2.4.2. Model Update Method

To improve the applicability of the model, a sample addition algorithm was proposed to update the model. Combined with wavelength position offset properties, wavelength fusion was performed. The KS algorithm calculates the Euclidean distance and adds the two samples with the largest distance to the calibration set. Then, the distance minimum (Dm) value between each remaining sample and the selected calibration set sample is calculated. The sample with the largest Dm is added to the calibration set until the calibration set reaches the specified number of samples. Therefore, samples with large spectral differences between different cultivation modes can be extracted for use as a calibration set using the KS algorithm. In this study, based on the Euclidean distance between samples, the KS algorithm was used to select new samples in turn, and the RMSECV of the established PLSR model was calculated.
