Optical method for predicting total soluble solids in pears using radial basis function networks

摘要

Radial basis function networks (RBFN) have been widely used for function approximation and pattern classification as an alternative to conventional artificial neural networks. In this paper, reflectance spectroscopy and chemical measurements of total soluble solids (TSS)content were used to develop a nondestructive technique for predicting the TSS and a relationship was also established between the TSS content in pears determined by diffuse reflectance spectra (4200-12500cm~(-1)) and by chemical measurements. The effectiveness of the radial basis function networks of nonlinear calibration model was presented and compared with the linear algorithms of the partial least squares calibration models. The results show that the relatively coefficient of determination (r) of prediction obtained with linear partial least squares and the nonlinear radial basis function networks are 0.72, 0.83 and the root mean square error of prediction are 0.49, 0.45 respectively. Our results revealed that the calibration model of radial basis function networks produced better prediction of TSS than the model of partial least squares when the samples consist of multi-components.