Hybrid gas concentration detection method based on temperature modulation technique

Abstract

The application discloses a mixed gas concentration detection method based on temperature modulation technology and relates to the field of semiconductor gas sensing.The application solves the problem that the concentration of mixed gas is difficult to predict and is difficult to analyze more accurately.The application expands the gas-sensitive response data features in the scale of the spatial domain through the temperature modulation technology, adopts an LSTM neural network model to process time sequence data, fully mines the feature information contained in the sequence data, and significantly improves the recognition accuracy of the mixed gas concentration in combination of the two.The application is mainly used for detecting the concentration of H2 and CO gases released in the transformer oil during the fault.

Description

Technical Field

[0001] This invention relates to the field of semiconductor gas sensing. Background Technology

[0002] As a crucial component of power transmission and distribution, the reliable operation of transformers is critical to the safety and stability of the power system. Currently, dissolved gas analysis (DGA) is considered the most effective method for transformer fault diagnosis. Monitoring the decomposition gases in transformer oil through a gas detection system helps achieve online monitoring of the transformer's operating status. When a transformer fault occurs, CO and H2 gases are released sequentially from the transformer oil. These two gases can be considered two typical characteristic dissolved gases in transformer oil; therefore, real-time detection of CO and H2 gases is of great significance for the safe and stable operation of transformer equipment. Currently, gas chromatography is commonly used to separate and detect dissolved gases in transformer oil. The basic steps of dissolved gas extraction include transformer oil sampling, purification, and degassing separation of dissolved gases. Subsequently, the gas chromatograph separates the mixture into individual components, which are then sent to infrared spectroscopy or gas sensors for concentration detection. However, because CO and H2 gases move at similar speeds in the chromatographic column, their elution times after passing through the column are relatively close, leading to some overlap of signal peaks detected by the sensor. Therefore, certain errors may occur in actual detection.

[0003] In recent years, artificial olfaction systems, composed of metal oxide (MOS) gas sensor arrays and artificial intelligence algorithms, have demonstrated significant advantages in the field of real-time gas detection. MOS gas sensors are characterized by fast detection speed, simple operation, and low cost. Artificial intelligence algorithms possess strong feature extraction and recognition capabilities, spontaneously extracting feature information from data and updating model parameters through self-learning, exhibiting strong generalization ability. Currently, artificial olfaction systems are playing an important role in industrial pollution monitoring, food safety, and the medical field. However, current research on the prediction of mixed gases often focuses only on the prediction of gas types, neglecting the crucial indicator of concentration, making more accurate analysis difficult. Therefore, developing a method for predicting the concentration of mixed gases has significant application value. Summary of the Invention

[0004] The purpose of this invention is to solve the problem that it is difficult to make more accurate analysis of the concentration prediction of mixed gases. This invention provides a method for detecting the concentration of mixed gases based on temperature modulation technology.

[0005] A method for detecting the concentration of mixed gases based on temperature modulation technology, the method comprising the following steps:

[0006] Gas preparation: Preparation of mixed gases of various concentrations, wherein the mixed gases consist of H2 and CO;

[0007] Data acquisition: The gas-sensitive response data of the mixed gas collected by the sensor array under different temperature modulation modes in each working cycle is used as a training sample. The sensor array is made up of multiple metal oxide gas sensors of different types arranged in an array. The training sample is a data matrix composed of the gas-sensitive response data collected by four metal oxide gas sensors under different temperature modulation modes in each working cycle.

[0008] Data processing: Preprocessing the training samples and labeling each gas type and concentration;

[0009] Model training: Input the training samples into the LSTM neural network model to predict the concentration of each type of gas in the mixed gas, thereby training the LSTM neural network model;

[0010] Inference phase: The concentration of the mixed gas is predicted using the trained LSTM neural network model.

[0011] Preferably, each metal oxide gas sensor includes a substrate and a gas-sensitive part; the gas-sensitive part is fabricated by mixing a doped metal target and a SnO2 target.

[0012] The substrate is an Al2O3 ceramic sheet, with positive and negative gold electrodes printed on the top surface and positive and negative heating electrodes printed on the bottom surface. The sensitive part is connected to the positive and negative gold electrodes, and the positive and negative gold electrodes serve as the data output terminals of the metal oxide gas sensor, while the positive and negative heating electrodes serve as the temperature modulation terminals of the metal oxide gas sensor.

[0013] Preferably, the heating electrode is made of ruthenium oxide.

[0014] Preferably, the sensor array includes four types of metal oxide gas sensors, and the metals doped in the SnO2 target of the four types of metal oxide gas sensors are In, Ni, Pd and Co, respectively.

[0015] Preferably, the LSTM neural network model includes one input layer, two LSTM layers, one fully connected layer, and one output layer arranged sequentially.

[0016] The received gas-sensitive response data is sequentially fed into two LSTM layers through the input layer for feature extraction. The extracted features are then integrated through a fully connected layer and output through the output layer.

[0017] Preferably, the fully connected layer is composed of a combination of linear transformation and activation function.

[0018] Preferably, the output layer is implemented using a linear regression activation function.

[0019] Preferably, the preprocessing of training samples is performed by sequentially performing smoothing and normalization.

[0020] A mixed gas concentration detection device based on temperature modulation technology includes a storage device, a processor, and a computer program stored in the storage device and executable on the processor. The processor executes the computer program to implement the mixed gas concentration detection method based on temperature modulation technology.

[0021] A computer-readable storage device stores a computer program that, when executed, implements the method for detecting the concentration of a mixed gas based on temperature modulation technology.

[0022] Advantages of this invention:

[0023] This invention provides a method for detecting the concentration of mixed gases based on temperature modulation technology. Compared to existing methods for predicting mixed gas concentrations, this invention significantly improves the accuracy of mixed gas concentration identification by expanding the features of gas-sensitive response data at the spatial domain scale through temperature modulation technology. Simultaneously, it employs an LSTM neural network model to process time-series data, fully extracting the feature information contained within the sequence data. This identification method boasts advantages such as high identification accuracy and high prediction precision. Using this method, gas concentration prediction tasks under complex backgrounds can be accomplished, providing an effective method for detecting gases released from transformer oil.

[0024] The gas concentration detection method of the present invention has the advantages of being fast, efficient, and convenient, and is of great significance to the development of the field of gas identification. Attached Figure Description

[0025] Figure 1 This is a flowchart of the mixed gas concentration detection method based on temperature modulation technology described in this invention;

[0026] Figure 2 These are scanning electron microscope images of SnO2 target materials doped with four different metals; among them, Figure 2 a is an electron microscope image of In-doped SnO2. Figure 2 b is an electron microscope image of Ni-doped SnO2. Figure 2 c is an electron microscope image of Pd-doped SnO2. Figure 2 d is an electron microscope image of Co-doped SnO2; and 100nm in the image represents the scale bar.

[0027] Figure 3 This is a graph showing the temperature change of the sensor array under temperature modulation mode.

[0028] Figure 4This is a comparison chart of the predicted and actual concentrations of CO gas in a gas mixture.

[0029] Figure 5 This is a comparison chart of the predicted and actual concentrations of H2 gas in a gas mixture.

[0030] Figure 6 This is a graph showing the gas-sensing response data obtained by four metal oxide gas sensors detecting a mixture of CO and H2 under temperature modulation mode. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0033] Detailed Implementation Method 1, see [link / reference] Figure 1 This embodiment describes a method for detecting the concentration of a mixed gas based on temperature modulation technology, which includes the following steps:

[0034] Gas preparation: Preparation of mixed gases of various concentrations, wherein the mixed gases consist of H2 and CO;

[0035] Data acquisition: The gas-sensitive response data of the mixed gas collected by the sensor array under different temperature modulation modes in each working cycle is used as a training sample. The sensor array is made up of multiple metal oxide gas sensors of different types arranged in an array. The training sample is a data matrix composed of data collected by four metal oxide gas sensors under different temperature modulation modes in each working cycle.

[0036] Data processing: Preprocessing the training samples and labeling each gas type and concentration;

[0037] Model training: Input the training samples into the LSTM neural network model to predict the concentration of each type of gas in the mixed gas, thereby training the LSTM neural network model;

[0038] Inference phase: The concentration of the mixed gas is predicted using the trained LSTM neural network model.

[0039] In this embodiment, temperature modulation technology is used to expand the feature dimensions of the gas-sensitive response data. Each training sample uses one working cycle of a sensor array as the data extraction period, and each training sample includes the data output by four metal oxide gas sensors under each temperature modulation. The data collected under each temperature modulation mode within the working cycle is used as a feature dimension. At the same time, a long short-term memory neural network model (LSTM neural network model) is used to extract the features of the gas-sensitive response data, fully mining the feature information contained in the sequence data, thereby improving the prediction accuracy of the mixed gas concentration.

[0040] Figure 3 The table gives a duty cycle of 500 seconds for the sensor array under temperature modulation.

[0041] Furthermore, each metal oxide gas sensor includes a substrate and a gas-sensitive element;

[0042] The gas-sensitive part is fabricated by mixing a doped metal target and a SnO2 target.

[0043] The substrate is an Al2O3 ceramic sheet, with positive and negative gold electrodes printed on the top surface and positive and negative heating electrodes printed on the bottom surface. The sensitive part is connected to the positive and negative gold electrodes, and the positive and negative gold electrodes serve as the data output terminals of the metal oxide gas sensor, while the positive and negative heating electrodes serve as the temperature modulation terminals of the metal oxide gas sensor.

[0044] This preferred embodiment provides a specific structure for a metal oxide gas sensor, which is simple to fabricate and easy to implement.

[0045] Further heating electrodes are implemented using ruthenium oxide.

[0046] Furthermore, the sensor array includes four types of metal oxide gas sensors, and the SnO2 targets of these four sensors are doped with In, Ni, Pd, and Co, respectively. See also Figure 2 Scanning electron microscope images of four metal-doped SnO2 target materials are presented. It can be seen from the images that the morphology of the four thin film materials is basically the same, and they are uniform films. The doping elements have no significant effect on the morphology of the film.

[0047] See Figure 6 The response curves of gas-sensing response data acquired by metal oxide gas sensors doped with In, Ni, Pd, and Co in that order are presented. The gas-sensing response data represents the resistance change of the metal oxide gas sensors.

[0048] Furthermore, the LSTM neural network model includes one input layer, two LSTM layers, one fully connected layer, and one output layer arranged sequentially.

[0049] The received gas-sensitive response data is sequentially fed into two LSTM layers through the input layer for feature extraction. The extracted features are then integrated through a fully connected layer and output through the output layer.

[0050] Specifically, the fully connected layer is composed of a combination of linear transformation and activation function; the output layer is implemented using a linear regression activation function, which establishes a mapping relationship between input data and concentration to predict gas concentration.

[0051] LSTM uses special "gate" mechanisms (forget gate, input gate, output gate) to control the flow of information, thereby solving the gradient vanishing or gradient explosion problems that occur when traditional RNNs process long sequences, and can capture long-term dependencies more effectively.

[0052] Furthermore, the preprocessing of the training samples is achieved by performing smoothing and normalization sequentially. Specifically, the data is scaled to the range [0,1] using the Min-Max normalization method to eliminate errors caused by the unit of measurement.

[0053] Specific Implementation Method 2: A mixed gas concentration detection device based on temperature modulation technology, comprising a storage device, a processor, and a computer program stored in the storage device and executable on the processor, wherein the processor executes the computer program to implement the mixed gas concentration detection method based on temperature modulation technology as described above.

[0054] Specific Implementation Method 3: A computer-readable storage device, the storage device storing a computer program, which, when executed, implements the mixed gas concentration detection method based on temperature modulation technology as described above.

[0055] Verification experiment:

[0056] 1. Fabrication of the sensor array. Magnetron sputtering was employed to co-sputter multiple targets to fabricate doped metal oxide gas-sensitive materials. The substrate target was SnO2, and the doped metal targets included In, Ni, Pd, and Co. The sputtering power of the SnO2 target was 100W, and the sputtering power of the doped metal targets was 10W. The sputtering pressure was 3.4 Pa, the sputtering time was 20 min, the sputtering temperature was 100℃, the Ar / O2 ratio was 60 / 30, and the bias voltage was 50V. A sensor array with four types of metal oxide gas sensors was obtained.

[0057] 2. The prepared sensor array is used to collect gas-sensing data on the mixed gas formed by CO and H2;

[0058] 3. Perform preprocessing such as smoothing and normalization on the collected gas-sensitive response data, and assign category and concentration labels; divide the data into training set and test set;

[0059] 4. The training set is used to train the LSTM neural network model, and the output is the concentration value of each type of gas; the data in the test set is input into the trained LSTM neural network model to predict the concentration of H2 gas and CO gas.

[0060] See Figure 4 and Figure 5 Comparison charts of predicted CO and H2 gas concentrations with actual concentrations are provided. Figure 4 and Figure 5 Most of the predicted concentrations are the same as the actual concentrations, and the predicted concentrations of the parts that are different are also very close to the actual concentrations, which proves that the prediction accuracy of the present invention is high and the prediction method of the present invention is effective.

[0061] Compared with existing technologies, the advantages of this invention are as follows: The use of temperature modulation technology expands the feature dimensions of gas-sensitive data, enriches the characteristics of gas-sensitive performance, and thus improves the prediction accuracy of mixed gases. The use of an LSTM neural network model to process time-series data fully mines the feature information contained in the sequence data, further improving the accuracy of the prediction method.

[0062] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A method for detecting the concentration of a mixed gas based on temperature modulation technology, characterized in that, The method includes the following steps: Gas preparation: Preparation of mixed gases of various concentrations, wherein the mixed gases consist of H2 and CO; Data acquisition: The gas-sensitive response data of the mixed gas collected by the sensor array under different temperature modulation modes in each working cycle is used as a training sample. The sensor array is made up of multiple metal oxide gas sensors of different types arranged in an array. The training sample is a data matrix composed of the gas-sensitive response data collected by four metal oxide gas sensors under different temperature modulation modes in each working cycle. Each metal oxide gas sensor includes a substrate and a gas-sensing part; the gas-sensing part is fabricated by mixing a doped metal target and a SnO2 target. The substrate is an Al2O3 ceramic sheet, with positive and negative gold electrodes printed on the top surface and positive and negative heating electrodes printed on the bottom surface. The sensitive part is connected to the positive and negative gold electrodes, and the positive and negative gold electrodes serve as the data output terminals of the metal oxide gas sensor, while the positive and negative heating electrodes serve as the temperature modulation terminals of the metal oxide gas sensor. The sensor array includes four metal oxide gas sensors, and the SnO2 targets of the four metal oxide gas sensors are doped with metals In, Ni, Pd and Co, respectively. Data processing: Preprocessing the training samples and labeling each gas type and concentration; Model training: Input the training samples into the LSTM neural network model to predict the concentration of each type of gas in the mixed gas, thereby training the LSTM neural network model; Inference phase: The concentration of the mixed gas is predicted using the trained LSTM neural network model.

2. The method for detecting the concentration of a mixed gas based on temperature modulation technology according to claim 1, characterized in that, The heating electrode is made of ruthenium oxide.

3. The method for detecting the concentration of a mixed gas based on temperature modulation technology according to claim 1, characterized in that, The LSTM neural network model consists of one input layer, two LSTM layers, one fully connected layer, and one output layer arranged sequentially. The received gas-sensitive response data is sequentially fed into two LSTM layers through the input layer for feature extraction. The extracted features are then integrated through a fully connected layer and output through the output layer.

4. The method for detecting the concentration of a mixed gas based on temperature modulation technology according to claim 3, characterized in that, Fully connected layers are composed of a combination of linear transformations and activation functions.

5. The method for detecting the concentration of a mixed gas based on temperature modulation technology according to claim 3, characterized in that, The output layer is implemented using a linear regression activation function.

6. The method for detecting the concentration of a mixed gas based on temperature modulation technology according to claim 1, characterized in that, The preprocessing of training samples is achieved by performing smoothing and normalization processes sequentially.

7. A mixed gas concentration detection device based on temperature modulation technology, comprising a storage device, a processor, and a computer program stored in the storage device and executable on the processor, characterized in that, The processor executes a computer program to implement the mixed gas concentration detection method based on temperature modulation technology as described in any one of claims 1 to 6.

8. A computer-readable storage device storing a computer program, characterized in that, When the computer program is executed, it implements the mixed gas concentration detection method based on temperature modulation technology as described in any one of claims 1 to 6.

Images