Process Control System Development And Model Optimization For Hot Strip Finishing Mill

The process automation control system is the important part of the computer control system, and it is the most effective control method to ensure the thickness precision of the hot strip. As one of the main quality indexes of hot strip production, the thickness control strategy is the key research of rolling automation. Based on a building project of hot strip production line, taking the precision of thickness control as the key point, the theoretical research was carried out and optimized methods was taken into the thickness control models, finally a comprehensive automatic control system for hot strip rolling was developed and applied to the practical production, which had good performance. The main research contents are as follows:(1) The multi-processing, multi-tasking process control system for hot strip rolling was established. In the view of the characters of hot strip rolling, the control system which meets the actual requirements was developed. The multi-process structure and one task one thread model were used in the control system. The stability of the system was improved greatly and the coupling between the various functional modules was reduced, with the comprehensive control system, the communication process, database process, tracking process and model calculation process were final developed.(2) The multi-samples data processing and corresponding model self-learning were researched. The communication between basic automation system and process automation system was achieve by the OPC and Socket streaming socket technology, and the data collection and process platform was set up. A multi-samples data processing algorithm was put forward, in which the variation coefficient of the samples was received to calculate the data precision, with the method the higher accuracy source data was received, a corresponding model self-learning strategy was put forward, and the model precision was enhanced.(3) The temperature set-up models and multi-objective optimization strategy were researched. The temperature theoretical models which contain the air cooling model, water cooling model and deformation zone temperature model were deeply analyzed. Based on the theoretical models, the temperature distribution model in the entrance and inner of the finishing rolling zone were build. In view of the defects of the traditional temperature model self-learning, an algorithm was put forward to optimize the self-learning model, in which the multi-objective function was solved with the Nelder-Mead method, final the temperature distribution in the finishing rolling zone was received, and with the method the accurate temperature distribution was received.(4) The high precision thickness related models were build, and the rolling force model and the roll gap position model was given, also the main influenced factors of the model was analyzed. Combined with the actual production process, the rolling force model was optimized, and the prediction precision of the rolling force was enhanced when the thickness layer changed. In view of abnormal actual data, the concept of the velocity adjustment coefficient was put forward, and with which the gap position model was optimized, and the self-learning efficiency was enhanced, and the thickness precision was guaranteed.(5) The intelligent algorithm was used to optimize the gap position model. Starting from the construction of the gap position model, the reason of low thickness precision in the start of a new rolling or long time rolling stop was analyzed, and a new self-learning model based on static error was put forward to optimize the model, in which the model self-learning coefficient was divided into two parts, and with the method, the control effect was obvious; the temperature change of the work roll was analyzed by the ANSYS model, and the roll thermal expansion curves with different water cooling heat transfer coefficient were received, and an estimated method was used to determine the model parameters, which has the general applicability for different rolling line, the thickness precision in a long time stop was enhanced; also the threading adaption method was optimized, with a reasonable detective stand, the success rate of the threading adaption was enhanced and the thickness precision was enhanced.(6) The results of the thickness control effect were analyzed with the actual control data, and the thickness precision was in a high range. The process automation control system has been applied successfully in a hot strip production line. It is beneficial to the computer control level of hot strip rolling in China.