Agricultural product quality analysis method and device

Abstract

The invention provides an agricultural product quality analysis method and device. The agricultural product quality analysis method comprises the steps that spectral data are acquired, and reference value data are determined; the data are divided into a correction set and an external verification set; a composite variable selection algorithm is combined with the Monte Carlo sampling method to select a key variable; and a mathematical model is established based on the selected variable, and blind sample verification is carried out by using the external verification set. Based on the method, anagricultural product quality analysis device based on a digital light processing core is developed, and comprises an optical system, a control system, a circuit system and a data storage and processing system. The device stores data collected by the optical system, extracts key variable data according to the variable selection result, calls the mathematical model to perform data calculation basedon the key variable for a sample to be tested to acquire quality data, and carries out analyzing / classifying according to the quality data. According to the invention, spectral data information can begreatly simplified; the working efficiency of an analysis instrument is improved; and a technical reference is provided for non-destructive rapid and accurate classification of the agricultural product quality.

Description

technical field [0001] The invention belongs to the field of spectral analysis, and in particular relates to a method and device for analyzing the quality of agricultural products. Background technique [0002] Spectral analysis has the technical characteristics of fast, efficient, non-destructive, and environment-friendly, and has always been an important analysis method in the field of fast and non-destructive analysis. In the field of industrial and agricultural production, multispectral analysis represented by near-infrared spectroscopy has become a popular research field in recent years. Taking near-infrared spectroscopy as an example, because it is caused by the combined and doubled frequency absorption of hydrogen-containing groups, it often presents the characteristics of mixed spectral peaks and the inability to obtain a single characteristic peak of the substance. In view of the above characteristics, traditional spectral analysis often uses full-spectrum multivar...

AI Technical Summary

Problems solved by technology

[0009] The concept of the present invention is as follows: Aiming at the problem that there is a large amount of redundant information in the spectral data, which increases the difficulty of the analysis work hardware and software, in order to realize the non-destructive and fast analysis and / or grading of the quality of agricultural products, a spectral variable selection method is provided: obtain spectral data and measure Reference value data; divide the data after removing outliers into a correction set and an external verification set; based on a series of correction data generated by the Monte-Carlo (MC) sampling method, the UVE (UniformativeVariable Elimination, non-informative variable Elimination) algorithm combined with CARS (Competitive Adaptive Reweighted Sampling, combined with competitive adaptive reweighted sampling) algorithm, the data is screened for secondary key variables and the frequency statistical analysis of the selected variables is performed, and the key variables are screened in depth. Establish a mathematical model for key variables and use an external verification set for blind sample verification; use the model comprehensive index P F Evaluate the built model; based on the key variable selection results, develop a non-destructive fast analyzer for agricultural product quality, the analyzer stores the data collected by the optical system, extracts key variable data according to the variable selection results, and calls the mathematical model for the Perform data calculation based on key variables to obtain quality data, and perform non-destructive, fast and accurate analysis and / or grading of agricultural product quality based on the quality data; thus overcoming the problems of low efficiency of agricultural product quality inspection encountered in current practical work

Images