Font Size: a A A

Study On Schedule-Optimization Practice For Eight-Roller Five-Stand Tandem Cold Rolling Mill

Posted on:2010-08-04Degree:MasterType:Thesis
Country:ChinaCandidate:C LiuFull Text:PDF
GTID:2121360302459036Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
Rolling schedule is the core content of rolling process, a reasonable schedule can give full play to the equipment capacity, reduce energy consumption, ensure the accuracy of strip products, so that rolling process achieves the best state. With the rapid development of artificial intelligence technology, intelligent optimization algorithm is applied to rolling schedule more and more, replacing the traditional experience.There are many factors affect product quality in rolling process, this paper sets up objective functions, such as equal relatively load objective function, anti-skid objective function, poses multi-objective programming problem. Because of conflicts among the objective functions, generally, the optimal solution can not be simultaneously achieved. Adopting the ideal point method to find effective solutions, change multi-objective optimization into a single-objective optimization with weight coefficient, so that the corresponding solution of the target closest to various single optimal solution, all parties to achieve the desired effect.Considering limits of rolling equipment, process conditions, identify constraints. Apply sequential unconstrained minimization technique (SUMT) to change constraints into unconstraint, restrict solutions within reasonable limits.Particle swarm optimization (PSO) and genetic algorithm hybrid particle swarm optimization (GAPSO) are used to optimize rolling schedule, respectively, for multi-variable, nonlinear, strong coupling, the rolling time-varying systems.Combination with ideal point method and the penalty function to optimize objective functions, in order to provide basic automation level pressure, speed, roll gap, rolling force, etc. settings. This paper analyzes the two algorithms , selects control parameters and gives a comparison of results before and after optimization. Optimized schedule improves the productivity of the cold rolling strip, reduces energy consumption, while ensures product quality.It was difficult to accurately express the actual rolling force for the previous theoretical model. this paper takes actual rolling force as teaching information, using particle swarm algorithm to optimize BP neural network, training rolling force model, so that rolling force prediction to be accurate, for rolling schedule calculation.
Keywords/Search Tags:tandem cold rolling, rolling schedule, multiple objective optimization, SUMT, PSO, GAPSO, neural network
PDF Full Text Request
Related items