欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
以土壤pH、野外实测光谱以及多元散射校正（MSC）预处理后的光谱数据为基础，利用数学方法（主成分回归PCA、偏最小二成回归PLSR、BP神经网络模型）分别建立了土壤pH的预测模型。结果表明：土壤实测光谱和经过MSC方法预处理的光谱数据均与pH存在良好的相关性，并呈极显著水平，后者的相关性更高。PCA和PLSR两种土壤pH估测模型均具有良好的预测能力。BP神经网络模型则因输入变量多，预测精度较低。但利用PCA和PLSR模型所获得主成分，作为BP神经网络的输入变量所建立的复合模型，可明显提高模型稳定性和预测能力。
英文摘要：
Based on soil pH data, measured VIS-NIR reflectance and the data pretreated by multiplicative scatter correction（MSC） at the given spots, soil pH prediction models were established by using principle components a- nalysis-PCA, partial least squares regression -PLSR and back propagation -BP. The results showed that soil pH had a good correlation with both the original reflectance and the spectral data pretreated by MSC. The correction be- tween the soil pH and the spectral data pretreated by MSC was more obvious. PCA and PLSR soil pH prediction models both have good predictability on soil alkalinization. BP neural network model had a lower forecasting preci- sion because of the amount of input variables. However, using the principal components obtained from PCA and PI.SR models as input variables, the predictability and the stability of the BP neural network model can be signifi- cantly improved. Compared with PLSR, BP and PCA, the prediction results of PLSR-BP model is the best.
同期刊论文项目
同项目期刊论文
期刊信息
位置：