Optimization Of Key Technologies For Electronic Nose Instruments And Their Evaluation On Sensory Intensity Of Air Odor

As an important analytical instrument in the field of intelligent sensory science and rapid detection of food,electronic nose has been developed rapidly because of the simple operation,fast detection and analysis,high sensitivity,real-time monitoring,etc.It has been widely used in the process of food raw materials,processing and production process control,food authenticity identification,product quality evaluation,environment environmental monitoring,bio-medical and other fields.Aiming at optimization of electronic nose stability,this paper studied the electronic nose signal offset correction and the optimization of electronic nose correction based on chromatographic separation.Based on this,a correlation model between electronic nose response signal and odor gas concentration was established and applied to the monitoring of air odor.This article was discussed as the following three aspects:(1)On the basis of the situation that the data shift in varying degrees in the process of actual research and application of electronic nose,this study designed the standard calibration curve to calibrate the electronic nose data,and analyze the calibration test results.Five kinds of typical gases which are sensitive to the sensor of electronic nose were selected.Then the test results were optimized by different data preprocessing methods,obtaining the sensor with good response to the measured material and establishing the standard calibration curve.The calibration curve was used to correct the sample data collected at different times.Finally,the correction of the calibration curve was carried out by analyzing the stability and consistency of the corrected data,and the correction of the drift of the electronic nose data was achieved.The experimental results showed that the response signal of most of sensors has a good and stable linear relationship with the concentration of the standard gas after the pretreatment of the resistance data of the experimental data,and the result is better than other methods.Therefore,the resistance value transformation was selected as the data preprocessing method.By calibrating the experimental data,it was found that the variation coefficients of most sensor data were reduced from 20%to 5%,which indicated that the stability of the sample data has been greatly improved after the calibration of calibration curve.(2)The sample processing optimization system based on the chromatographic separation method and the electronic nose sensor array is was designed,and the volatile components of milk and dairy products are tested by this system,to analyze the test results by the principal component analysis algorithm and support vector machine algorithm.Different types of chromatographic packed column were added into the electronic nasal carrier gas system and six different milk and dairy samples were tested.After analysis of the results obtained from the type of chromatographic packed column for milk and dairy sample testing,the optimized electronic nose was selected.This electronic nose was used to collect the data of six kinds of milk and dairy products,and to establish the identification model of different dairy products by artificial intelligence algorithm support vector machine.Through this model,the identification and classification of unknown milk samples can be realized,laying the foundation for rapid detection of dairy products and artificial intelligence senses.The experimental results showed that the classification ability of the electronic nose was improved greatly by adding the chromatographic packed column in the electronic nasal carrier gas system,and that of adding non-polar packed column was the best for the milk dairy sample.Using the established identification model for milk and dairy products,the accuracy rate of unknown milk samples was more than 90%.(3)The construction and simple application of the analysis model of the odor strength analysis of the optimized electronic nose were studied.Sensory olfactory group was made based on the standard sniffer screening method then sensory evaluation of five kinds of odor gases such as ammonia,carbon disulfide,hydrogen sulfide,methyl mercaptan and trimethylamine were test by the group according to the reasonable odor strength grade standard.,The relationships between odor concentration,odor strength,and gas concentration and the correlation model between gas concentration and odor concentration were obtained by analyzing the sensory evaluation results,to construct the model of gas concentration and odor intensity by sensory method.Secondly,the gas samples corresponding to the concentration of the five kinds of odor gas were tested by electronic nose,and the correlation model between the response value of the electronic nose and the gas concentration was constructed.Finally,the relationship between the response value of the electronic nose and the grade intensity of the odor gas was obtained by combination of the sensory evaluation results.And the model was used to analyze the odor atmosphere.The experimental results showed that the odor strength analysis model constructed by sensory evaluation combined with electronic nose was basically accurate for the analysis of environmental gas.