关键词: 大规模优化, 变量分组, 加性可分, 优化器, 协同进化

Abstract: The rapid development of big data has increased the scale of problem in the field of the optimization, and brought great challenges to the original optimization methods.The article makes a research the new method of large-scale optimization based on Divide-and-Conquer strategy. Firstly, based on the principle of additive separability, an improved variable grouping method is proposed. This method uses the random access point, and paired the detection of interdependency among all the variables. At the same time, by making full use of the interdependency information of learning, the large groups of separable variables were re-grouped. Secondly, based on the improved differential evolution algorithm to design a new subcomponent optimizer, to enhance the subspace optimization performance;Finally, this two kinds of improvements are introduced to co-evolutionary framework to construct a DECC-NDG-CUDE algorithm. Two experiments of grouping and optimization were made on 10 large-scale optimization problems.Grouping experimental results show the interdependency between variables can be effectively identified by the new method of grouping; The optimization experiments show that the performance of DECC-NDG-CUDE is better than two state-of-the-art algorithm DECC-DG and DECCG .So the DECC-NDG-CUDE is an effective algorithm for solving large-scale optimization problems.

Key words: Keywords: large-scale optimization, variable grouping, additive separability, optimizer, Cooperative Co-Evolution