论文部分内容阅读
本文以肥东县土壤墒情为例,选用BP神经网络模型,对土壤墒情预测进行探究。先介绍神经网络相关理论知识,然后针对模型建立需要确定的参数做了详细说明。着重比较了隔日和五日两种预测结果的精度,结果表明隔日预报精度要高于五日预测精度,得出结论 BP神经网络短期土壤墒情预报能取得较高的精度,对农业生产具有实际指导意义。
In this paper, the soil moisture in Feidong County is taken as an example, BP neural network model is selected to investigate the soil moisture prediction. First introduce the neural network related theoretical knowledge, and then set the parameters needed to determine the model to do a detailed description. The precision of the prediction results of every other day and that of the fifth is emphatically compared. The results show that the forecast accuracy of every other day is higher than the forecast accuracy of the fifth day, and the conclusion is drawn that the short-term soil moisture prediction of BP neural network can obtain higher precision and has practical guidance for agricultural production significance.