A method for predicting the comprehensive quality of dried apple products based on apple raw material indexes

Abstract

The invention discloses a method for predicting the comprehensive quality of a dried product based on an apple raw material index. The method comprises the following steps of determining an evaluationindex of the dried product and establishing a corresponding relationship between the evaluation index of the dried product and the comprehensive quality; determining an apple raw material evaluationindex related to the dry product evaluation index; obtaining the apple raw material evaluation index data of the training sample, obtaining the dry product evaluation index data of the training sample, calculating the comprehensive quality of each training sample, taking the apple raw material evaluation index data of the training sample as the input layer, taking the comprehensive quality as theoutput layer, training to obtain the neural network learning model; using the neural network learning model and the apple raw material evaluation index data of the dried product to be tested to predict the comprehensive quality of the dried product to be tested. The map relationship between the input and the output is not determined in advance, so that the method of the invention has the advantageof fast prediction speed and high precision.

Description

technical field [0001] The invention relates to the related field of apple processing. More specifically, the present invention relates to a method for predicting the comprehensive quality of dried products based on apple raw material indexes. Background technique [0002] Apple ranks first among the four major fruits in the world. my country is the largest producer and consumer of apples in the world. Apples are widely planted in my country and are also rich in varieties. In addition to being used as fresh food, apples can also be squeezed into juice and made into dried products. However, the current supply of fresh food is saturated, and the development momentum of processing and transformation is insufficient. Diversified processing technologies are urgently needed. As a new type of snack food, dried apple products are delicious and healthy, can enrich the apple processing industry, drive the processing and transformation of raw materials, and have good development prosp...

AI Technical Summary

Problems solved by technology

At present, the research on the comprehensive quality evaluation of products mostly uses AHP, gray correlation and other methods for separate analysis. For example, Wu Houjiu et al. used the percentage method to judge the quantitative evaluation method of sweet orange variety processing suitability, and Wang Yongyang et al. predicted grape quality based on the apparent characteristics. Sensory quality of table grapes, such methods can only evaluate the processing characteristics of the variety studied, and cannot predict the processing performance of unknown raw material samples

Existing methods for correlating raw material characteristics and product quality only use linear models such as discriminant function and linear regression. Dealing with the complex nonlinear correspondence between indicators has limitations, the scope of application is small, and the accuracy of predicting the processing performance of unknown samples is low.

Images