Differential evolution has shown success in solving different optimization problems. However, its performance depends on the control parameters and search operators. Different from existing approaches, in this paper, a new framework which dynamically configures the appropriate choices of operators and parameters is introduced, in which the success of a search operator is linked to the proper combination of control parameters (scaling factor and crossover rate). Also, an adaptation of the population size is adopted. The performance of the proposed algorithm is assessed using a well-known set of constrained problems with the experimental results demonstrating that it is superior to state-of-the-art algorithms.
展开▼
机译:差分演变在解决不同的优化问题方面取得了成功。但是,其性能取决于控制参数和搜索操作员。 Different from existing approaches, in this paper, a new framework which dynamically configures the appropriate choices of operators and parameters is introduced, in which the success of a search operator is linked to the proper combination of control parameters (scaling factor and crossover rate).此外,采用了人口大小的改编。使用众所周知的一组约束问题评估所提出的算法的性能,实验结果表明它优于最先进的算法。
展开▼