Study On Rolling Schedule Optimization Based On Improved Differential Evolution Algorithm

Rolling schedule optimization of cold tandem mill is a nonlinear constrained multi-objective optimization problem, in which both reduce energy consumption and improve the quality of the mill’s production are considered. Making reasonable rolling schedule can prolong service life of rolling mill, reduce the failure probability, as well as, improve the working efficiency under ensuring the product quality. In this paper, based on the study of traditional differential evolution algorithm, a new Self-adaptive Differential Evolution Algorithm(SE-DE) was presented according to the demand of production practice, and was applied in multi-objective optimization problem.Firstly, an external archive is utilized for opposition-based learning, which can effectively increase diversity of the population and expand the searching space, by this way, more high quality solutions can be selected for candidate solutions. Then, making an analysis and research of all kinds of adaptive selection mechanism, in order to reduce limitations of fixed parameters and improve the adaptability to different problems, a roulette wheel based F selection scheme is introduced to adaptively select the mutation factor(F), it can be adaptively controlled based on the success of offspring/trial solutions generated. Finally, an optimization factor α is proposed to select the crossover strategy, the combination of binomial and exponential crossover can effectively balance the exploration and exploitation ability of the algorithm. The performance of SE-DE is compared with the other five DE algorithms including DE, SADE, ODE, NDE and MDE-p BX. The comparison is carried out for a set of 30-, 50- and 100-dimensional test functions from CEC-2005. The results show that our algorithm is better than, or at least comparable to, the algorithms from other literature.Taking the five tandem lines for strip cold rolling as the example, the structure of control system, the equipment parameters and technological requirements were analyzed, describes in detail the related mathematical models of rolling parameters. And on this basis, selecting the equal relative load and the slip rated as the objective functions, the optimization by the SE-DE algorithm is performed on a five-stand cold tandem rolling production line. Simulation results verify the effectiveness and superiority of the proposed algorithm, and it also plays a guiding role for multi-objective optimization problem of other fields.