论文部分内容阅读
目的实现对相同型号和不同型号水果便携近红外仪进行模型传递。方法采用标准变量变换(SNV)预处理,无信息变量消除法(UVE)及与连续投影算法相结合(UVE-SPA)进行特征波长选择后,再使用直接校正法(DS)进行模型传递。结果对于相同型号的仪器K1,K2,经SNV-UVE-SPA-DS算法处理后,建立在K1上的苹果糖度校正模型成功传递到了K2上,R=0.9730,RMSEP=0.2375°Brix,结果略优于截距法;而经过SNV-UVE-DS算法,传递到与K1不同型号J仪器上的模型预测结果为R=0.9375,RMSEP=0.3458°Brix。结论合适的波长选择结合DS算法可以有效地用于相同型号及不同型号水果便携仪上的模型传递。
Objective To achieve the same model and different models of fruit near infrared portable model for transmission. Methods Using standard variable transformation (SNV) pretreatment, no information variable elimination (UVE) method and continuous projection algorithm (UVE-SPA) for feature wavelength selection, we use direct correction method (DS) for model transfer. Results For the same model instruments K1 and K2, the model based on SNV-UVE-SPA-DS algorithm was successfully transferred to K2 with R = 0.9730 and RMSEP = 0.2375 ° Brix In the intercept method, the model predicted by SNV-UVE-DS algorithm and transmitted to J instrument with K1 is R = 0.9375 and RMSEP = 0.3458 ° Brix. Conclusion The appropriate wavelength selection combined with DS algorithm can be effectively used for model transfer on the same model and different models of fruit portable instrument.