基于遗传算法的堤坝反压优化设计

Optimizing anti-press design of dike and dam based on genetic algorithm

堤坝反压设计采用试算法或图解法时,存在试算工作量巨大或图表繁杂等局限,精度受限。针对上述局限性,尝试采用具有高度并行、随机、自适应搜索的遗传算法建立堤坝反压优化设计模型。该方法和模型参数不发生直接联系,其求解过程中控制一个解群,从而大大提高了搜索效率,并且可以避免陷入局部极值。实例分析表明,该模型精度较高,为堤坝反压优化设计提出了一种全新算法。

Trial and error method or diagrammatic method was used to anti?press design of dike and dam. The former was very heavy in working load, the latter was miscellaneous in chart and table, and thus led to restrict the precision. Aiming at the localization, and taking advantages of high parallel, random and adaptive searching of genetic algorithm(GA), the model of optimizing anti?press design based on GA is established. Without direct relation to parameters, an outcome group of the model is controlled in seeking result, thus the method not only is high in searching efficiency, but also can avoid running into local extremum. The study through an illustration shows that the model is high in precision. Thus a new arithmetic is put forward for optimizing anti?press design of dike and dam.