Prediction of Fresh Seed Melon Quality Characteristics Based on Electrical Characteristics and an ANFIS Model

This study aimed to analyze the relationships between the electrical properties of fresh seed melon pulp and storage conditions, thereby providing a rapid electrical method for quality evaluation. The electrical parameters of fresh seed melon were measured using the parallel-plate electrode method at different storage temperatures of 4 °C, 8 °C, 12 °C, 16 °C, and 20 °C, and storage times of 0, 2, 4, 6, and 8 h. The relationships between electrical parameters and quality attributes at different frequencies were further investigated. An adaptive neuro-fuzzy inference system (ANFIS) model was established to predict the quality characteristics of fresh seed melon, with electrical parameters used as input variables and quality characteristics used as output variables. Eight membership function models were constructed and compared to select the optimal prediction model. The results showed that, with increasing frequency, the impedance (Z), capacitance (Cp), and resistance (Rp) of fresh seed melon decreased, while conductance (G) and reactance (X) increased at different storage temperatures. At different storage times, Z, quality factor (Q), Cp, and Rp decreased, while G and X increased with increasing frequency. The variation ranges of Z, Cp, Rp, G, and X gradually decreased at higher frequencies. Significant correlations between electrical parameters and quality characteristics were observed at the characteristic test frequency of 163.28 kHz. The ANFIS results showed that gauss2mf was the optimal model for predicting cohesiveness (Co, R2 = 0.9491), pimf was the optimal model for predicting chewiness (Ch, R2 = 0.9595), and gbellmf was the optimal model for predicting resilience (Re, R2 = 0.9596). These results indicate that the combination of electrical properties and the ANFIS model has potential for evaluating the quality characteristics of fresh seed melon under the present experimental conditions. This study provides a theoretical basis for quality detection and storage preservation of fresh seed melon and offers a reference for the development of rapid electrical quality-evaluation techniques for seed melon pulp.