Fast non-destructive detection method and device of food smell based on gas sensor array technology

Abstract

Some sample of relevant food is estimated by some professionals with their sensory organ or tested through conventional physical and chemical analysis to establish correlative data base. Some other sample is determined with sensor array through carrying the smell of the food to the sensor array reaction chamber via carrier gas, amplifying and filtering the signal from the sensor array, A/D conversion in A/D converter and data processing in computer. The result of computer processing is related with those in the established data base and one knowledge library is obtained through learning forthe determination of truth and quality of tested sample. The said method is fast and simple, needs no pre-treatment of the sample and organic solvent extraction, and has the artificial intelligent identification to unknown sample. The present invention facilitates the physical and chemical analysis of food products.

Description

Technical field [0001] The invention relates to a detection method and device for food odor, in particular to a fast nondestructive detection method and device for food aroma and flavor based on gas sensor array and pattern recognition. Background technique [0002] Many of the unique qualities and characteristics of foods can be traced to the chemicals that create their odors. To date, odor assessment has been done by trained professionals, but individual assessments vary widely, are highly subjective, and in some cases involve health and safety issues. In recent years, gas chromatography is the most widely used method for the determination of food odor. Its determination requires preparation and processing of samples, selection of appropriate extraction solvents, and appropriate chromatographic separation conditions. The analysis time usually ranges from several tens of minutes to one hour. Because the concentration of volatile matter is generally relatively low, detecti...

AI Technical Summary

Problems solved by technology

Since the odor concentration emitted by food is generally all relatively low, the inventions described above are used to identify the volatilized odor of food and have the following problems: 1) the environment tested is not controlled during detection, and in an open test environment We know that the gas sensor is particularly sensitive to the composition, temperature, humidity, and partial pressure of oxygen in the surrounding test environment, so the sensor's response to the odor emitted by the same food is a value in the morning, and a value at noon, not to mention There is a difference between winter and summer, and in different places, the partial pressure of oxygen in the air environment is different and the composition is different, so the noise generated will be larger than the required signal

(2) Feature extraction is simple, only staying in the time domain to take feature values, and some only take stable values, which will waste a lot of useful information

(3) The analysis method used is too simple, most of which use error backpropagation BP neural network, but this kind of network has limited convergence ability and recognition ability, and can only identify several simple gases with large differences, which is not suitable for detection food volatile odor

[0004] On the device, most of the above several patents use the vacuum pump intermittent suction device. This test device looks like an animal nose, but it wi

Images