论文部分内容阅读
以改良盐碱土壤、提供入渗参数为研究目的,在山西省北部的4种盐碱荒地进行了系列入渗试验和基本理化参数测定试验。基于误差反向传播算法(Back Propagation算法),建立了盐碱地土壤基本理化参数与Philip入渗模型参数之间的神经网络预报模型。预测所得Philip入渗模型参数的平均相对误差如下:稳渗率A为4.30%、吸渗率S为0.31%,预测值与实测值吻合程度高。研究结果表明,基于盐碱地土壤条件,选择土壤体积含水率、容重、质地、有机质含量、全盐量以及p H作为预报模型输入变量,Philip入渗模型参数为输出变量的BP神经网络的预报模型是可行的。
In order to improve the saline-alkali soil and provide infiltration parameters, a series of infiltration experiments and the determination of basic physical and chemical parameters were conducted in four kinds of saline-alkali wasteland in northern Shanxi Province. Based on the backpropagation algorithm, a neural network forecasting model was established between the soil salinization soil basic parameters and the Philip infiltration model parameters. The average relative error of predicted Philip infiltration model parameters is as follows: the steady infiltration rate A is 4.30%, the infiltration rate S is 0.31%, and the predicted value is in good agreement with the measured value. The results showed that the prediction model of BP neural network with Philip infiltration model parameters as output variables was selected based on the soil conditions of saline-alkali soil, soil moisture content, bulk density, texture, organic matter content, total salinity and p H as input variables. feasible.