【摘要】 为了有效地克服非线性、多参数、多极值的组合优化问题中遗传算法的早熟收敛现象，引入生物的小生境技术和自然界生物多种杂交方式并存、多个子群体并行以及灾变现象而提出了一种综合的快速基因搜索策略。该算法以多个子群体不同杂交方式并行搜索为主，并结合排序配对的受限交配技术和灾变算子等方法来实现快速的寻优过程。对地震资料波阻抗反演的理论模型试验表明：该算法的收敛能力和计算效率有了明显改善。更多还原

【Abstract】 hybrid fast genetic algorithm (H2O) is proposed on the basis of various characteristics of nature evolution. This algorithm is combined with niche technician, a variety of crossover schemes, multi-populations and catastrophism phenomenons. Last sense, it can efficiently eliminate premature convergence of GA on the nonlinear multi-parameter optimization problem. Through testing seism wave impedance inversion, the result is obviously improved compared with the convention genetic algorithm in the convergence ability and computation efficiency.更多还原