System and method for detecting milk freshness based on electronic nose

Abstract

The invention discloses a system and a method for detecting milk freshness based on an electronic nose and aims to overcome the defects in quick detection of the raw milk freshness. The system comprises a flavor fingerprint information acquisition module, a signal preprocessing module, a character extracting and screening module and a pattern recognition module, wherein the flavor fingerprint information acquisition module is used for reading voltage signals responding to gas of a gas sensitive array into an internal memorizer in a digital manner, implementing real-time digital display and waveform display of the read data, and storing and recycling the displayed data; the signal preprocessing module is used for preprocessing the original signals by using a filter function, for example, removing noise of the original signals and extracting basic signals; the characteristic extracting and screening module is used for extracting characteristics of the filtered and de-noised signals and screening the characteristics according to a principal component analysis method; the pattern recognition module is used for training and identifying the detected samples according to a pattern recognition algorithm. The system and the method are used for quickly detecting the raw milk freshness based on the electronic nose and have guiding significance to control of flavor and quality of milk products. Compared with the conventional method, the method provided by the invention is more efficient and convenient, faster, and lower in manpower cost.

Description

technical field [0001] The invention relates to a detection system and method based on an electronic nose, in particular to a milk freshness detection system and method based on an electronic nose. Background technique [0002] It is very important to detect flavor substances in milk and determine the source of raw milk and its products for the quality control of dairy products. At present, the chemical methods used by most enterprises to detect the flavor of raw milk and products are cumbersome and have poor real-time performance. The physical methods are expensive and require high technical requirements, which are not suitable for rapid analysis and difficult to meet the requirements of the current food industry. [0003] Electronic nose technology is a brand-new gas identification technology. It uses a gas sensor array, combined with neural network, support vector machine and other pattern recognition algorithms, to achieve qualitative or quantitative analysis of each gas...

AI Technical Summary

Problems solved by technology

At present, the chemical methods used by most enterprises to detect the flavor of raw milk and products are cumbersome and have poor real-time performance. The physical methods are expensive and have high technical requirements, which do not meet the needs of rapid analysis and are difficult to meet the requirements of the current food industry.

Images