Analysis of influential points in decreasing curves applied to corn kernel drying

Abstract

The objective of this work was to identify and evaluate influential points for the analysis of corn kernel drying curves. Classical influential points derived from nonlinear models fitted to drying kinetics were analyzed. In addition, a new influential point was proposed to determine the time to interrupt drying, when moisture content reached 13%. The Lewis, Overhults, Page, three-parameter simple exponential, and Midilli models were applied to corn kernel drying data. The selection of the best models was based on the adjusted coefficient of determination and the Akaike information criterion. The times associated with the influential points were estimated using the confidence intervals obtained with the bias-corrected and accelerated bootstrap method. Among the evaluated models, the Page model allowed the identification of all influential points and provided a physically consistent interpretation of drying behavior. Initial influential points represented a higher kinetic efficiency, while the asymptotic deceleration point indicated transition to a stabilization regime. The proposed influential point shows narrow confidence intervals and a low model dependence, indicating robustness and operational relevance.