论文部分内容阅读
以35个实验小区,8个生育期的冬小麦田间测量冠层反射高光谱组成三维数据信息阵,采用多维偏最小二乘(multiway partial least squares,NPLS)算法建立了冬小麦亩产量估产模型。模型用于对13个实验小区产量预测分析,预测值与实际值的相关系数r为0.936 6,平均相对误差为4.44%,说明模型的预测效果较好,对实际的估产工作有一定的指导意义。
The three-dimensional data array of canopy reflectance hyperspectral data was measured in 35 experimental plots and 8 growing winter wheat fields. The multi-yield partial least squares (NPLS) algorithm was used to establish the yield estimation model of winter wheat. The model was used to predict the yield of 13 experimental plots. The correlation coefficient r between predicted and actual values was 0.936 6 and the average relative error was 4.44%, indicating that the model has a good forecasting effect and has some guiding significance for the actual yield estimation .