Prediction Modeling of Egg Shelf Life and Storage Time Based on Back Propagation (BP) Neural Network

Abstract

To investigate the shelf life of eggs from different chicken breeds stored at various temperatures, Haugh unit, air cell depth, yolk index, albumen pH and mass loss of eggs from Jingfen 6 and Hy-Line Grey laying hens stored under refrigerated (4 ℃) or room temperature (25 ℃) conditions were examined. Taking Haugh unit below 60 as the end of shelf life, the shelf life of eggs from both breeds was found to be 12 and 83 days under ambient and refrigerated storage conditions, respectively. To develop prediction models for egg shelf life and storage time using back propagation artificial neural network (BP-ANN), Haugh unit, the most important indicator of egg freshness, was taken as an input parameter, and the other input parameters were selected based on the results of Pearson correlation analysis and used in descending order of correlation with Haugh unit. The specific input parameters were determined based on the performance of the models on the prediction set, and the BP-ANN models with optimized number of neurons in the hidden layer were compared with the other machine learning models partial least squares regression (PLSR) and support vector regression (SVR) models. The results showed that the BP-ANN models had higher accuracy in predicting the remaining shelf life and storage time of eggs compared to the PLSR and SVR models. This study provides a reference for determining the shelf life of eggs at different storage temperatures and technical support for the rapid, accurate and simultaneous prediction of the remaining shelf life and storage time.

Keywords

• eggs; shelf life; storage time; back propagation neural network; prediction model