CN115563546A - Intelligent gas smell identification method, system, medium, equipment and terminal - Google Patents
Intelligent gas smell identification method, system, medium, equipment and terminal Download PDFInfo
- Publication number
- CN115563546A CN115563546A CN202210989922.4A CN202210989922A CN115563546A CN 115563546 A CN115563546 A CN 115563546A CN 202210989922 A CN202210989922 A CN 202210989922A CN 115563546 A CN115563546 A CN 115563546A
- Authority
- CN
- China
- Prior art keywords
- gas
- data set
- characteristic
- curve
- intelligent
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Investigating Or Analyzing Materials By The Use Of Fluid Adsorption Or Reactions (AREA)
Abstract
The invention belongs to the technical field of gas identification, and discloses a gas smell intelligent identification method, a system, a medium, equipment and a terminal, which are used for acquiring response recovery curve data of a gas sensor to gas identification; extracting characteristic values from the response recovery curve; preprocessing the characteristic value and standardizing the characteristic value by Z-score to obtain a characteristic data set; randomly dividing the characteristic data set into a training data set and a testing data set; and training the training data set by using a machine learning algorithm to obtain a comprehensive gas classification model, and detecting the comprehensive gas classification model by using the test data set. The invention trains gas characteristic data by utilizing five machine learning algorithms, predicts and classifies the gas data to be detected by integrating the gas classification model, can classify the gas by a machine learning mode, reduces the participation of artificial subjectivity, and accurately identifies the category of the gas in a short time, and the prediction accuracy of the artificial neural network is 95.5%.
Description
Technical Field
The invention belongs to the technical field of gas identification, and particularly relates to a gas smell intelligent identification method, system, medium, equipment and terminal.
Background
At present, volatile organic compound gas not only comes from waste gas discharged in the industrial production process, but also exists in gas released by using chemical materials in the house decoration process, and poses potential threats to the health and the life quality of people. In addition, the volatile organic compound gas can be used as a human disease marker, for example, the types and the contents of benzene series and acetone gas in the exhaled gas of a human body can be detected, so that whether a patient has the risk of lung cancer and diabetes can be preliminarily screened, and therefore, the volatile organic compound gas needs to be accurately identified.
Accurate and reliable monitoring of volatile organic compound gases in pollution sources and ambient air is a prerequisite guarantee for research and control of pollution. The industrial emission of volatile organic compound gas has the characteristics of point position dispersion, wide distribution range, large emission amount and the like, and has the problems of difficult discovery, monitoring and judgment. In the existing stage, qualitative detection and identification of industrial source volatile organic compound gas mostly depend on off-line analysis of large instruments such as a mass spectrometer, a chromatograph and the like, the on-line measurement cost is high, and the sampling analysis period is long.
In recent years, the gas sensor technology has been developed rapidly, and the sensor has the remarkable advantages of small size, low cost and the like, and is one of the important development directions for solving the problem. However, the detection of the sensor has the problems of cross interference, poor stability, baseline drift and the like, the extraction of characteristic values usually needs manual screening, and the online qualitative and quantitative analysis of gas components is difficult to realize. Therefore, an efficient method for intelligently identifying and rapidly predicting the smell of gas is needed to make up for the defects of the prior art.
Through the above analysis, the problems and defects of the prior art are as follows:
(1) In the present stage, qualitative detection and identification of industrial source volatile organic compound gas mostly depend on off-line analysis of large instruments such as a mass spectrometer, a chromatograph and the like, the on-line measurement cost is high, and the sampling analysis period is long.
(2) The problems of cross interference, poor stability, baseline drift and the like exist in the detection of the sensor, the extraction of the characteristic value is usually manually screened, and the online qualitative and quantitative analysis of the gas components is difficult to realize.
Disclosure of Invention
The invention provides a gas smell intelligent identification method, a system, a medium, equipment and a terminal, and particularly relates to a gas smell intelligent identification method, a system, a medium, equipment and a terminal based on machine learning.
The invention is realized in such a way that a gas smell intelligent identification method comprises the following steps:
and training the characteristic data set acquired by the gas sensor by using a machine learning algorithm to obtain a comprehensive classification model, and further obtaining a gas type prediction result by using the comprehensive classification model.
Further, the intelligent gas smell identification method comprises the following steps:
acquiring response recovery curve data of a gas sensor to gas identification;
step two, extracting characteristic values from the response recovery curve;
thirdly, preprocessing the characteristic value and standardizing the characteristic value by Z-score to obtain a characteristic data set;
step four, randomly dividing the characteristic data set into a training data set and a testing data set;
and fifthly, training the training data set by using a machine learning algorithm to obtain a comprehensive gas classification model, and detecting the comprehensive gas classification model by using the test data set.
Further, the characteristic values in the second step include the sum of response time, recovery time, steady-state time and curve integral area, the maximum response value, the maximum positive slope in the rising stage and the minimum negative slope in the falling stage.
Further, the response time is the time required for the output change to reach a stable value of 90% when the concentration of the gas contacted with the sensor changes in a step mode; the recovery time is the time required for the ventilation equilibrium state of the sensor to recover to 10% of the signal; the steady state time is the time at which the sensor response curve remains in equilibrium.
The sum of the integral areas of the curves is calculated by a Newton-Cotes product solving formula;
wherein h is an integral step length which is the distance between the horizontal coordinates of two adjacent points on the curve; n represents that the integral area is averagely divided into i +1 equal parts; f (x) i ) And f (x) i + 1) are the ith and (i + 1) th function values on the curve, respectively.
The maximum response value is the maximum value of the curve.
The maximum positive slope of the rise phase is calculated according to the following formula:
wherein, x is taken as a rising stage area.
The minimum negative slope calculation formula in the descending stage is as follows:
wherein, x is taken as a descending stage area.
Further, the preprocessing and Z-score normalization of the feature values extracted from the response recovery curve in the third step includes:
removing useless data to obtain a characteristic value according to preset useless characteristics;
the eigenvalues are de-dimensioned using Z-score normalization.
Further, the machine learning algorithm in the fifth step includes five machine learning algorithms including logistic regression, K nearest neighbor, random forest, support vector machine and artificial neural network.
Another object of the present invention is to provide a gas smell intelligent recognition system using the gas smell intelligent recognition method, the gas smell intelligent recognition system comprising:
the curve data acquisition module is used for acquiring response recovery curve data of the gas sensor to gas identification;
the characteristic value extraction module is used for extracting a characteristic value from the response recovery curve;
the characteristic data set acquisition module is used for preprocessing the characteristic values and standardizing Z-score to obtain a characteristic data set, and randomly dividing the characteristic data set into a training data set and a testing data set;
and the model construction module is used for training the training data set by using a machine learning algorithm to obtain a comprehensive gas classification model and detecting the comprehensive gas classification model by using the test data set.
Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and the computer program, when executed by the processor, causes the processor to execute the steps of the intelligent gas smell identification method.
Another object of the present invention is to provide a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the processor executes the steps of the intelligent identification method for smell sense of gas.
The invention also aims to provide an information data processing terminal which is used for realizing the intelligent gas smell identification system.
In combination with the technical solutions and the technical problems to be solved, please analyze the advantages and positive effects of the technical solutions to be protected in the present invention from the following aspects:
first, aiming at the technical problems and difficulties in solving the problems in the prior art, the technical problems to be solved by the technical scheme of the present invention are closely combined with results, data and the like in the research and development process, and some creative technical effects are brought after the problems are solved. The specific description is as follows:
the invention provides a gas smell intelligent recognition method based on machine learning, which trains a characteristic data set acquired by a gas sensor by using a machine learning algorithm to obtain a comprehensive classification model and further obtain a gas type prediction result. The comprehensive gas classification model based on the five algorithms uses five algorithms of logistic regression, nearest K, random forest, support vector machine and artificial neural network; accordingly, the integrated gas classification model can use five algorithm models to detect the gas to be detected, and an optimal result is obtained. Simulation experiments show that the prediction accuracy of the five algorithms provided by the invention reaches over 90 percent, wherein the prediction effect of the artificial neural network is the best, and the accuracy is 95.5 percent.
Secondly, considering the technical scheme as a whole or from the perspective of products, the technical effect and advantages of the technical scheme to be protected by the invention are specifically described as follows:
the gas classification method based on the comprehensive gas classification model utilizes five machine learning algorithms to train the gas characteristic data, carries out prediction classification on the gas data to be detected through the comprehensive gas classification model, can classify the gas in a machine learning mode, reduces human subjective participation, and can accurately identify the category of the gas in a short time.
Third, as an inventive supplementary proof of the claims of the present invention, there are also presented several important aspects:
(1) The technical scheme of the invention fills the technical blank in the industry at home and abroad:
regarding gas identification, data are often analyzed by using a certain machine learning algorithm, the accuracy of gas identification is improved by improving the form of the algorithm, and whether the algorithm is suitable for the situation is often ignored.
(2) The technical scheme of the invention solves the technical problems which are always desired to be solved but are not successfully achieved:
gas identification tends to suffer from several problems: first, the characteristic parameters extracted from the gas response recovery curve tend to have different dimensions and dimensional units; second, the collection of gas sensor data is very difficult, so the existing sensor data is often less; thirdly, the feature extraction usually needs manual extraction, which is time-consuming and not accurate enough. The present invention effectively solves these problems.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required to be used in the embodiments of the present invention will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of an intelligent gas smell identification method provided by an embodiment of the invention;
FIG. 2 is a graph illustrating a recovery of a response of a gas sensor to gas identification provided by an embodiment of the present invention;
FIG. 3 is a recognition confusion matrix diagram of a trained integrated gas recognition model according to an embodiment of the present invention; (a) is KNN, (b) is Logistic Reg, (c) is Random Forest, (d) is ANN, and (e) is SVM.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
In order to solve the problems in the prior art, the invention provides a method, a system, a medium, a device and a terminal for intelligently identifying smell of gas, and the invention is described in detail with reference to the accompanying drawings.
1. The embodiments are explained. This section is an explanatory embodiment expanding on the claims so as to fully understand how the present invention is embodied by those skilled in the art.
As shown in fig. 1, the intelligent identification method for smell of gas provided by the embodiment of the invention comprises the following steps:
s101, acquiring response recovery curve data of a gas sensor to gas identification;
s102, extracting a characteristic value from the response recovery curve;
s103, preprocessing the characteristic value and standardizing the Z-score to obtain a characteristic data set;
s104, randomly dividing the characteristic data set into a training data set and a testing data set;
and S105, training the training data set by using a machine learning algorithm to obtain a comprehensive gas classification model, and detecting the comprehensive gas classification model by using the test data set.
The characteristic values provided by the embodiment of the invention comprise response time, recovery time, steady-state time, sum of curve integral areas, maximum response value, maximum positive slope in an ascending stage and minimum negative slope in a descending stage.
The response time is the time required when the output change reaches a stable value of 90% when the concentration of gas contacted with the sensor changes in a step mode; the recovery time is the time required for the ventilation equilibrium state of the sensor to recover to 10% of the signal; the steady state time is the time at which the sensor response curve remains in equilibrium.
The sum of the integral areas of the curves is calculated by a Newton-Cotes product solving formula;
wherein h is an integral step length and is the distance between the abscissas of two adjacent points on the curve(ii) a n represents the average division of the integration area into i +1 equal parts; f (x) i ) And f (x) i + 1) are the ith and (i + 1) th function values on the curve, respectively.
The maximum response value is the maximum of the curve.
The maximum positive slope during the rise phase is calculated as follows:
wherein, x is taken as a rising stage area.
The minimum negative slope calculation formula in the descending stage is as follows:
wherein, x is taken as a descending stage area.
The preprocessing and Z-score standardization of the characteristic value provided by the embodiment of the invention comprises the following steps:
removing useless data to obtain a characteristic value according to preset useless characteristics;
the eigenvalues are de-dimensioned using Z-score normalization.
The comprehensive gas classification model based on the five algorithms provided by the embodiment of the invention uses five algorithms of logistic regression, nearest K, random forest, support vector machine and artificial neural network; accordingly, the comprehensive gas classification model can use five algorithm models to detect the gas to be detected and obtain the optimal result.
The gas smell intelligent identification system provided by the embodiment of the invention comprises:
the curve data acquisition module is used for acquiring response recovery curve data of the gas sensor to gas identification;
the characteristic value extraction module is used for extracting a characteristic value from the response recovery curve;
the characteristic data set acquisition module is used for preprocessing the characteristic values and standardizing Z-score to obtain a characteristic data set, and randomly dividing the characteristic data set into a training data set and a testing data set;
and the model construction module is used for training the training data set by using a machine learning algorithm to obtain a comprehensive gas classification model and detecting the comprehensive gas classification model by using the test data set.
2. Application examples. In order to prove the creativity and the technical value of the technical scheme of the invention, the part is an application example of the technical scheme of the claims to a specific product or related technology.
As shown in fig. 1, the intelligent gas smell identification method provided by the embodiment of the present invention includes the following steps:
s101, acquiring response recovery curve data of five types of volatile organic compound gases of a gas sensor, such as p-toluene, p-xylene, m-xylene, o-xylene, ethylbenzene and the like, under the same concentration;
s102, extracting the sum of response time, recovery time, steady-state time and curve integral area from a response recovery curve, and taking the maximum response value, the maximum positive slope of an ascending stage and the minimum negative slope of a descending stage as characteristic values;
s103, preprocessing the characteristic value and standardizing the Z-score to obtain a characteristic data set;
s104, randomly dividing the characteristic data set into a training data set and a testing data set;
and S105, training the training data set by using five machine learning algorithms of logistic regression, nearest K, random forest, support vector machine and artificial neural network respectively to obtain a comprehensive gas classification model based on the five algorithms.
3. Evidence of the relevant effects of the examples. The embodiment of the invention achieves some positive effects in the process of research and development or use, and has great advantages compared with the prior art, and the following contents are described by combining data, diagrams and the like in the test process.
The invention provides a gas smell intelligent identification method based on machine learning, which can classify gases in a machine learning mode and reduce artificial subjective participation, so that the category of the gases can be accurately identified in a short time.
In the present invention, the gas is exemplified by a volatile organic compound gas, wherein the volatile organic compound gas includes toluene, p-xylene, m-xylene, o-xylene, and ethylbenzene, and the identification and detection are performed based on the five types of gas components as the main body of the volatile organic compound gas.
As shown in fig. 1, an embodiment of the present invention provides a gas smell intelligent identification method based on machine learning, which specifically includes the following steps:
s1, collecting gas sensing response recovery curve data of volatile organic compound gas;
specifically, the same gas sensor is placed in an organic compound gas environment with the same gas concentration but different gas types, and multiple groups of gas sensor response recovery curve data are collected under the condition of the same prefabricated parameters, wherein the organic compound gas is selected from toluene, p-xylene, m-xylene, o-xylene and ethylbenzene.
S2, extracting the sum of response time, recovery time, steady-state time and curve integral area from a response recovery curve, and taking the maximum response value, the maximum positive slope of a rising stage and the minimum negative slope of a falling stage as characteristic values;
specifically, the obtained response curve is subjected to feature extraction as shown in fig. 3.
More specifically, the response time is the time required for the output change to reach a stable value of 90% when the concentration of the gas contacted with the sensor changes in a step manner, the recovery time is the time required for the ventilation equilibrium state of the sensor to recover to a 10% signal, and the steady-state time is the time for the response curve of the sensor to keep the equilibrium state.
Further, the sum of the curve integral areas is calculated by a Newton-Cotes product finding formula, and the specific calculation is as follows:
wherein h is an integral step length, namely the distance between the abscissas of two adjacent points on the curve; n represents the average division of the integration area into i +1 equal parts; f (x) i ) And f (x) i + 1) are the i-th and i + 1-th function values on the curve, respectively.
Further, the maximum response value is the maximum value of the curve.
Further, the maximum positive slope during the ramp-up phase is calculated as follows:
wherein, x is taken as a rising stage area.
The minimum negative slope calculation formula in the descending stage is as follows:
wherein, x is taken as a descending stage area.
S3, preprocessing the characteristic value and standardizing the Z-score according to the data in the S2 to obtain a characteristic data set;
specifically, the steps include:
s1, removing useless data to obtain a characteristic value according to preset useless characteristics;
s2, carrying out dimensionless processing on the characteristic value by utilizing Z-score standardization, wherein a calculation formula is as follows:
where μ is the mean of a feature in all samples and σ is the standard deviation of a feature under all samples.
S4, randomly performing the following steps of 4:1 into a training data set and a testing data set;
s5, training a training data set by using five machine learning algorithms of logistic regression, nearest K, random forest, support vector machine and artificial neural network respectively to obtain a comprehensive gas classification model based on the five algorithms;
specifically, the logistic regression algorithm is a two-classification algorithm, and there are two main methods for improving the algorithm to realize the multi-classification task: the first is to regard the classification problem of each category as a two-classification problem, namely, the classifier only needs to judge whether the sample belongs to the category or not, so the number of the classifiers is needed according to the number of the categories; secondly, the output form of Softmax is adopted, and the invention adopts the output form of Softmax.
Specifically, the K-nearest neighbor algorithm is an unparameterized algorithm and can be used for both classification and regression, and the core idea of the K-nearest neighbor algorithm is to find K samples that are nearest to the test sample in the feature space.
The algorithm flow is as follows:
inputting: training set samples and test samples
And (3) outputting: test sample classification results
1: calculating the distance between the test data and each training data
2: sorting according to increasing distance relation
3: selecting K points with minimum distance
4: determining the frequency of occurrence of the category in which the K points are located (classification), determining the weighting value of the K points (regression)
5: and returning the category with the highest frequency in the K points as the category (classification) of the test data, and returning the weighted value of the K points as the predicted value of the test data.
In particular, support Vector Machines (SVMs) are binary algorithms. Given a set of training sample sets, if the sample data sets are two-dimensional and are scattered on a plane, a straight line needs to be found to segment the data sets. There are many lines that can be separated, and the line with the best generalization capability and the strongest robustness is to be found. This is a point on a plane, if in three-dimensional space, one plane needs to be found; if the dimension exceeds three dimensions, a hyperplane needs to be found.
In particular, a random forest is a classifier that contains multiple decision trees, and the class of its output is dependent on the mode of the class of the individual tree output. The algorithm flow is as follows:
1. for a sample with the sample capacity of N, the replacement extraction is carried out for N times, and 1 sample is extracted each time, so that N samples are finally formed. And training by using the N samples to obtain a decision tree, wherein the decision tree is used as the sample at the root node of the decision tree.
2. When each sample has M attributes, when each node of the decision tree needs to be split, M attributes are randomly selected from the M attributes, and the condition M < < M is met. Then, a certain policy (for example, information gain) is applied to select 1 attribute from the m attributes as the splitting attribute of the node.
3. And repeating the step 2, and continuing the splitting attribute until the splitting can not be continued any more, namely, the leaf node is reached. Note that pruning is not performed throughout the decision tree formation process.
4. And (4) establishing a large number of decision trees according to the steps 1-3 to form a random forest.
In particular, artificial Neural Networks (ANN) are a supervised machine learning technique for classifying or regressing data. An ANN, which is connected to a large number of neurons that process information and produce accurate results, typically runs on layers that are connected to nodes, each node being associated with an active function, the ANN comprising three layers, respectively referred to as an input layer, one or more hidden layers, and an output layer. ANN very easily classifies or regresses complex and non-linear data sets and, like other methods, does not have any restrictions on the inputs, but has higher computational requirements.
FIG. 2 is a recognition confusion matrix diagram of five algorithms corresponding to the trained integrated gas classification model.
And S6, sending the test data set into a trained comprehensive gas classification model to realize the rapid identification of the volatile organic compound gas.
Specifically, the prediction accuracy of the five algorithms reaches over 90%, wherein the prediction effect of the artificial neural network is the best, and the accuracy is 95.5%.
It should be noted that the embodiments of the present invention can be realized by hardware, software, or a combination of software and hardware. The hardware portions may be implemented using dedicated logic; the software portions may be stored in a memory and executed by a suitable instruction execution system, such as a microprocessor or specially designed hardware. Those skilled in the art will appreciate that the apparatus and methods described above may be implemented using computer executable instructions and/or embodied in processor control code, such code being provided on a carrier medium such as a disk, CD-or DVD-ROM, programmable memory such as read only memory (firmware), or a data carrier such as an optical or electronic signal carrier, for example. The apparatus of the present invention and its modules may be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, or software executed by various types of processors, or a combination of hardware circuits and software, e.g., firmware.
The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. An intelligent gas smell identification method is characterized by comprising the following steps:
and training the characteristic data set acquired by the gas sensor by using a machine learning algorithm to obtain a comprehensive classification model, and further obtaining a gas type prediction result by using the comprehensive classification model.
2. The intelligent gas smell identification method as claimed in claim 1, wherein the intelligent gas smell identification method comprises the following steps:
acquiring response recovery curve data of a gas sensor to gas identification;
step two, extracting characteristic values from the response recovery curve;
thirdly, preprocessing the characteristic value and standardizing the characteristic value by Z-score to obtain a characteristic data set;
step four, randomly dividing the characteristic data set into a training data set and a testing data set;
and fifthly, training the training data set by using a machine learning algorithm to obtain a comprehensive gas classification model, and detecting the comprehensive gas classification model by using the test data set.
3. The intelligent gas olfactory identification method as set forth in claim 2 wherein the characteristic values in step two include the sum of response time, recovery time, steady state time, curve integral area, maximum response value, maximum positive slope in rising phase and minimum negative slope in falling phase.
4. The intelligent gas olfactory identification method as claimed in claim 3, wherein the response time is the time required for the output to change to 90% of a stable value when the concentration of the gas contacted with the sensor changes in a step manner; the recovery time is the time required for the ventilation equilibrium state of the sensor to recover to 10% of the signal; the steady state time is the time for the sensor response curve to keep in an equilibrium state;
the sum of the integral areas of the curves is calculated by a Newton-Cotes product solving formula;
wherein h is an integral step length which is the distance between the horizontal coordinates of two adjacent points on the curve; n represents the average division of the integration area into i +1 equal parts; f (x) i ) And f (x) i + 1) are the ith and (i + 1) th function values on the curve, respectively;
the maximum response value is the maximum value of the curve;
the maximum positive slope of the rise phase is calculated according to the following formula:
wherein, x is taken as a rising stage area;
the minimum negative slope calculation formula in the descending stage is as follows:
wherein, x is taken as a descending stage area.
5. The intelligent gas olfaction identification method according to claim 2, wherein the preprocessing and Z-score normalization of the feature values extracted from the response recovery curve in the third step includes:
removing useless data to obtain a characteristic value according to preset useless characteristics;
the eigenvalues are de-dimensioned using Z-score normalization.
6. The intelligent gas olfaction identification method according to claim 2, wherein the machine learning algorithm in the fifth step includes five machine learning algorithms including logistic regression, K nearest neighbor, random forest, support vector machine and artificial neural network.
7. A gas smell intelligent recognition system using the gas smell intelligent recognition method according to any one of claims 1 to 6, the gas smell intelligent recognition system comprising:
the curve data acquisition module is used for acquiring response recovery curve data of the gas sensor to gas identification;
the characteristic value extraction module is used for extracting a characteristic value from the response recovery curve;
the characteristic data set acquisition module is used for preprocessing the characteristic values and standardizing Z-score to obtain a characteristic data set, and randomly dividing the characteristic data set into a training data set and a testing data set;
and the model construction module is used for training the training data set by utilizing a machine learning algorithm to obtain a comprehensive gas classification model and detecting the comprehensive gas classification model by utilizing the test data set.
8. A computer device, characterized in that the computer device comprises a memory and a processor, the memory stores a computer program, and the computer program is executed by the processor, so that the processor executes the steps of the gas olfaction intelligent identification method according to any one of claims 1 to 6.
9. A computer-readable storage medium, storing a computer program which, when executed by a processor, causes the processor to carry out the steps of the intelligent gas olfactory sensation identification method according to any one of claims 1 to 6.
10. An information data processing terminal, characterized in that the information data processing terminal is used for realizing the gas olfaction intelligent recognition system as claimed in claim 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210989922.4A CN115563546A (en) | 2022-08-17 | 2022-08-17 | Intelligent gas smell identification method, system, medium, equipment and terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210989922.4A CN115563546A (en) | 2022-08-17 | 2022-08-17 | Intelligent gas smell identification method, system, medium, equipment and terminal |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115563546A true CN115563546A (en) | 2023-01-03 |
Family
ID=84739391
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210989922.4A Pending CN115563546A (en) | 2022-08-17 | 2022-08-17 | Intelligent gas smell identification method, system, medium, equipment and terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115563546A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114496116A (en) * | 2020-10-26 | 2022-05-13 | 国际商业机器公司 | Volatile organic compound detection and classification |
| CN116718648A (en) * | 2023-08-11 | 2023-09-08 | 合肥中科国探智能科技有限公司 | Method for detecting and identifying thermal runaway gas of battery and alarm device thereof |
-
2022
- 2022-08-17 CN CN202210989922.4A patent/CN115563546A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114496116A (en) * | 2020-10-26 | 2022-05-13 | 国际商业机器公司 | Volatile organic compound detection and classification |
| CN116718648A (en) * | 2023-08-11 | 2023-09-08 | 合肥中科国探智能科技有限公司 | Method for detecting and identifying thermal runaway gas of battery and alarm device thereof |
| CN116718648B (en) * | 2023-08-11 | 2023-11-10 | 合肥中科国探智能科技有限公司 | Method for detecting and identifying thermal runaway gas of battery and alarm device thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115563546A (en) | Intelligent gas smell identification method, system, medium, equipment and terminal | |
| CN112580749B (en) | Intelligent fire detection method based on machine olfaction technology | |
| Gao et al. | Research and improvement of isolation forest in detection of local anomaly points | |
| CN114707598A (en) | Mixed gas identification method, system, terminal equipment and readable storage medium | |
| CN113903409A (en) | Molecular data processing method, model construction and prediction method and related device | |
| Sharma et al. | Forecasting and prediction of air pollutants concentrates using machine learning techniques: the case of India | |
| Sylligardos et al. | Choose wisely: An extensive evaluation of model selection for anomaly detection in time series | |
| CN114970977A (en) | Abnormal data detection method and system for digital urban air quality monitoring data | |
| CN110956331A (en) | Method, system and device for predicting operation state of digital factory | |
| CN113378935B (en) | Intelligent olfactory sensation identification method for gas | |
| Singh et al. | Multiclass imbalanced big data classification utilizing spark cluster | |
| Shaloo et al. | Real-Time Color Detection for Automated Production Lines Using CNN-Based Machine Learning | |
| Mohammad et al. | Air Pollution Comparison RFM Model Using Machine Learning Approach | |
| Licen et al. | Self-organizing map algorithm as a tool for analysis, visualization and interpretation of electronic nose high dimensional raw data | |
| Sugetha et al. | Performance evaluation of classifiers for analysis of road accidents | |
| Rizwan et al. | Comparative Investigation of Machine Learning and Deep Learning Methods for Univariate AQI Forecasting | |
| Lima et al. | A comparative study on concept drift detectors for regression | |
| Ahmed et al. | A review of machine learning models in the air quality research | |
| Habib et al. | A classification model for gas drift problem | |
| Asha et al. | Implementation of speculate modules and performance evaluation of data mining clustering techniques on air quality index and health index to predict high-risk air polluted stations of a metropolitan city using R programming | |
| Kiersztyn et al. | Random clustering-based outlier detector | |
| Lazzerini et al. | Fros: a fuzzy logic-based recogniser of olfactory signals | |
| Jaleel et al. | Classification of Gas Sensor Data Using Multiclass SVM | |
| Jaleel et al. | A Voting ensemble technique for gas classification | |
| Luque et al. | Reduction of the size of datasets by using evolutionary feature selection: the case of noise in a modern city |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |