Multi-objective Evolutionary Algorithm And The Application In Its Application In Different Dimension Cold Rolling Down Strategy

With the continuous progress of science and technology,rolling steel products are required to be better and better in the market,the study of press strategy have helped the product to meet market requirements,at the same time reducing energy consumption.In this paper,a 1250 mm five-stand cold tandem mill is used as the research background,and the improved strategy is better optimized by the improved algorithm,the results of optimization can give the engineers more choices.The main contents of this paper are as follows:Firstly,the basic concepts of multi-objective optimization problems are introduced,the important concepts are analyzed,such as Pareto domination and Optimal solution set.Then,the mathematical functions of the strategy are analyzed,and four objective functions which needed to be studied are determined.Secondly,the core principle of nondominated sorting genetic algorithm is described in detail,and the existing problems in the original algorithm are analyzed.Based on the discovered problems,a two-dimensional information cycle squeezing strategy with order of magnitude is designed.Through 30 independent repetitive simulation experiments,the experimental results show that this strategy can better balance the population distribution and convergence.The improved algorithm is applied to the two-dimensional and three-dimensional cold rolling reduction strategy,and a better Pareto optimal solution set can be obtained for decision-making reference.Thirdly,some theories,such as reference point(preference)theory and neighbor strategy,are proposed to analyze the high-dimensional multi-objective evolutionary algorithm.In order to better adapt different reference points,an improved mutation strategy is proposed.For the purpose of verifing the validity of this strategy,the simulation results contain the calculation of the evaluation index value for the high-dimensional problem.In this paper,the convergence of the test function itself is used to measure the degree of convergence.Then,the improved high-dimensional multi-objective evolutionary algorithm is used to optimize the four-dimensional reduction strategy,and the solution process of the optimal solution set is described in detail.Finally,I summarize the research results and innovation of this paper in the conclusion of the article,and point out the further work of this subject and the future research direction.