Trusted Evolutionary Algorithm for Global Optimization

In this paper, the initial population is generated by orthogonal design method which can generate a uniformly distributed initial population in the search space, and a new crossover operator based on Latin square design and an approximate method of principle component analysis is presented. The approximate method first determine the sensitive components and then the crossover operator is mainly used to these components. As a result, the crossover operator has the ability of local search and thus can exploit the search space efficiently. Furthermore, a mutation operator based on quantization is presented. Based on these, a trusted evolutionary algorithm for no differentiable global optimization is proposed and its global convergence can be guaranteed, which indicated the proposed algorithm is trusted. At last, the numerical simulations are made for some standard test functions. The performance of the proposed algorithm is compared with that of two widely-cited algorithms. The results indicate the proposed algorithm is effective and has better performance than the compared algorithms for these test functions.