研究目的:工程水文地质参数是工程优化设计的基础。针对传统工程水文参数获取存在人为辨识误差 和求参耗时的缺陷,开发遗传算法模型,解决求参问题。
研究方法:利用某大型地下工程降水抽水试验取得的实测资料,采用遗传算法反演了工程地质层的渗透 系数,并与两点法计算结果进行了对比。
研究结果:只要能给出正演计算的模型,形成遗传反演系统问题,就能够直接求解,计算的结果与传统计 算方法接近。
研究结论:遗传算法作为一种全局最优化方法,具有简单、快捷的特点,计算效率高、无须人为干扰、能够 快速搜索到可行性解的优点,可广泛用于工程水文地质参数的求解。
Research purposes: Hydrogeological parameter is one of the most important factors for project optimization design. Considering the generic parameter identification existing time consuming and unavoidable man-made error, the author applies genetic algorithm to identify the hydrogeological parameter to overcome the limitation of traditional method.
Research methods: Basing on the actual pumping test data measured in the field, inversion-analysis program is used which is based on the genetic algorithms to identify the hydrogeological parameters. The results obtained by genetic algorithms and two-pointes method are compared.
Research results: The results have indicated that two kinds of parameter inversion method can get the similar results.
Research conclusions: As one kind of optimization methods , the genetic algorithm is simple , universal and also has the characteristic of good adaptability and high efficiency on parameter inversion, and which can be widely employed to hydrogeological parameter inversion.