改进的AGA及其在约束函数优化中的应用

提出了一种改进的基于自适应惩罚函数的AGA。一方面，采用分裂选择算子，增加了潜在优良个体的生存概率；另一方面，引入基于优势遗传的交叉概率和变异概率，防止了算法的早熟收敛。此外，应用改进的最优保存策略，保证了算法的收敛性和收敛解的有效性。通过对约束函数优化的仿真计算，证明该算法具有快速收敛和鲁棒性好的特点。

Improved adaptive genetic algorithm and its application in constrained function optimization

An Adaptive Genetic Algorithm（AGA） based on Adaptive Penalty Function（AGA-APF） has been proposed.On the one hand,the disruptive selection operator is employed to enhance the survival probability of potential better individuals in the population.On the other hand,the probabilities of crossover and mutation based on superiority inheritance are introduced so as to prevent the algorithm from premature convergence.What＇s more,the improved optimal reserved strategy is applied in AGA-APF,which can guarantee the convergence of the algorithm and validity of the convergent solutions.The simulation results of the con strained function optimization have demonstrated that AGA-APF can converge to optimal solutions rapidly and own higher robustness.