现行的电容谷物水分仪只使用一种形式的数学模型来拟合不同谷物的水分，从而导致测量误差不够稳定。基于对我国主产谷物的水分的大量测试和统计研究，本文证明不同的谷物或植物种子表现为形式不同的最佳拟合方程。从而为较高性能的水分仪的软件设计提供新的理论依据。 Current capacitive grain moisture analyzers use only one type of mathematical model to fit the moisture of different grains, resulting in less stable measurement errors. Based on a large number of tests and statistical studies on the moisture content of grain produced in China, this paper proves that different grain or plant seeds behave as best fit equations with different forms. Thus providing a new theoretical basis for software design of higher performance moisture meter.