Near-Infrared Spectroscopy Analysis For Simple Sysrup And Molasses Of Sugarcane

The optimization model of near infrared spectroscopy analysis for brix and apparent purity of simple syrup and molasses of sugarcane were established by near infrared spectroscopy (NIR) and chemometrics method, respectively. For optimization of spectrum wavelengths combination (continuous band or discrete wavelengths combination), and study the stability of the model.Firstly, some methods of NIR transmission, diffuse reflectance spectroscopy experiment for simple syrup and molasses of sugarcane were studied.Then, the calibration set and prediction set was divided into based on the predictive effect of the best single wavelength prediction bias(SWPB) model with all samples, making the calibration set and the prediction set is similar to a model representation. And each of the indicators (Brix and simple purity) to meet the conditions of the simple syrup and molasses of sugarcane was given in 50 division of calibration set and prediction set, and the different division was modeled to discuss the calibration and prediction model on the stability of the calibration set and prediction set partitioning. On this condition, the calibration models were established by single-point regression, direct PLS, PLS with Savitzky-Golay smoothing, moving window PLS, and moving window MLR with fixed interval, respectively. The optimum objective was the root mean square error of predication (RMSEP).For optimization of spectrum wavelengths combination (continuous band or discrete wavelengths combination), and study the stability of the model. Of which the optimal model prediction RRMSEP the values of simple syrup brix,simple syrup apparent purity, molasses brix and molasses apparent purity of sugarcane were 0.94%,0.83%,1.61%, 3.33%, respectively. They provided an important basis for developing the small-scale dedicated analytical instruments.