the adaptive learning methods, BP neural networks and adaptive learning methods of mathematical models are combined together to constructed an intelligent system of rolling force preset in FGC.The simulation results show that by using adaptive learning methods, the preset accuracy of rolling force has been increased. The combination of BP neural network and adaptive learning methods made the preset accuracy of rolling force increased highly. The high precision preset of rolling force was achieved, and the quality requirements of the strip's head and tail are met. |
