应用遗传算法拟合浑水入渗经验公式  

Fitted empirical formula for muddy water infiltration using genetic algorithm

作　　者：白丹[1] 李占斌[1] 

机构地区：[1]西安理工大学水资源研究所,西安710048

出　　处：《农业工程学报》2003年第3期76-79,共4页Transactions of the Chinese Society of Agricultural Engineering

摘　　要：探索浑水入渗规律 ,对浑水灌溉、流域降雨产汇流、土壤侵蚀机理的研究与应用具有一定的价值。该文首先分析了影响浑水入渗主要因素的基础上 ,提出了浑水入渗试验方案。根据浑水入渗试验资料 ,认为入渗时间和物理性粘粒含量是影响浑水入渗的主要因素 ,引入粘性指数来反映浑水中物理性粘粒含量的大小 ,在此基础上 ,提出了浑水入渗经验公式。应用遗传算法 ,以浑水入渗经验公式中待求参数为优化变量 ,采用二进制编码 ,拟合这些待求参数。分析结果表明 :拟合的浑水入渗经验公式精度较高 ;浑水与清水入渗的差异主要是由浑水中物理性粘粒的存在引起的 ;浑水中物理性粘粒含量越大 ,累积入渗量就越小。The experimental design plans were determined on analyzing main factors to affect infiltration capability of muddy water. Based on the experimental data, it is considered that the main factors are infiltration time and soil clay content, the concept of soil clay index was presented to denote soil clay content in muddy water, and muddy water infiltration expression was presented. The concerned parameters in the expression are regarded as the optimal variables and determined by binary system code method of the genetic algorithm. It showed that muddy water infiltration expression was relatively accurate, and the reason for muddy water infiltration differing from rinsing was the clay in muddy water, and the more the soil clay index was, the less the accumulation infiltration amount was.

关 键 词：浑水 入渗规律 粘性指数 遗传算法 

分 类 号：S275[农业科学—农业水土工程] TV142[农业科学—农业工程]