Quality Analysis of Table Cauarium album L. Based on Fuzzy Mathematics Sensory Evaluation, Physicochemical Properties and Electronic Tongue

Abstract

In order to comprehensively and objectively evaluate and analyze the quality of table Chinese olive (Cauarium album L.) fruits from different varieties (lines), a rapid method for the detection of the quality characteristics of table Chinese olive was established. In this study, the fruit quality of 10 different varieties (lines) of Chinese olive was evaluated by fuzzy mathematics sensory evaluation and multi-frequency pulsed electronic tongue and its correlation with physicochemical indicators was analyzed. On this basis, a quality prediction model of table Chinese olive was developed and evaluated in terms of coefficient of determination (R2) and average relative error (δ). Sensory evaluation on a nine-point hedonic scale using the Delphi method showed that the total weight of sweetness and astringency were 76.9%, both of which were the major quality characteristics of table Chinese olive fruit. The correlation analysis indicated that phenol-to-sugar ratio was significantly correlated with aftertaste sweetness and astringency (P < 0.01). According to principal component analysis (PCA), the best electrode combination for multi-frequency pulse electronic tongue to distinguish the fruit quality of different varieties of Chinese olive was palladium (10 Hz) and titanium (10 Hz). Multiple stepwise regression was used to develop a model describing the relationship between sensory evaluation score or physicochemical indicators and electronic tongue signal eigenvalues. It was found that the physicochemical indicators could be effectively predicted from the electronic tongue signal eigenvalues and the prediction performance for phenol-to-sugar ratio (R2 = 0.832, δ = 10.89%) and soluble sugar content (R2 = 0.831, δ = 5.75%) was best, followed by total phenol content (R2 = 0.783, δ = 12.08%). The sensory evaluation score could be effectively predicted by the regression model based on phenol-to-sugar ratio or electronic tongue signal eigenvalues, and the predictive effect of the model based on phenol-to-sugar ratio was better than that of the one based on tongue signal eigenvalues. However, the R2 values of the two models were low, only 0.589 and 0.542, respectively. These results provide a reference for the rapid detection of Chinese olive quality characteristics and the selection of excellent varieties.

Keywords

• cauarium album l.; fruit quality; fuzzy mathematics; sensory evaluation; electronic tongue; physicochemical properties; regression model