Font Size: a A A

Study On The Rolling Schedule Optimization And Model Self-Learning Of Five Tandem Cold Rolling Mills

Posted on:2008-10-22Degree:MasterType:Thesis
Country:ChinaCandidate:F P DouFull Text:PDF
GTID:2121360212495245Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
Plate and strip production plays an important role in national economy. The gauge and shape are the main accuracy standards of plate-strip product. The rolling schedule is the main technological substance in plate and strip production, and it is the fundamental guarantee for the rolling mill capacity, product quality and accuracy and the quality of shape. Presently, the procedure only based on the experience is not satisfying. Therefore, to offer a more rational schedule, the field experience should be concluded into theory for more extensive significance.Firstly, after researching the processing model and controlling model of the tandem cold rolling mill, we proposed the main parameter models for designing rolling schedule. On considering the defects of rolling force model, the Neural Network (NN) is used to predict the rolling force instead. The rolling schedule can be revised on-line based on models'self-learning off-line. Secondly, different object functions and constrain conditions are proposed for different requirements based on the analysis of the characters of the cold rolling. Further more, two optimization methods are offered for optimizing the rolling schedule. With these, we optimized the field rolling schedule and show the results of the after and before the optimization.Lastly, we analyze the variables online disturbance and the influence to the rolling procedure, then, propose the corresponding control methods which are adaptive and self-learning. According to the classification of the material, relation table has been established in database, which can connect special type steel with neural network's parameters accordingly and save the result of network training. In the meantime, the table of rolling process is proposed to convenient the schedule design in the future.
Keywords/Search Tags:Tandem cold rolling mill, Rolling schedule, Neural network, Genetic algorithm, Adaptive, Self-learning, Model base
PDF Full Text Request
Related items