优劣复取舍遗传算法

为加快遗传算法的收敛速度 ,满足优化控制实时性的要求 ,对遗传算法的机理进行了研究 ,提出了遗传算子操作结果的优劣复取舍原则。并以此原则为基础 ,经过改进 ,加入调整适应度、动态调整变异概率和局部优化等方法 ,形成了优劣复取舍遗传算法。理论分析和实例计算结果表明 ,该算法能有效消除遗传算法本身带来的局部极值点 ,解决成熟前收敛的缺陷 ,与其它算法相比 ,具有收敛速度快、寻优能力强等特点。

Genetic algorithm based on better-worse re-selection

To increase the convergence speed of the genetic algorithm for real time optimum control, analysis of the genetic algorithm mechanism was used to propose a principle for selecting results generated by the genetic operators. This principle and other related methods, including adjusting fitness, dynamic adjusting mutation probability and local optimization, were used to form a genetic algorithm based on better worse re selection. Theoretical and computational results show that the modified algorithm can effectively eliminate the local convergence produced by the genetic algorithm and solve the premature convergence problem. The algorithm convergences relatively fast and effectively finds the optimal solution.