Polymer product odor detection method, device and electronic nose system

Through the combination of gas chromatography analysis and machine learning model, the problem of insufficient polymer odor distinction in the existing technology is solved, and accurate and reliable distinction of polymer product odors is achieved, which improves the accuracy and reliability of polymer odor analysis. .

CN114527198BActive Publication Date: 2024-11-22CHINA PETROLEUM & CHEMICAL CORP +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202011190832.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-10-30
Publication Date
2024-11-22
Estimated Expiration
2040-10-30

AI Technical Summary

Technical Problem

The existing electronic nose technology has shortcomings in the differentiation of polymer odors, and it is impossible to accurately and reliably distinguish similar polymer products odors. The sensor array has a single response characteristics, and the existing odor recognition algorithm has a narrow applicability.

Method used

The odor data of polymer products were extracted by gas chromatography analysis method, and analyzed through pre-trained odor analysis model. Combined with machine learning algorithms such as support vector machines and neural networks, odor attributes and grade models are constructed to improve the distinction of polymer odors. .

Benefits of technology

A richer, accurate and reliable distinction of odor substances emitted by polymer products is achieved, and the odor contribution of long-chain alkanes, alcohols, aldehydes, ketones and other substances with different molecular weights can be effectively distinguished, and the distinction of polymer odor is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114527198B_ABST
    Figure CN114527198B_ABST
Patent Text Reader

Abstract

The present invention discloses a method, device and electronic nose system for detecting the odor of a polymer product. The method comprises: collecting the gas to be tested emitted by the polymer product to be tested; performing gas chromatography analysis on the gas to be tested to obtain a gas chromatogram of the gas to be tested; extracting the odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested; inputting the odor data of the polymer product to be tested into a pre-trained odor analysis model for analysis, and outputting the detection information of the polymer product to be tested, so as to achieve separation of the odor substances emitted by the polymer product using a chromatographic method, obtain more abundant, accurate and reliable odor data, and effectively distinguish the contribution of long-chain alkanes, alcohols, aldehydes, ketones and other substances with different molecular weights emitted by the polymer product to the odor. The technical solution of the present invention can accurately and reliably improve the differentiation of polymer odors.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of electronic noses, and in particular relates to a method and equipment for detecting the odor of a polymer product and an electronic nose system. Background Art

[0002] Polymers usually have excellent comprehensive properties and are inexpensive, and are widely used in various products in daily life. With the improvement of living standards and the enhancement of environmental awareness, consumers are paying more and more attention to the odor produced by polymer products, and their tolerance is getting lower and lower. In application fields such as packaging, storage, automobiles and building materials, the odor of polymer raw materials and products has become one of the important indicators that all parties pay attention to. It can be said that the size of the odor largely determines the quality and experience of the product, and directly affects the competitiveness of the product in the market. Therefore, it is very necessary to develop and improve fast and accurate odor identification and evaluation methods for polymers to control product odor from the source.

[0003] Possible sources of polymer odor include residual catalysts and unreacted products, various additives used to improve performance, and related aging and degradation products, etc., with complex compositions. It is very difficult to accurately qualitatively and quantitatively analyze such complex odor substances using traditional instrumental analysis methods. It is also difficult to associate odor substances with odor magnitude, and it is impossible to directly compare the odor of materials. At present, the odor evaluation method for polymers is still mainly based on manual olfactory sensory analysis, that is, trained olfactors rate and describe the materials based on the odor they smell. The sensory analysis method is simple and intuitive in principle, but its results are highly subjective and random, with poor repeatability and low test efficiency.

[0004] Electronic nose is a new type of odor analysis technology. Its analysis process is similar to that of biological sensory systems. First, sensors are used to extract the characteristic information of mixed volatile gases for processing. Then, pattern recognition algorithms are used to establish an analysis model to achieve the identification of odor information. Electronic nose has the advantages of simple and fast analysis and low cost. It has begun to be used in food, medicine, environment and other fields.

[0005] However, the existing electronic nose technology is not established for the odor characteristics of polymers, and there are many problems in evaluating polymers. First, most electronic noses use multiple gas sensors such as metal oxide semiconductors (MOS) to collect data. For polymers with similar odor types, only one or a few sensors often contribute greatly to the results. The response characteristics of the sensor array are single, and the degree of differentiation of polymer odors is not high; secondly, the existing odor recognition algorithms, such as principal component analysis (PCA) and linear discriminant analysis (LDA), cannot finely distinguish similar polymer products, and can only distinguish odor samples with large differences, and their applicability is narrow.

[0006] Therefore, how to accurately and reliably improve the differentiation of polymer odors is a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention

[0007] The main purpose of the present invention is to provide a polymer product odor detection method, device and electronic nose system to accurately and reliably improve the differentiation of polymer odors.

[0008] In view of the above problems, the present invention provides a method for detecting the odor of a polymer product, comprising:

[0009] Collect the gas to be tested emitted by the polymer product to be tested;

[0010] Performing gas chromatography analysis on the gas to be tested to obtain a gas chromatogram of the gas to be tested;

[0011] Extracting odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested;

[0012] The odor data of the polymer product to be tested is input into a pre-trained odor analysis model for analysis, and the detection information of the polymer product to be tested is output.

[0013] Furthermore, in the above-mentioned polymer product odor detection method, the training process of the odor analysis model includes the following steps:

[0014] The current analysis model is trained based on the odor data of the training sample to obtain a current optimized analysis model; the current analysis model is a pre-trained model or a previous optimized analysis model obtained after the previous training; wherein the odor data of the training sample includes artificial odor evaluation data of the training sample, and gas chromatography analysis of the sample gas emitted by the training sample to obtain a gas chromatogram of the sample gas, and odor data extracted based on the gas chromatogram of the sample gas;

[0015] The current optimized analysis model is tested using the odor data of the test sample to obtain the current test result; wherein the odor data of the test sample includes artificial odor evaluation data of the test sample, and gas chromatography analysis of the test gas emitted by the test sample to obtain a gas chromatogram of the test gas, and odor data extracted based on the gas chromatogram of the test gas;

[0016] If the current test result indicates that the preset standard condition is met, the odor analysis model is determined according to the current optimization model;

[0017] If the current test result indicates that the preset standard condition is not met, the next training is performed until the next test result indicates that the preset standard condition is met.

[0018] Furthermore, in the above-mentioned polymer product odor detection method, the odor analysis model includes an odor attribute model and an odor level model; correspondingly, the current analysis model includes a current attribute model and a current level model;

[0019] The current analysis model is trained based on the odor data of the training sample to obtain the current optimized analysis model, including:

[0020] Based on a support vector machine or a neural network algorithm, the current attribute model is trained to obtain a current optimized attribute model;

[0021] Based on the support vector regression algorithm or the partial least squares regression algorithm, the current grade model is trained to obtain the current optimized grade model;

[0022] The current optimization attribute model and the current optimization level model are combined to obtain a current optimization analysis model.

[0023] Furthermore, in the above-mentioned polymer product odor detection method, the test sample is obtained by traversing and selecting one group of original samples from the K groups of original samples after the original samples are divided into K groups; the training sample is obtained by traversing and selecting K-1 groups of original samples from the K groups of original samples;

[0024] Correspondingly, the current optimization models are K, and determining the odor analysis model according to the current optimization model includes:

[0025] Based on the K current optimized models, the hyperparameters of the models are selected or optimized, the odor analysis model is constructed, and the accuracy of the odor analysis model is evaluated.

[0026] Furthermore, in the above-mentioned polymer product odor detection method, the artificial odor evaluation data is the odor perception attribute description and intensity level obtained by the odorist evaluating the training samples and the test samples according to the prescribed method.

[0027] Furthermore, in the above-mentioned polymer product odor detection method, extracting odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested includes:

[0028] Based on the gas chromatogram of the gas to be tested, extracting the overall response information and / or the peak information of each gas chromatogram as the odor data of the polymer product to be tested;

[0029] The overall response information is used to describe information about all emitted gaseous substances contained in the polymer product to be tested; and the spectrum peak information is used to describe concentration information about each emitted gaseous substance of the polymer product to be tested.

[0030] Furthermore, in the above-mentioned polymer product odor detection method, based on the gas chromatogram of the gas to be tested, extracting the overall response information and / or peak information of each gas chromatogram as the odor data of the polymer product to be tested, comprises:

[0031] Based on the clustering algorithm, the spectral peaks of the polymer products to be tested are clustered to obtain chromatographic peaks representing the same odor substance in different polymer products to be tested; based on the chromatographic peak of each odor substance, the overall response information and / or spectral peak information of each gas chromatograph is extracted as the odor data of the polymer products to be tested.

[0032] Furthermore, in the above-mentioned polymer product odor detection method, the detection information of the polymer product to be tested includes the odor attributes and / or odor evaluation level of the polymer product to be tested; the odor attributes include at least one of odor category, odor label and odor description information.

[0033] The present invention also provides a polymer product odor detection device, comprising a memory and a processor;

[0034] The memory stores a computer program, and when the computer program is executed by the processor, the steps of the method for detecting a polymer product as described in any one of the above items are implemented.

[0035] The present invention also provides an electronic nose system, comprising:

[0036] A sampling device, used to collect the gas to be tested emitted by the polymer product to be tested;

[0037] A chromatographic separation device, used for performing gas chromatography analysis on the gas to be tested to obtain a gas chromatogram of the gas to be tested;

[0038] A detection device, used for extracting odor data of the polymer product to be tested based on a gas chromatogram of the gas to be tested;

[0039] The odor analysis device is used to input the odor data of the polymer product to be tested into a pre-trained odor analysis model for analysis, and output detection information of the polymer product to be tested.

[0040] Compared with the prior art, one or more embodiments of the above scheme may have the following advantages or beneficial effects:

[0041] The polymer product odor detection method, device, and electronic nose system of the present invention perform gas chromatography analysis on the gas to be tested emitted by the collected polymer product to be tested, obtain the gas chromatogram of the gas to be tested, extract the odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested, input the odor data of the polymer product to be tested into a pre-trained odor analysis model for analysis, and output the detection information of the polymer product to be tested, so as to achieve the separation of odor substances emitted by polymer products using chromatographic methods, obtain more abundant, accurate, and reliable odor data, and effectively distinguish the contribution of long-chain alkanes, alcohols, aldehydes, ketones, and other substances with different molecular weights emitted by polymer products to the odor. The technical solution of the present invention can accurately and reliably improve the differentiation of polymer odors.

[0042] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0044] Figure 1 This is a flow chart of an embodiment of a method for detecting a polymer product of the present invention;

[0045] Figure 2 Flowchart of the training process of the odor analysis model;

[0046] Figure 3 Schematic diagram of the test chromatogram of a sample and the characteristic chromatographic peaks selected by the clustering method;

[0047] Figure 4 This is a comparison chart between the rating results of the odor analysis model and the artificial odor rating;

[0048] Figure 5is a schematic structural diagram of an electronic nose system embodiment of the present invention;

[0049] Figure 6 It is a schematic structural diagram of an embodiment of a polymer product odor detection device of the present invention. DETAILED DESCRIPTION

[0050] The following will describe the implementation methods of the present invention in detail with reference to the accompanying drawings and embodiments, so that the implementation process of how the present invention applies technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly. It should be noted that as long as there is no conflict, the various embodiments of the present invention and the various features in the embodiments can be combined with each other, and the technical solutions formed are all within the protection scope of the present invention.

[0051] Embodiment 1

[0052] In order to solve the above technical problems existing in the prior art, an embodiment of the present invention provides a method for detecting a polymer product.

[0053] Figure 1 The following is a flow chart of an embodiment of the method for detecting a polymer product of the present invention, as shown in FIG. Figure 1 As shown, the detection method of the polymer product of this embodiment may specifically include the following steps:

[0054] 100. Collect the gas to be tested emitted by the polymer product to be tested;

[0055] In this embodiment, the polymer product to be tested may be pre-treated to make the polymer product to be tested emit gas, which may be referred to as the gas to be tested. In this embodiment, 3 to 5 parallel polymer products to be tested are preferred.

[0056] Specifically, a certain weight of the polymer product to be tested can be placed in a clean, odorless glass bottle or air bag, heated and shaken in an environment of 60°C to 120°C for 30min to 120min. The purpose of treating the polymer product to be tested at high temperature is to accelerate the diffusion of odor substances from the surface. The specific temperature and time selected are related to the type and purpose of the product.

[0057] After the polymer product to be tested emits the gas to be tested, sampling can be performed using a headspace bottle / vacuum bottle / gas bag thermal desorption method to obtain the gas to be tested emitted by the polymer product to be tested.

[0058] 101. Performing gas chromatography analysis on the gas to be tested to obtain a gas chromatogram of the gas to be tested;

[0059] In this embodiment, after collecting the gas to be tested emitted by the polymer product to be tested, the mixture can be separated by a chromatographic column using the differences in boiling points, polarities and adsorption properties of the substances, thereby obtaining a gas chromatogram of the gas to be tested.

[0060] It should be noted that if the polymer product to be tested contains aldehyde and ketone odor substances, solid phase adsorption / air bag injection can be used for analysis by high performance liquid chromatography.

[0061] 102. Extracting odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested;

[0062] Specifically, based on the gas chromatogram of the gas to be tested, the overall response information and / or the spectrum peak information of each gas chromatogram can be extracted as the odor data of the polymer product to be tested. The overall response information is used to describe the information of all emitted gas substances contained in the polymer product to be tested; and the spectrum peak information is used to describe the concentration information of each emitted gas substance of the polymer product to be tested.

[0063] In a specific implementation process, the retention time of each peak in the gas chromatogram of the gas to be tested corresponds to different emitted gas substances. The response information of each gas chromatogram describes the emitted gas substance contained in the polymer product to be tested, which is proportional to the concentration of the substance. The data can be organized into s i :

[0064] s i ∈R p ,i=1,...,n

[0065] Among them, s i is the response information of the gas chromatograph, n is the number of polymer products to be tested, and p is the number of data points of the gas chromatogram curve.

[0066] This data contains complete information about the substances emitted by the sample and can be used as the odor data of the polymer product to be tested. In order to reduce the amount of calculation of the odor analysis model, integrate the collected data, and reduce the interference of background noise in the spectrum, further process the data to calculate the peak information of each chromatographic peak, where the peak information includes retention time, peak height or peak area. It can be obtained:

[0067] t i ∈R q ,i=1,...n

[0068] Among them, t i is the peak height or peak area of ​​each chromatographic peak, n is the number of polymer products to be tested, and q is the number of different chromatographic peaks in the spectrum.

[0069] Since the retention time of the same gas substance will vary within a certain range during each sampling process, in this embodiment, a clustering method, such as a hierarchical clustering method, is used to automatically cluster the spectral peaks of the polymer product to be tested, so as to obtain chromatographic peaks representing the same odor substance in different polymer products to be tested, and based on the chromatographic peak of each odor substance, the overall response information and / or spectral peak information of each gas chromatograph is extracted as the odor data of the polymer product to be tested, thereby eliminating the need for tedious and difficult qualitative and quantitative work of the substance, simplifying the analysis process and improving efficiency.

[0070] When rating the odor perception intensity of the tested polymer products, the response information and peak information of each gas chromatograph are linearly related to the concentration of the odor substance, while the odor perception intensity is power-law related to the concentration of the odor substance. i Processing to obtain odor data x i :x i =log(t i ),i=1,...,n.

[0071] 103. Input the odor data of the polymer product to be tested into a pre-trained odor analysis model for analysis, and output detection information of the polymer product to be tested.

[0072] In a specific implementation process, the training process of the odor analysis model can refer to Figure 2 The flowchart shown in FIG. Figure 2 The flowchart of the training process of the odor analysis model is as follows: Figure 2 As shown, the training process of the odor analysis model may specifically include the following steps:

[0073] 200. Training the current analysis model based on the odor data of the training sample to obtain a current optimized analysis model;

[0074] In this embodiment, the odor data of the training sample is obtained by performing gas chromatography analysis on the gas emitted by the training sample (referred to as sample gas herein) to obtain a gas chromatogram of the sample gas, and the odor data is extracted based on the gas chromatogram of the sample gas.

[0075] Specifically, representative polymer products can be collected as samples. The number of samples should be sufficient and can represent the possible types and property variation ranges of the unknown polymer products to be measured. The samples are pretreated: a certain weight of polymer products are placed in a clean and odorless glass bottle or air bag, heated and shaken in an environment of 60°C to 120°C, and treated for 30min to 120min. The purpose of treating the samples at high temperature is to accelerate the diffusion rate of odor substances out of the surface. The specific temperature and time selected are related to the type and purpose. When training the odor analysis model, each training sample should include 5 to 10 parallel samples to determine the variation range of the sample data and the repeatability of the test, to ensure that the collected data truly reflects the characteristics of the sample. The sample gas emitted by the pretreated training sample is subjected to gas chromatography analysis to obtain the gas chromatogram of the sample gas, and the odor data of the training sample is extracted based on the gas chromatogram of the sample gas.

[0076] In this embodiment, the process of extracting the odor data of the training sample based on the gas chromatogram of the sample gas is the same as the process of extracting the odor data of the polymer product to be tested. Please refer to the above-mentioned related records for details, which will not be repeated here.

[0077] In practical applications, the training samples can be divided into two groups. The first group is subjected to subjective evaluation of artificial odor levels, and the second group extracts odor data of the training samples in the above manner. The two groups of training samples are matched one by one to label the training samples.

[0078] In this embodiment, 5 to 10 odorists are preferably used to evaluate the gas emitted by the pre-treated training samples to obtain the odor intensity and odor description of the training samples. Odorists cannot test continuously for a long time and should ensure sufficient recovery and rest time; the training samples should be tested repeatedly until the odor intensity and odor description obtained meet the error requirements; each round of testing should hide the labels of the training samples and shuffle the order for blind testing. The obtained subjective evaluation criteria for odor levels can be shown in Table 1:

[0079] Table 1

[0080]

[0081] After extracting the odor data of the training sample, the current analysis model is input for training to obtain the current optimized analysis model. If it is the first time to extract the odor data of the training sample, the current analysis model is the preset training model, that is, the initial model. If it is not the first time to extract the odor data of the training sample, the current analysis model is the last optimized analysis model obtained after the last training.

[0082] It should be noted that, in this embodiment, the odor analysis model includes an odor attribute model and an odor level model; correspondingly, the current analysis model includes a current attribute model and a current level model. In this embodiment, the current attribute model can be trained based on a support vector machine or a neural network algorithm to obtain a current optimized attribute model; the current level model can be trained based on a support vector regression algorithm or a partial least squares regression algorithm to obtain a current optimized level model; the current optimized attribute model and the current optimized level model are combined to obtain a current optimized analysis model. .

[0083] In this embodiment, nonlinear machine learning models such as support vector machines, neural network algorithms, support vector regression algorithms, or partial least squares regression algorithms are used, which are more in line with the characteristics of the odor of polymer products, can better describe the odor differences between samples, have high recognition accuracy, and can objectively and uniformly evaluate the odor of polymer products.

[0084] 201. Testing the current optimization analysis model using the odor data of the test sample to obtain a current test result;

[0085] In this embodiment, the odor data of the test sample is obtained by performing gas chromatography analysis on the gas emitted by the test sample (referred to as test gas herein) to obtain a gas chromatogram of the sample gas, and extracting the odor data based on the gas chromatogram of the test gas.

[0086] Specifically, representative polymer products can be collected as samples. The number of samples should be sufficient and can represent the possible types and property variation ranges of the unknown polymer products to be tested. The samples are pretreated: a certain weight of polymer products is placed in a clean and odorless glass bottle or air bag, heated and shaken in an environment of 60°C to 120°C, and treated for 30min to 120min. The purpose of treating the samples at high temperature is to accelerate the diffusion rate of odor substances out of the surface. The specific temperature and time selected are related to the type and purpose. When testing the odor analysis model, each test sample should include 5 to 10 parallel samples to determine the variation range of the sample data and the repeatability of the test, to ensure that the collected data truly reflects the characteristics of the sample. The sample gas emitted by the pretreated test sample is subjected to gas chromatography analysis to obtain the gas chromatogram of the sample gas, and the odor data of the test sample is extracted based on the gas chromatogram of the sample gas.

[0087] In this embodiment, the process of extracting the odor data of the test sample based on the gas chromatogram of the sample gas is the same as the process of extracting the odor data of the polymer product to be tested. Please refer to the above-mentioned related records for details, which will not be repeated here.

[0088] In practical applications, the test samples can be divided into two groups. The first group is subjected to subjective evaluation of artificial odor levels, and the second group extracts odor data of the test samples in the above manner. The two groups of test samples are matched one by one to mark the test samples.

[0089] After extracting the odor data of the test sample, the odor data of the test sample is input into the current optimization analysis model for testing to obtain the current test result.

[0090] 202. Determine whether the current test result meets the preset standard condition. If yes, execute step 203. If no, return to step 200.

[0091] In this embodiment, it can be determined whether the current test result meets the preset standard condition. For example, the current test result that can be obtained can be the correlation and / or deviation value with the labeling information of the test sample. If the value of the correlation with the labeling information of the test sample is greater than or equal to the first threshold, it can be considered that the current test result indicates that the preset standard condition is met. If the value of the correlation with the labeling information of the test sample is less than the first threshold, it can be considered that the current test result indicates that the preset standard condition is not met. Alternatively, if the deviation value with the labeling information of the test sample is greater than or equal to the second threshold, it can be considered that the current test result indicates that the preset standard condition is met. If the value of the correlation with the labeling information of the test sample is less than the second threshold, it can be considered that the current test result indicates that the preset standard condition is not met. If the current test result indicates that the preset standard condition is met, step 203 can be executed. If the current test result indicates that the preset standard condition is not met, return to step 200 and perform the next training until the next test result indicates that the preset standard condition is met, and the training is stopped.

[0092] 203. Determine the odor analysis model according to the current optimization model.

[0093] If the current test result indicates that the preset compliance condition is met, the odor analysis model is determined according to the current optimization model.

[0094] In practical applications, a large amount of data is often required when training an odor analysis model. In this embodiment, when the amount of data is small, a cross-validation training method can be used to train the odor analysis model. Therefore, in this embodiment, the test sample is obtained by traversing and selecting a group of original samples from the K groups of original samples after the original samples are divided into K groups; the training sample is obtained by traversing and selecting K-1 groups of original samples from the K groups of original samples. For example, K can be 4, so that the first 3 groups of original samples are used as training samples, and the 4th group of original samples are used as test samples, and the same method is used to traverse all original samples, thereby increasing the number of training samples, and after multiple training and testing, the obtained odor analysis model is more accurate and reliable.

[0095] After the above cross-validation training, the current optimization model of the verification training can be obtained. In this way, when the odor analysis model is determined according to the current optimization model, the hyperparameters of the model can be selected or optimized based on K current optimization models, the odor analysis model can be constructed, and the accuracy of the odor analysis model can be evaluated.

[0096] For example, the detection information of the polymer product to be tested includes the odor attribute and / or odor evaluation level of the polymer product to be tested; the odor attribute includes at least one of odor category, odor label and odor description information.

[0097] The detection method of the polymer product of this embodiment performs gas chromatography analysis on the gas to be tested emitted by the collected polymer product to be tested, obtains the gas chromatogram of the gas to be tested, extracts the odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested, and inputs the odor data of the polymer product to be tested into a pre-trained odor analysis model for analysis, outputs the detection information of the polymer product to be tested, and realizes the separation of odor substances emitted by the polymer product using a chromatographic method, obtains richer, more accurate and reliable odor data, and can effectively distinguish the contribution of long-chain alkanes, alcohols, aldehydes, ketones and other substances with different molecular weights emitted by the polymer product to the odor. The technical solution of the present invention can accurately and reliably improve the differentiation of polymer odors.

[0098] It should be noted that the method of the embodiment of the present invention can be performed by a single device, such as a computer or a server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present invention, and the multiple devices will interact with each other to complete the method.

[0099] Embodiment 2

[0100] The technical solution of the present invention is described below with specific examples:

[0101] Select 17 representative impact copolymer polypropylene products. Follow the steps below to establish the odor analysis model:

[0102] (1) Obtaining electronic nose odor feature data: Each training sample includes 5 parallel samples. Take 1g of pellets and place them in a 20mL headspace bottle. Heat and shake at 80℃ for 60min. Use an automatic sampler to extract the headspace gas. Use thermal desorption / capillary gas chromatography / hydrogen flame ionization detector to analyze. Mark the retention time of each chromatographic peak and record the peak height.

[0103] The hierarchical clustering shortest distance algorithm was used, and the retention time difference threshold was set to 0.1 s. All sample chromatographic peaks were grouped so that the chromatographic peaks representing the same odor substance on different samples were used as an odor feature dimension, such as Figure 3 As shown, we get t i ∈R q ,i=1,...n. Among them, Figure 3 Schematic diagram of the test chromatogram of a sample and the characteristic chromatographic peaks selected by the clustering method.

[0104] (2) With reference to the odor test standard PV3900 method, five odorists evaluated the odor level of the pretreated samples. The subjective evaluation criteria are shown in Table 1. Half-point values ​​can be given, and the average value of the odor level is taken as the result. The test is repeated until the deviation of the average value of the odor level is no more than 0.1 level.

[0105] (3) With t i , i=1,...,n as training vector data, polymer product brand as label, linear discriminant analysis model and support vector machine model are established respectively, where the support vector machine uses RBF kernel function exp(-γ||xx′|| 2 ), the value of γ is 1 / (n·var), where var is the variance of the training data. The kernel function can map the original features to a high-dimensional space, and can project some features that are linearly inseparable in low dimensions to a linearly separable space, thereby improving the accuracy of recognition. Using the cross-validation algorithm, some samples are randomly selected as validation data, and the remaining samples are used as training data. The accuracy of the two classification models is shown in Table 2. The linear discriminant analysis model cannot accurately distinguish several samples with similar properties, while the support vector machine model has a higher recognition accuracy.

[0106] Table 2

[0107]

[0108] (4) log(t i),i=1,...,n as training vector data and artificial odor level as target value, a support vector regression model is established, and the RBF kernel function is also used for nonlinear mapping. Figure 4 The difference between the rating results of the electronic nose analysis model and the artificial odor rating was compared, and the correlation coefficient between the two was 0.9, and the maximum deviation was 0.1. The model can better evaluate the odor properties of this category of polypropylene. Figure 4 This is a comparison chart between the odor analysis model rating results and the manual odor rating.

[0109] Embodiment 3

[0110] In order to solve the above technical problems existing in the prior art, an embodiment of the present invention provides an electronic nose system.

[0111] Figure 5 FIG. 1 is a schematic diagram of the structure of an electronic nose system embodiment of the present invention. Figure 5 As shown, the electronic nose system of this embodiment may include a sampling device 50 , a chromatographic separation device 51 , a detection device 52 and an odor analysis device 53 .

[0112] The sampling device 50 is used to collect the gas to be tested emitted by the polymer product to be tested;

[0113] In this embodiment, the polymer product to be tested may be pre-treated to make the polymer product to be tested emit gas, which may be referred to as the gas to be tested. In this embodiment, 3 to 5 parallel polymer products to be tested are preferred.

[0114] Specifically, a certain weight of the polymer product to be tested can be placed in a clean, odorless glass bottle or air bag, heated and shaken in an environment of 60°C to 120°C for 30min to 120min. The purpose of treating the polymer product to be tested at high temperature is to accelerate the diffusion of odor substances from the surface. The specific temperature and time selected are related to the type and purpose of the product.

[0115] After the polymer product to be tested emits the gas to be tested, sampling can be performed using a headspace bottle / vacuum bottle / gas bag thermal desorption method to obtain the gas to be tested emitted by the polymer product to be tested.

[0116] A chromatographic separation device 51 is used to perform gas chromatography analysis on the gas to be tested to obtain a gas chromatogram of the gas to be tested;

[0117] In this embodiment, after collecting the gas to be tested emitted by the polymer product to be tested, the mixture can be separated by a chromatographic column using the differences in boiling points, polarities and adsorption properties of the substances, thereby obtaining a gas chromatogram of the gas to be tested.

[0118] It should be noted that if the polymer product to be tested contains aldehyde and ketone odor substances, solid phase adsorption / air bag injection can be used for analysis by high performance liquid chromatography.

[0119] A detection device 52, for extracting odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested;

[0120] Specifically, based on the gas chromatogram of the gas to be tested, the overall response information and / or spectral peak information of each gas chromatogram can be extracted as the odor data of the polymer product to be tested; wherein the overall response information is used to describe the information of all emitted gas substances contained in the polymer product to be tested; and the spectral peak information is used to describe the concentration information of each emitted gas substance of the polymer product to be tested. .

[0121] In a specific implementation process, the retention time of each peak in the gas chromatogram of the gas to be tested corresponds to different emitted gas substances. The response information of each gas chromatogram describes the emitted gas substance contained in the polymer product to be tested, which is proportional to the concentration of the substance. The data can be organized into s i :

[0122] s i ∈R p ,i=1,...,n

[0123] Among them, s i is the response information of the gas chromatograph, n is the number of polymer products to be tested, and p is the number of data points of the gas chromatogram curve.

[0124] This data contains complete information about the substances emitted by the sample and can be used as the odor data of the polymer product to be tested. In order to reduce the amount of calculation of the odor analysis model, integrate the collected data, and reduce the interference of background noise in the spectrum, further process the data to calculate the peak information of each chromatographic peak, where the peak information includes retention time, peak height or peak area. It can be obtained:

[0125] t i ∈R q ,i=1,...n

[0126] Among them, t i is the peak height or peak area of ​​each chromatographic peak, n is the number of polymer products to be tested, and q is the number of different chromatographic peaks in the spectrum.

[0127] Since the retention time of the same gas substance will vary within a certain range during each sampling process, in this embodiment, a clustering method, such as a hierarchical clustering method, is used to automatically cluster the spectral peaks of the polymer product to be tested, so as to obtain chromatographic peaks representing the same odor substance in different polymer products to be tested, and based on the chromatographic peak of each odor substance, the overall response information and / or spectral peak information of each gas chromatograph is extracted as the odor data of the polymer product to be tested, thereby eliminating the need for tedious and difficult qualitative and quantitative work of the substance, simplifying the analysis process and improving efficiency.

[0128] When rating the odor perception intensity of the tested polymer products, the response information and peak information of each gas chromatograph are linearly related to the concentration of the odor substance, while the odor perception intensity is power-law related to the concentration of the odor substance. i Processing to obtain odor data x i :x i =log(t i ),i=1,...,n.

[0129] The odor analysis device 53 is used to input the odor data of the polymer product to be tested into a pre-trained odor analysis model for analysis, and output detection information of the polymer product to be tested.

[0130] In this embodiment, after the odor data of the polymer product to be tested is obtained, the odor data of the polymer product to be tested can be input into a pre-trained odor analysis model for analysis, so that the detection information of the polymer product to be tested can be output.

[0131] For example, the detection information of the polymer product to be tested includes the odor attribute and / or odor evaluation level of the polymer product to be tested; the odor attribute includes at least one of odor category, odor label and odor description information.

[0132] The polymer product detection device 52 of this embodiment performs gas chromatography analysis on the gas to be tested emitted by the collected polymer product to be tested, obtains the gas chromatogram of the gas to be tested, extracts the odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested, inputs the odor data of the polymer product to be tested into a pre-trained odor analysis model for analysis, and outputs the detection information of the polymer product to be tested, thereby realizing the separation of odor substances emitted by polymer products using chromatographic methods, obtaining richer, more accurate and reliable odor data, and being able to effectively distinguish the contribution of long-chain alkanes, alcohols, aldehydes, ketones and other substances with different molecular weights emitted by polymer products to the odor. The technical solution of the present invention can accurately and reliably improve the differentiation of polymer odors.

[0133] Furthermore, in the above embodiment, the training process of the odor analysis model is as follows:

[0134] The current analysis model is trained based on the odor data of the training sample to obtain the current optimized analysis model; the current optimized analysis model is tested using the odor data of the test sample to obtain the current test result; if the current test result indicates that the preset standard conditions are met, the odor analysis model is determined according to the current optimized model; if the current test result indicates that the preset standard conditions are not met, the next training is performed until the next test result indicates that the preset standard conditions are met.

[0135] In this embodiment, the current analysis model is a pre-trained model or a last optimized analysis model obtained after the last training; wherein, the odor data of the training sample includes artificial odor evaluation data of the training sample, and, gas chromatography analysis is performed on the sample gas emitted by the training sample to obtain a gas chromatogram of the sample gas, and odor data is extracted based on the gas chromatogram of the sample gas; wherein, the odor data of the test sample includes artificial odor evaluation data of the test sample, and, gas chromatography analysis is performed on the test gas emitted by the test sample to obtain a gas chromatogram of the test gas, and odor data is extracted based on the gas chromatogram of the test gas.

[0136] In practical applications, a large amount of data is often required when training an odor analysis model. In this embodiment, when the amount of data is small, a cross-validation training method can be used to train the odor analysis model. Therefore, in this embodiment, the test sample is obtained by traversing and selecting a group of original samples from the K groups of original samples after the original samples are divided into K groups; the training sample is obtained by traversing and selecting K-1 groups of original samples from the K groups of original samples. For example, K can be 4, so that the first 3 groups of original samples are used as training samples, and the 4th group of original samples are used as test samples, and the same method is used to traverse all original samples, thereby increasing the number of training samples, and after multiple training and testing, the obtained odor analysis model is more accurate and reliable.

[0137] After the above cross-validation training, the current optimization model of the verification training can be obtained. In this way, when the odor analysis model is determined according to the current optimization model, the hyperparameters of the model can be selected or optimized based on K current optimization models, the odor analysis model can be constructed, and the accuracy of the odor analysis model can be evaluated.

[0138] It should be noted that the apparatus of the above embodiment is used to implement the corresponding method in the above embodiment and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0139] Embodiment 4

[0140] In order to solve the above technical problems existing in the prior art, an embodiment of the present invention provides a polymer product odor detection device.

[0141] Figure 6 FIG. 1 is a schematic diagram of a polymer product odor detection device according to an embodiment of the present invention. Figure 6 As shown, the device may include: a processor 1010 and a memory 1020. As known to those skilled in the art, the device may also include an input / output interface 1030, a communication interface 1040 and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030 and the communication interface 1040 are connected to each other in communication within the device through the bus 1050.

[0142] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0143] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0144] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0145] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).

[0146] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0147] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.

[0148] In a specific implementation process, a computer program is stored on the memory of the polymer product odor detection device provided in this embodiment, and when the computer program is executed by the processor, the following steps are implemented:

[0149] Collect the gas to be tested emitted by the polymer product to be tested;

[0150] Performing gas chromatography analysis on the gas to be tested to obtain a gas chromatogram of the gas to be tested;

[0151] Extracting odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested;

[0152] The odor data of the polymer product to be tested is input into a pre-trained odor analysis model for analysis, and the detection information of the polymer product to be tested is output.

[0153] Furthermore, when the computer program is executed by the processor, the following steps may be implemented:

[0154] The current analysis model is trained based on the odor data of the training sample to obtain a current optimized analysis model; the current analysis model is a pre-trained model or a previous optimized analysis model obtained after the previous training; wherein the odor data of the training sample includes artificial odor evaluation data of the training sample, and gas chromatography analysis of the sample gas emitted by the training sample to obtain a gas chromatogram of the sample gas, and odor data extracted based on the gas chromatogram of the sample gas;

[0155] The current optimized analysis model is tested using the odor data of the test sample to obtain the current test result; wherein the odor data of the test sample includes artificial odor evaluation data of the test sample, and gas chromatography analysis of the test gas emitted by the test sample to obtain a gas chromatogram of the test gas, and odor data extracted based on the gas chromatogram of the test gas;

[0156] If the current test result indicates that the preset standard condition is met, the odor analysis model is determined according to the current optimization model;

[0157] If the current test result indicates that the preset standard condition is not met, the next training is performed until the next test result indicates that the preset standard condition is met.

[0158] Furthermore, the odor analysis model includes an odor attribute model and an odor evaluation model; correspondingly, the current analysis model includes a current attribute model and a current evaluation model; when the computer program is executed by the processor, the following steps may also be implemented:

[0159] Based on a support vector machine or a neural network algorithm, the current attribute model is trained to obtain a current optimized attribute model;

[0160] Based on the support vector regression algorithm or the partial least squares regression algorithm, the current grade model is trained to obtain the current optimized grade model;

[0161] The current optimization attribute model and the current optimization level model are combined to obtain a current optimization analysis model.

[0162] Furthermore, the test sample is obtained by traversally selecting one group of original samples from the K groups of original samples after dividing the original samples into K groups; the training sample is obtained by traversally selecting K-1 groups of original samples from the K groups of original samples;

[0163] When the computer program is executed by the processor, the following steps may also be performed:

[0164] Based on the K current optimized models, the hyperparameters of the model are selected or optimized, the odor analysis model is constructed, and the accuracy of the odor analysis model is evaluated.

[0165] Furthermore, the artificial odor evaluation data is the description of odor perception attributes and intensity levels obtained by olfactory detectives evaluating training samples and test samples according to a prescribed method.

[0166] Furthermore, when the computer program is executed by the processor, the following steps may be implemented:

[0167] Based on the gas chromatogram of the gas to be tested, extracting the overall response information and / or the peak information of each gas chromatogram as the odor data of the polymer product to be tested;

[0168] The overall response information is used to describe information about all emitted gaseous substances contained in the polymer product to be tested; and the spectrum peak information is used to describe concentration information about each emitted gaseous substance of the polymer product to be tested.

[0169] Furthermore, when the computer program is executed by the processor, the following steps may be implemented:

[0170] Based on the clustering algorithm, the spectral peaks of the polymer products to be tested are clustered to obtain chromatographic peaks representing the same odor substance in different polymer products to be tested; based on the chromatographic peak of each odor substance, the overall response information and / or spectral peak information of each gas chromatograph is extracted as the odor data of the polymer products to be tested.

[0171] Furthermore, in the above embodiment, the detection information of the polymer product to be tested includes the odor attribute and / or odor evaluation level of the polymer product to be tested; the odor attribute includes at least one of odor category, odor label and odor description information.

[0172] Embodiment 5

[0173] In order to solve the above technical problems existing in the prior art, an embodiment of the present invention provides a storage medium.

[0174] The storage medium of this embodiment stores a computer program, and when the computer program is executed by the processor, the following steps are implemented:

[0175] Collect the gas to be tested emitted by the polymer product to be tested;

[0176] Performing gas chromatography analysis on the gas to be tested to obtain a gas chromatogram of the gas to be tested;

[0177] Extracting odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested;

[0178] The odor data of the polymer product to be tested is input into a pre-trained odor analysis model for analysis, and the detection information of the polymer product to be tested is output.

[0179] Furthermore, when the computer program is executed by the processor, the following steps may be implemented:

[0180] The current analysis model is trained based on the odor data of the training sample to obtain a current optimized analysis model; the current analysis model is a pre-trained model or a previous optimized analysis model obtained after the previous training; wherein the odor data of the training sample includes artificial odor evaluation data of the training sample, and gas chromatography analysis of the sample gas emitted by the training sample to obtain a gas chromatogram of the sample gas, and odor data extracted based on the gas chromatogram of the sample gas;

[0181] The current optimized analysis model is tested using the odor data of the test sample to obtain the current test result; wherein the odor data of the test sample includes artificial odor evaluation data of the test sample, and gas chromatography analysis of the test gas emitted by the test sample to obtain a gas chromatogram of the test gas, and odor data extracted based on the gas chromatogram of the test gas;

[0182] If the current test result indicates that the preset standard condition is met, the odor analysis model is determined according to the current optimization model;

[0183] If the current test result indicates that the preset standard condition is not met, the next training is performed until the next test result indicates that the preset standard condition is met.

[0184] Furthermore, the odor analysis model includes an odor attribute model and an odor evaluation model; correspondingly, the current analysis model includes a current attribute model and a current evaluation model; when the computer program is executed by the processor, the following steps may also be implemented:

[0185] Based on a support vector machine or a neural network algorithm, the current attribute model is trained to obtain a current optimized attribute model;

[0186] Based on the support vector regression algorithm or the partial least squares regression algorithm, the current grade model is trained to obtain the current optimized grade model;

[0187] The current optimization attribute model and the current optimization level model are combined to obtain a current optimization analysis model.

[0188] Furthermore, the test sample is obtained by traversally selecting one group of original samples from the K groups of original samples after dividing the original samples into K groups; the training sample is obtained by traversally selecting K-1 groups of original samples from the K groups of original samples;

[0189] When the computer program is executed by the processor, the following steps may also be performed:

[0190] Based on the K current optimized models, the hyperparameters of the models are selected or optimized, the odor analysis model is constructed, and the accuracy of the odor analysis model is evaluated.

[0191] Furthermore, the artificial odor evaluation data is the description of odor perception attributes and intensity levels obtained by olfactory detectives evaluating training samples and test samples according to a prescribed method.

[0192] Furthermore, when the computer program is executed by the processor, the following steps may be implemented:

[0193] Based on the gas chromatogram of the gas to be tested, extracting the overall response information and / or the peak information of each gas chromatogram as the odor data of the polymer product to be tested;

[0194] The overall response information is used to describe information about all emitted gaseous substances contained in the polymer product to be tested; and the spectrum peak information is used to describe concentration information about each emitted gaseous substance of the polymer product to be tested.

[0195] Furthermore, when the computer program is executed by the processor, the following steps may be implemented:

[0196] Based on the clustering algorithm, the spectral peaks of the polymer products to be tested are clustered to obtain chromatographic peaks representing the same odor substance in different polymer products to be tested; based on the chromatographic peak of each odor substance, the overall response information and / or spectral peak information of each gas chromatograph is extracted as the odor data of the polymer products to be tested.

[0197] Furthermore, in the above embodiment, the detection information of the polymer product to be tested includes the odor attribute and / or odor evaluation level of the polymer product to be tested; the odor attribute includes at least one of odor category, odor label and odor description information.

[0198] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0199] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0200] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0201] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0202] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0203] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module 32, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0204] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0205] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0206] Although the embodiments disclosed in the present invention are as above, the contents described are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. A method for detecting the odor of a polymer product, characterized in that: include: Collect the gas to be tested emitted by the polymer product to be tested; Performing gas chromatography analysis on the gas to be tested to obtain a gas chromatogram of the gas to be tested; Extracting odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested, comprising: extracting overall response information and / or peak information of each gas chromatogram as the odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested; The overall response information is used to describe the information of all emitted gas substances contained in the polymer product to be tested, which is proportional to the concentration of the substance. The data can be organized into s i : s i ∈R p ,i=1,...,n Among them, s i is the response information of the gas chromatograph, n is the number of polymer products to be tested, and p is the number of data points of the gas chromatogram curve; The spectral peak information is used to describe the concentration information of each emitted gas substance of the polymer product to be tested, and the spectral peak information of each chromatographic peak is calculated, wherein the spectral peak information includes the retention time, and the peak height or peak area, and the following can be obtained: t i ∈R q ,i=1,...n Among them, t i is the peak height or peak area of ​​each chromatographic peak, n is the number of polymer products to be tested, and q is the number of different chromatographic peaks in the spectrum; The odor data of the polymer product to be tested is input into a pre-trained odor analysis model for analysis, and the detection information of the polymer product to be tested is output.

2. The method for detecting odor of a polymer product according to claim 1, characterized in that: The training process of the odor analysis model includes the following steps: The current analysis model is trained based on the odor data of the training sample to obtain a current optimized analysis model; the current analysis model is a pre-trained model or a previous optimized analysis model obtained after the previous training; wherein the odor data of the training sample includes artificial odor evaluation data of the training sample, and gas chromatography analysis of the sample gas emitted by the training sample to obtain a gas chromatogram of the sample gas, and odor data extracted based on the gas chromatogram of the sample gas; The current optimized analysis model is tested using the odor data of the test sample to obtain the current test result; wherein the odor data of the test sample includes artificial odor evaluation data of the test sample, and gas chromatography analysis of the test gas emitted by the test sample to obtain a gas chromatogram of the test gas, and odor data extracted based on the gas chromatogram of the test gas; If the current test result indicates that the preset standard condition is met, the odor analysis model is determined according to the current optimization model; If the current test result indicates that the preset standard condition is not met, the next training is performed until the next test result indicates that the preset standard condition is met.

3. The method for detecting odor of a polymer product according to claim 2, characterized in that: The odor analysis model includes an odor attribute model and an odor level model; correspondingly, the current analysis model includes a current attribute model and a current level model; The current analysis model is trained based on the odor data of the training sample to obtain the current optimized analysis model, including: Based on a support vector machine or a neural network algorithm, the current attribute model is trained to obtain a current optimized attribute model; Based on the support vector regression algorithm or the partial least squares regression algorithm, the current grade model is trained to obtain the current optimized grade model; The current optimization attribute model and the current optimization level model are combined to obtain a current optimization analysis model.

4. The method for detecting odor of a polymer product according to claim 2, characterized in that: The test sample is obtained by traversing and selecting one group of original samples from the K groups of original samples after the original samples are divided into K groups; the training sample is obtained by traversing and selecting K-1 groups of original samples from the K groups of original samples; Correspondingly, the current optimization models are K, and determining the odor analysis model according to the current optimization model includes: Based on the K current optimized models, the hyperparameters of the model are selected or optimized, the odor analysis model is constructed, and the accuracy of the odor analysis model is evaluated.

5. The method for detecting odor of a polymer product according to claim 2, characterized in that: The artificial odor evaluation data is the description of odor perception attributes and intensity levels obtained by olfactory judges evaluating training samples and test samples according to prescribed methods.

6. The method for detecting odor of a polymer product according to claim 1, characterized in that: Based on the gas chromatogram of the gas to be tested, extracting the overall response information and / or the peak information of each gas chromatogram as the odor data of the polymer product to be tested, including: Based on the clustering algorithm, the spectral peaks of the polymer products to be tested are clustered to obtain chromatographic peaks representing the same odor substance in different polymer products to be tested; based on the chromatographic peak of each odor substance, the overall response information and / or spectral peak information of each gas chromatograph is extracted as the odor data of the polymer products to be tested.

7. The method for detecting the odor of a polymer product according to any one of claims 1 to 6, characterized in that: The detection information of the polymer product to be tested includes the odor attribute and / or odor evaluation grade of the polymer product to be tested; the odor attribute includes at least one of odor category, odor label and odor description information.

8. A polymer product odor detection device, characterized in that: including memory and processor; The memory stores a computer program, and when the computer program is executed by the processor, the steps of the method for detecting a polymer product according to any one of claims 1 to 6 are implemented.

9. An electronic nose system, characterized in that: include: A sampling device, used to collect the gas to be tested emitted by the polymer product to be tested; A chromatographic separation device, used for performing gas chromatography analysis on the gas to be tested to obtain a gas chromatogram of the gas to be tested; A detection device, used for extracting odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested, comprising: extracting overall response information and / or peak information of each gas chromatogram as the odor data of the polymer product to be tested based on the gas chromatogram of the gas to be tested; The overall response information is used to describe the information of all emitted gas substances contained in the polymer product to be tested, which is proportional to the concentration of the substance. The data can be organized into s i : s i ∈R p ,i=1,...,n Among them, s i is the response information of the gas chromatograph, n is the number of polymer products to be tested, and p is the number of data points of the gas chromatogram curve; The spectral peak information is used to describe the concentration information of each emitted gas substance of the polymer product to be tested, and the spectral peak information of each chromatographic peak is calculated, wherein the spectral peak information includes the retention time, and the peak height or peak area, and the following can be obtained: t i ∈R q ,i=1,...n Among them, t i is the peak height or peak area of ​​each chromatographic peak, n is the number of polymer products to be tested, and q is the number of different chromatographic peaks in the spectrum; The odor analysis device is used to input the odor data of the polymer product to be tested into a pre-trained odor analysis model for analysis, and output detection information of the polymer product to be tested.