一种用于匹配场反演的遗传算法
A new genetic algorithm for matched-field inversion

遗传算法在接近全局最优解时,存在搜索速度变慢、过早收敛、个体的多样性减少很快、甚至陷入局部最优解等问题。通过在遗传算法中引入模拟退火因子、混沌因子和多样性测度因子,在很大程度上克服了原有遗传算法的早熟、局部搜索能力差的缺点。同时,又能发挥原有遗传算法的强大的全局搜索能力,保证了改进后的混合遗传算法能较好地收敛于其全局最优值。
When approaching the global optimal solution, some shortcomings of the genetic algorithm, such as slow search speed, premature convergence, quick reduction of the diversity of individuals, and even getting into the trouble for local optimal solution, are highlighted. By introducing the simulated annealing factor, chaos factor and diversity measure factor into the genetic algorithm, the original shortcomings, such as the premature convergence and the poor local search capability, are greatly overcome, and meanwhile, the original powerful global search capability of genetic algorithm is maintained. So the hybrid genetic algorithm improved by all the measures can better converge at its global optimal value.