论文部分内容阅读
建立了种群中最佳个体的马尔可夫链模型 ,定义了仅包含所有全局最优解的状态子集 ;根据从任意状态转移至该子集的概率的极限分布 ,给出了独立于搜索曲面拓扑结构的遗传算法全局收敛性的精确定义 ;提出并严格地证明了与编码方式和选择策略无关的、统一的全局收敛性判据定理 .对几种不同的遗传算法进行全局收敛性分析的结果表明 ,统一的判断方法具有普遍的适用性 .
The Markov chain model of the best individuals in the population is established, and a subset of states that contains only all the global optimal solutions is defined. Based on the distribution of the probabilities of the transition from any state to the subset, Topological structure of the global convergence of the genetic algorithm of the precise definition; proposed and strictly proved that the coding method and selection strategy has nothing to do with the unified global convergence criterion theorem. Several different genetic algorithms for global convergence analysis results It shows that a unified method of judgment has universal applicability.