Analysis and prediction of storability and storage quality of pear fruit based on clustering and time series analysis

Abstract

Objective: The storability and storage quality of three pear varieties were evaluated to provide theoretical support for variety breeding and fresh-keeping technology. Methods: Zaoshengxinshui, HJ18 and HJ67 pear with similar genetic background were used as test materials. Fruit with the same maturity were stored in a cold storage with a temperature of （1.0±0.5） ℃ and a relative humidity of 85%~90% for 60 days. The fruit texture, sugar and acid and volatile substances were determined every 10 days, divide storage period and quality based on clustering, principal components and time series analysis. Results: The results showed that the fruit texture, sugar, acid and volatile substances of the three pear varieties were regulated by storage time, and there was a certain correlation among different indexes. Fourteen quality evaluation parameters could be simplified into 8 indexes: firmness, chewiness, compactness, brittleness, elasticity, malic acid, citric acid, soluble solids content （SSC） and fructose. The contribution rate of sensors S2, 6, 7, 8 and 9 was larger. Based on fourteen quality indexes, HJ67 was in a relatively independent space with Zaosheng Xinshui and HJ18 during storage. Based on the response values of ten sensors, the three varieties had spatial intersection, which had not been effectively distinguished. The storage time of three pear varieties was divided into different intervals, among which HJ67 was divided into four periods: 0~20 days, 30 days, 40~50 days and 60 days based on texture, nonvolatile and volatile substances. Conclusion: It is concluded that the texture, sugars and acids contents and volatile matter of HJ67 are significantly different from those of Zaoshengxinshui and HJ18, that has a rich flavor. The change of fruit texture of three pear varieties can be well fitted by exponential equation, and the change of sugar and acid content can be well described by nonlinear equation. The prediction errors of the two prediction models are low. Based on principal component analysis, cluster analysis and time series analysis, the storage period and comprehensive quality of three pear varieties can be comprehensively and objectively evaluated.

Keywords

• pear
• time series analysis
• cluster analysis
• storability
• storage quality