Agricultural product quality analysis method and device

Abstract

The invention provides a method and device for analyzing the quality of agricultural products. The method for analyzing the quality of agricultural products includes: acquiring spectral data and determining reference value data; dividing the data into a calibration set and an external verification set; using a composite variable selection algorithm in combination with a Monte Carlo sampling method to select key variables; establishing a mathematical model based on the selected variables and The external validation set was used for blind validation. Based on the above method, an agricultural product quality analysis device based on a digital light processing kernel is developed, including an optical system, a control system, a circuit system, and a data storage and processing system. The device stores the data collected by the optical system, extracts the key variable data according to the variable selection result, calls the mathematical model to perform data calculation based on the key variable for the sample to be tested, obtains the quality data, and performs analysis / grading according to the quality data. The invention can greatly simplify spectral data information, improve the working efficiency of analytical instruments, and provide technical reference for non-destructive, fast and accurate grading of agricultural products.

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