Apple Quality Evaluation Based on Entropy Weight Method, Grey Relational Degree Method and Low-field Nuclear Magnetic Resonance Detection

Abstract

To study the quality characteristics of different apple varieties and establish a comprehensive evaluation model of apple quality, taking five varieties of apples (Tianshui Huaniu, Aksu Tangxin, Marshal Huang, Cream Fuji, and Luochuan Red Fuji) as the research object, the four texture characteristics, including hardness, adhesion, chewability, cohesion, and four physical and chemical indicators, including water content, titratable acid (TA), soluble sugar (SS), and soluble solid content (SSC) were tested. Combining the low-field nuclear magnetic resonance detection technology, the correlation between the water distribution and the physicochemical and texture characteristics of apple was explored, and the main indicators for evaluating apple quality were established by principal component analysis. Based on the entropy weight method, each core index was given weight, and a grey correlation degree evaluation model was established. The results showed that there were significant differences in various indexes of different varieties of apples (P<0.05), and there was a high correlation between their water distribution and texture characteristics and physical and chemical indexes. The spin-spin relaxation time T22 (immobilized water), T21 (bound water) and TA, SS, SSC were established as the core indexes. The weight calculated by entropy weight method showed that the sum of T22 and T21 was 35.31%, accounting for the largest proportion, indicating that the water distribution had the greatest impact on apple quality. The grey correlation analysis showed that the quality of Tianshui Huaniu and Aksu Tangxin was better. The method adopted in this study could quickly and accurately establish the quality evaluation model of apples, and provide a new method for the quality evaluation of fruits and vegetables including apples.

Keywords

• apple
• entropy weight method
• grey relational analysis
• low-field nuclear magnetic resonance
• principal component analysis
• quality evaluation