【摘要】 通过对遗传算法算子的行为分析,给出概率动态变化的交叉和变异算子。同时,将遗传算法与随机模拟结合,设计GASSII遗传模拟混合智能优化算法。随机模拟采用缩减方差、加速收敛的分层抽样技术,保证抽样遍布整个搜索空间,突破个体分布“畸形”交叉对遗传算法进化能力的限制,有效发挥遗传算法的隐含并行搜索特性,在发生不成熟收敛时能从当前局部解中跳出,搜索转向具有更高适应度的抽样解,迅速收敛到全局最优解,从而有效避免早熟现象,提高了寻优的效率与精度。Benchmark问题评测结果显示出该算法的有效性。更多还原

【Abstract】 Analyzing operators’ behaviors of Genetic Algorithm,this paper proposes dynamic adaptive crossover and mutation operators.Genetic Algorithm with Stochastic Simulation of Stratified Sampling Intelligently Integrated(GASS II) is further constructed.Stratified sampling technique is adopted to ensure that samples spread all over the search space and latent parallel ability facilitate the grabbling process to overcome prematurity and improve the optimization efficiency and precision,which are revealed in performance evaluation of Benchmark problems.更多还原