Prediction and optimization of fruit quality of peach based on artificial neural network

Abstract

Mineral nutrition is the material basis for growth and quality improvement of fruit trees. In this study, artificial neural networks (ANNs) and Multiple linear regression (MLR) models were used to study the influence of mineral nutrient elements in fruits on the peach fruit quality. The results showed that four established ANN models of the fruit quality with the Log-sigmoid transfer function and the Levenberg-Marquardt back-propagation training function can obtain the highest accuracy (R2 =0.9243, 0.9260, 0.9536, and 0.9644 respectively). The results of the sensitivity and redundancy analysis indicated that the nitrogen, phosphorus, potassium, magnesium, and iron content in fruits had the greatest contribution to these fruit quality indicators of the peach, which was of great significance to the fertilization of peach orchards and the improvement of peach fruit quality.