Mass modeling of potato cultivars with different shape index by physical characteristics

摘要

The correlation between physical parameters like linear dimensions, projected areas,volumes, and mass of potato cultivars is imperative for predicting the quality besidesthe development of post-harvest machineries especially grading systems. Therefore,this investigation was envisaged to determine the correlation between the mass andproperties like dimensions viz. length (l), width (w), thickness (t), geometric mean diameter(Gmd), first projected area (FPA), second projected area (SPA), third projected area(TPA), criteria area (Cae), oblate spheroid volume (Vobsp), ellipsoid spheroid volume(Vellsp), and shape index (SI) of potato cultivars cv. Milva, Jelly, and Sante. Based on theSI, potato tubers were classified as round (100–160), oval (161–240), and long (241–340), respectively. The predictive modeling was done using 171 linear regressionmodels and the models having the highest coefficient of determination (R2) and lowestregression standard error (RSE) and root mean square error (RMSE) were recommended.A total of 27 model equations based on dimensions and projected area wererecommended for the estimation of the mass of all three potato tubers. These modelequations find application for developing an effective grading setup, further augmentingits prospective utilization. Results revealed that the linear models based on lwt(m = k_1l + k_2w + k_3t + k_4) were recommended for all the SI of the cultivars with R~2varied from .942 to .965 (cv. Milva), .949 to .975 (cv. Jelly), and .946 to .956(cv. Sante). The regression models based on projected area (m = k_1FPA + k_2SPA +-k_3TPA + k_4) were recommended with R2 varied from .956 to .974 (cv. Milva), .959 to.982 (cv. Jelly), and .957 to .977 (cv. Sante). The detailed information about the recommendedmass models based on the engineering properties of potatoes could be imperativefor the efficient design of an integrated and automated grading system.