水文模型参数的快速率定是山洪以及中小河流洪水预报预警中的重要研究内容之一。参数自动优选功能可以极大地提高水文模型的使用效率,随机搜索算法是大多数参数全局优选算法基础,但却因为耗时较长而应用较少。以Brooks的自适应随机搜索...水文模型参数的快速率定是山洪以及中小河流洪水预报预警中的重要研究内容之一。参数自动优选功能可以极大地提高水文模型的使用效率,随机搜索算法是大多数参数全局优选算法基础,但却因为耗时较长而应用较少。以Brooks的自适应随机搜索算法(adaptive random search method)为对象、利用.NET的Parallel对象进行了CPU并行改造,利用英伟达的CUDA对象进行了GPU+CPU并行改造,并以缅甸境内其培河子流域上的新安江模型为优选对象,比较了优选效果和计算效率。研究表明:ARS算法和SCE-UA的优选结果相当,并行改造后的ARS算法计算效率有显著提高。研究成果对水文模型应用时参数优选算法的比选具有重要的参考价值。展开更多
Genetic algorithm is a widely used optimization method. Crossover and mutation are two Basicl operatorsof the genetic algorithm. On the basis of analyzing the principles of simple genetic algorithm and discussing its ...Genetic algorithm is a widely used optimization method. Crossover and mutation are two Basicl operatorsof the genetic algorithm. On the basis of analyzing the principles of simple genetic algorithm and discussing its exist-ing problems of crossover point and mutation bit, this paper presents a way of the adaptive multiple bit mutation ge-netic algorithm , which not only can keep the population diversity but also has quicker convergence speed. The resultsof the multi-modal function optimization show that the adaptive multiple bit mutation genetic algorithm is practicaland efficient.展开更多
文摘水文模型参数的快速率定是山洪以及中小河流洪水预报预警中的重要研究内容之一。参数自动优选功能可以极大地提高水文模型的使用效率,随机搜索算法是大多数参数全局优选算法基础,但却因为耗时较长而应用较少。以Brooks的自适应随机搜索算法(adaptive random search method)为对象、利用.NET的Parallel对象进行了CPU并行改造,利用英伟达的CUDA对象进行了GPU+CPU并行改造,并以缅甸境内其培河子流域上的新安江模型为优选对象,比较了优选效果和计算效率。研究表明:ARS算法和SCE-UA的优选结果相当,并行改造后的ARS算法计算效率有显著提高。研究成果对水文模型应用时参数优选算法的比选具有重要的参考价值。
文摘Genetic algorithm is a widely used optimization method. Crossover and mutation are two Basicl operatorsof the genetic algorithm. On the basis of analyzing the principles of simple genetic algorithm and discussing its exist-ing problems of crossover point and mutation bit, this paper presents a way of the adaptive multiple bit mutation ge-netic algorithm , which not only can keep the population diversity but also has quicker convergence speed. The resultsof the multi-modal function optimization show that the adaptive multiple bit mutation genetic algorithm is practicaland efficient.