A gas detection system and a gas detection method thereof

Abstract

The present invention discloses a gas detection system, which relates to the field of sensor technology and is used to solve the technical problems of weak signal, single gas type and low measurement accuracy in the prior art when detecting low volume fraction gas. The gas detection system comprises: a gas sensor and a data processor; the output end of the gas sensor is connected to the input end of the data processor; the gas sensor collects three gases respectively through three groups of circuits including gas-sensitive field effect tubes; the data processor calculates the concentrations of the three gases according to the output signals of the gas sensor. The present invention can overcome the problem of weak signal when detecting low volume fraction gas in traditional semiconductor gas sensors.

Description

Technical Field

[0001] The invention relates to the field of sensor technology, and in particular to a gas detector for detecting three-component toxic and harmful gases and a detection method thereof. Background Art

[0002] NO2, SO2 and CO are the main harmful gases in air pollution, which pose a serious threat to human health, environment and industrial safety.

[0003] A high concentration of NO2 in the environment can cause symptoms such as coughing and dyspnea. In severe cases, it can lead to pulmonary edema and respiratory damage, and may also cause chronic respiratory inflammation, neurasthenia syndrome, and tooth erosion. NO2 and SO2 are also one of the main factors in the formation of acid rain, which can destroy the ecological balance of water bodies, accelerate soil acidification, affect crop growth, and cause corrosion damage to buildings and cultural relics. CO, as a common industrial gas, may also cause poisoning accidents to workers if ventilation is poor or protective measures are not in place. Therefore, the demand for accurate detection of these three gases is becoming stronger and stronger. Real-time detection and alarm are of great practical significance to the safe development of the industrial field and environmental governance.

[0004] At present, the commonly used types of gas sensors are: catalytic combustion gas sensors, electrochemical gas sensors, optical gas sensors, and semiconductor gas sensors. Catalytic combustion gas sensors can only detect combustible gases, but not non-combustible gases; electrochemical gas sensors are prone to aging, have a short service life, are greatly affected by environmental factors, and have a low cost-performance ratio; optical gas sensors are susceptible to fluctuations in light source intensity, damage to optical fibers, and interference from the external environment; semiconductor gas sensors have the advantages of high sensitivity, long service life, fast response time, simple manufacturing methods, and low cost. However, traditional semiconductor gas sensors have the problems of weak signals when detecting low volume fraction gases and a single type of detected gas. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a gas detection system and a detection method thereof to solve the technical problems in the prior art of weak signal, single detection gas type and low measurement accuracy when the gas has a low volume fraction.

[0006] The present invention provides a gas detection system, comprising: a gas sensor and a data processor; the output end of the gas sensor is connected to the input end of the data processor;

[0007] The gas sensor collects three gases respectively through three groups of circuits including gas-sensitive field effect tubes; the data processor calculates the concentrations of the three gases according to the output signals of the gas sensors.

[0008] Furthermore, the circuit in the gas sensor includes: six acquisition units, the six acquisition units form three groups of circuits including gas-sensitive field-effect tubes; the circuit of the acquisition unit specifically includes: four gas-sensitive field-effect tubes, four non-gas-sensitive field-effect tubes and two triodes, the base of the first triode is respectively connected to the drain of the first gas-sensitive field-effect tube and the source of the first non-gas-sensitive field-effect tube; the collector of the first triode is connected to the source of the second gas-sensitive field-effect tube and serves as an output end of the acquisition unit; the emitter of the first triode is connected to the drain of the second non-gas-sensitive field-effect tube; the base of the second triode is respectively connected to the drain of the third non-gas-sensitive field-effect tube and the source of the third gas-sensitive field-effect tube The collector of the second triode is connected to the source of the fourth non-gas-sensitive field effect tube and serves as another output end of the acquisition unit; the emitter of the second triode is connected to the drain of the fourth gas-sensitive field effect tube; the drain of the first non-gas-sensitive field effect tube, the drain of the second gas-sensitive field effect tube, the drain of the fourth non-gas-sensitive field effect tube and the drain of the third gas-sensitive field effect tube are connected to an external power supply; the source of the first gas-sensitive field effect tube, the source of the second non-gas-sensitive field effect tube, the source of the fourth gas-sensitive field effect tube and the source of the third non-gas-sensitive field effect tube are grounded; the gates of the four gas-sensitive field effect tubes and the gates of the four non-gas-sensitive field effect tubes are all connected to an external power supply;

[0009] The six collection units are divided into three groups to collect the concentrations of three gases. Two collection units form a group and use the same gas-sensitive field effect tube to collect the corresponding gas concentrations.

[0010] Furthermore, the six acquisition units are arranged in an array on the PCB circuit board, and the acquisition units in the same group are arranged relative to each other.

[0011] Furthermore, the gas sensor also includes a temperature and humidity sensor, an output end of the temperature and humidity sensor is connected to an input end of a data processor, and the data processor optimizes the calculated gas concentration according to temperature data collected by the temperature and humidity sensor.

[0012] The present invention also provides a detection method of a gas detection system, which is applicable to the above-mentioned gas detection system. The detection method of the gas detection system comprises the following steps:

[0013] Step 1: Obtain the preliminary gas calibration formula of the gas sensor for each gas;

[0014] Step 2: Obtain the ambient temperature and humidity, and optimize the temperature compensation of the preliminary gas calibration formula through a neural network based on the ambient temperature and humidity;

[0015] Step 3: Optimize the gas cross sensitivity of the gas calibration formula optimized in step 2 by using the least squares support vector machine to obtain the final gas calibration formula;

[0016] Step 4: Use the gas sensor to collect the voltage data corresponding to each gas in the environment in real time, and calculate the concentration of each gas based on the voltage data using the final gas calibration formula.

[0017] Furthermore, in the detection method of the gas detection system, the gas sensor performs average processing on the output voltage of the collection unit that collects the same gas, and uses the voltage after the average processing as the voltage data corresponding to the gas concentration.

[0018] Furthermore, the neural network in step 2 is: PSO-BP neural network.

[0019] Furthermore, the step 3 also includes: optimizing the penalty parameter and kernel parameter of the least squares support vector machine by using an adaptive particle swarm algorithm.

[0020] Beneficial effects of the present invention:

[0021] The acquisition unit of the present invention adopts a differential amplifier circuit, which can effectively suppress zero-point drift, reduce the error caused by changes in ambient temperature and humidity on the experimental results, reduce the interference of environmental factors, and improve the stability and accuracy of the measurement results. At the same time, the differential amplifier circuit can also improve the strength and clarity of the signal, and can amplify the change of the output signal caused by the change of resistance in the circuit, thereby overcoming the problem of weak signal when detecting low volume fraction gas in traditional semiconductor gas sensors. At the same time, since the differential amplifier circuit can amplify the signal, there is no need to design an additional signal amplifier circuit, which simplifies the circuit structure and reduces the production cost.

[0022] The present invention arranges six acquisition units in an array, with two acquisition units for each gas arranged relatively to each other. The measurement results of the two differential circuits corresponding to the three gases are averaged respectively, and curve fitting is performed on the voltage average and the standard gas concentration. The fitting curve is used as the fitting curve of the gas, which can reduce the uncertainty caused by the error of a single acquisition unit and further improve the detection accuracy of the gas.

[0023] The present invention adopts the PSO-BP neural network algorithm which is a combination of the BP neural network algorithm and the particle swarm algorithm to perform temperature compensation, which overcomes the problems that the BP neural network algorithm is prone to fall into the extreme point when solving the global extreme value of a relatively complex nonlinear function, resulting in low measurement accuracy, slow convergence speed, low learning efficiency, etc., further reduces the interference of temperature on measurement accuracy, and further improves gas detection accuracy.

[0024] The present invention adopts a least squares support vector machine model optimized by an adaptive particle swarm algorithm, which overcomes the non-optimal particle problem that may occur when using the adaptive particle swarm algorithm alone, as well as the memory consumption and complex parameter selection problems caused by matrix inversion when using LSSVM alone. It not only improves the prediction accuracy and generalization ability of the model, but also significantly improves the computational efficiency and global search ability. By executing the adaptive particle swarm algorithm, the optimal penalty parameters and kernel parameters can be obtained, and the least squares support vector machine quantitative analysis model can be reconstructed using the optimized parameters, which will greatly reduce the measurement error caused by gas cross-sensitivity and ensure the accuracy of gas measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. In the accompanying drawings:

[0026] Figure 1 is a block diagram of a gas collection system according to a specific embodiment of the present invention;

[0027] Figure 2 is a circuit diagram of a collection unit according to a specific embodiment of the present invention;

[0028] Figure 3 It is a gas calibration system in a specific embodiment of the present invention;

[0029] Figure 4 is a schematic diagram of a gas chamber in a gas calibration system in a specific embodiment of the present invention;

[0030] Figure 5 is a schematic diagram of a drying tube in an air chamber in a specific embodiment of the present invention;

[0031] Figure 6 is a flow chart of a detection method according to a specific embodiment of the present invention;

[0032] Figure 7 It is the BP neural network topology structure in the specific embodiment of the present invention;

[0033] Figure 8 is a flow chart of temperature optimization using a PSO-BP neural network in a specific embodiment of the present invention;

[0034] Fig. 9 It is a flow chart of the gas calibration process in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0036] The present invention is further illustrated below in conjunction with specific embodiments. Those skilled in the art should understand that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention, and modifications to various equivalent forms of the present invention fall within the scope defined by the appended claims of this application.

[0037] like Figure 1 As shown, the present invention provides a gas detection system, including: a gas sensor 13, a six-choice module, a temperature and humidity sensor, an A / D converter, an STM32 data processor, a GPRS module, an audible and visual alarm device, and a central terminal device.

[0038] The circuit in the gas sensor 13 includes: six collection units, the surface of the round gas sensor is divided into six sector-shaped areas, which are, clockwise, a collection unit for detecting NO2 gas, a collection unit for detecting SO2 gas, a collection unit for detecting CO gas, a collection unit for detecting NO2 gas, a collection unit for detecting SO2 gas, and a collection unit for detecting CO gas. The collection units for collecting the same gas are arranged relatively to each other.

[0039] The output end of the gas sensor 13 is connected to the input end of the six-select-one module, the output end of the six-select-one module is connected to the input end of the A / D conversion module, and the output end of the A / D conversion module is connected to the STM32 data processor; the six-select-one module is used to sequentially connect the six acquisition units in the gas sensor to the A / D conversion module for sampling by the A / D conversion module;

[0040] The temperature and humidity sensor can collect the temperature and humidity of the air chamber or the environment, and the collected temperature and humidity are sent to the STM32 data processor; when the harmful gas reaches the alarm concentration, the STM32 data processor alarms through the sound and light alarm device; the STM32 data processor can upload the data to the central terminal device through the GPRS module connected to it.

[0041] like Figure 2As shown, the acquisition unit is a differential amplifier circuit, including: four gas-sensitive field effect transistors, four non-gas-sensitive field effect transistors and two transistors, the base of the transistor Q1 is connected to the drain of the gas-sensitive field effect transistor M5 and the source of the non-gas-sensitive field effect transistor M1 respectively; the collector of the transistor Q1 is connected to the source of the gas-sensitive field effect transistor M2 and serves as an output end of the acquisition unit; the emitter of the transistor Q1 is connected to the drain of the non-gas-sensitive field effect transistor M6; the transistor Q2 The base of the transistor is connected to the drain of the non-gas-sensitive field effect transistor M8 and the source of the gas-sensitive field effect transistor M4 respectively; the collector of the transistor Q2 is connected to the source of the non-gas-sensitive field effect transistor M3 and serves as another output end of the acquisition unit; the emitter of the transistor Q2 is connected to the drain of the gas-sensitive field effect transistor M7; the drain of the non-gas-sensitive field effect transistor M1, the drain of the gas-sensitive field effect transistor M2, the drain of the non-gas-sensitive field effect transistor M3 and the drain of the gas-sensitive field effect transistor M4 are connected to the external power supply U dd The source of the gas-sensitive field effect tube M5, the source of the non-gas-sensitive field effect tube M6, the source of the gas-sensitive field effect tube M7 and the source of the non-gas-sensitive field effect tube M8 are grounded; the gates of the four gas-sensitive field effect tubes and the gates of the four non-gas-sensitive field effect tubes are all connected to the external power supply U dd ;

[0042] Through the external power supply U dd Ensure that the gate-source voltage of all field effect tubes is consistent, because the conductivity and resistance of the gas-sensitive field effect tube are not only related to the gas concentration in contact with the gas-sensitive material, but also to the applied gate voltage. In order to ensure that the changes in the conductivity and resistance of the gas-sensitive field effect tube are only due to the gas concentration, all field effect tubes apply the same gate voltage to control the variables. Different gas concentrations correspond to different conductivity and resistance values ​​of the gas-sensitive field effect tube, so the output signal corresponding to each concentration value is different. When the gas concentration changes, the resistance values ​​of the four gas-sensitive field effect tubes will change, resulting in changes in the voltage divider and the current size of each branch of the circuit, which in turn leads to different output values ​​of the differential circuit. When the gas concentration increases, the gas reacts chemically on the surface of the gas-sensitive material of the gas-sensitive field effect tube to generate more carriers, which in turn leads to a decrease in the resistance of the gas-sensitive field effect tube, a decrease in the base voltage of the transistor Q1, a decrease in the emitter voltage, and an increase in the emitter current I e1 becomes smaller, the collector current I c1 with I e1 There is a multiple relationship, I c1 Will follow I e1 , so I c1 also becomes smaller, so U c1 becomes larger, the base voltage of transistor Q2 increases, the emitter voltage increases, and the emitter current I e2 becomes larger, the collector current I c2 For I e2 There is a multiple relationship, Ic2 Will follow I e2 , so I c2 also becomes larger, so U c2 When the gas concentration decreases, the voltage signal of the differential output increases. Similarly, when the gas concentration decreases, the voltage signal of the differential output decreases. Therefore, the size of the gas concentration can be measured by the size of the differential output signal.

[0043] The six collection units are divided into three groups to collect the concentrations of three gases. Two collection units form a group and use the same gas-sensitive field effect tube to collect the corresponding gas concentrations.

[0044] like Figure 6 As shown, the present invention also provides a detection method of a gas detection system, comprising the following steps:

[0045] Step S1: Obtaining a preliminary gas calibration formula of the gas sensor for each gas;

[0046] like Figure 3 As shown, the gas calibration system includes: a gas chamber 1, a waste treatment device 2, a gas distribution chamber 3, a N2 gas cylinder 4, a NO2 gas cylinder 5, a SO2 gas cylinder 6, a CO gas cylinder 7, a data processor 8 and a host computer 9;

[0047] The four gas cylinders are connected to the gas distribution chamber 3 through pipelines and control valves 10, the gas distribution chamber 3 is connected to the gas inlet 11 of the gas chamber 1, the gas outlet 12 of the gas chamber is connected to the waste treatment device 2, and the data processor 8 is connected to the host computer 9;

[0048] like Figure 4 As shown, the gas sensor is a circular PCB board 13, and a semi-spherical shell 14 is provided on the upper cover to form an air chamber 1. The semi-spherical shell 14 is provided with an air inlet 11 and an air outlet 12. Figure 5 As shown, a gas drying tube 15 is provided at both the gas inlet and outlet. The gas drying tube 15 dries and filters the gas in and out of the gas chamber 1 through the PVDF filter membranes 151 at both ends and the 3A molecular sieve 152 filled inside.

[0049] like Fig. 9 As shown, the specific gas calibration steps are as follows:

[0050] Step S11: assembling the system and performing air tightness testing;

[0051] Step S12: 99.999% high-purity nitrogen is introduced into the air chamber to exhaust the gas inside the air chamber, and the air tightness of the device is indirectly checked by detecting the humidity using a temperature and humidity sensor;

[0052] Step S13: mixing one of the three gases with high-purity nitrogen in a gas distribution chamber at corresponding concentrations;

[0053] Step S14: setting the temperature and humidity of the detection environment, passing a gas with a concentration ratio obtained in step S13 into the gas chamber, maintaining its flow rate slightly greater than that of air for about two minutes, recording the voltage value output by the differential amplifier circuit of the acquisition unit, and then cleaning the gas chamber with high-purity nitrogen, repeating steps S13 and S14, obtaining the voltage value output by the differential amplifier circuit of the acquisition unit under different concentrations of the current gas; using the voltage value output by the differential amplifier circuit as the x-axis and the corresponding standard gas concentration as the y-axis, fitting a calibration curve between the voltage value and the standard gas concentration according to the collected data; one gas is collected by two acquisition units, and the calibration curves of the two acquisition units for the gas can be obtained, obtaining the voltage values ​​of the two differential amplifier circuits under the same concentration, averaging the two output voltage values, obtaining the corresponding mean voltages under all concentrations, and then fitting the mean voltage with the corresponding standard concentration, obtaining the final fitting curve and using it as the gas calibration formula / curve for the current gas;

[0054] Step S15: After all concentration tests of one gas are completed, the remaining two mixed gases are sequentially operated according to the above steps S13 and S14 to complete the calibration of the three gases and obtain the calibration formulas of the three gases.

[0055] Six differential amplifier circuits are divided into three categories by setting different types of gas-sensitive field effect tubes in the differential amplifier circuits, namely, two differential amplifier circuits for measuring NO2 gas, two differential amplifier circuits for measuring SO2 gas, and two differential amplifier circuits for measuring CO gas. Each gas corresponds to two measurement circuits, which are arranged in a matrix. The output signals of the two circuits are processed by averaging respectively, and the fitting curves of the averages of the two differential circuits corresponding to each of the three gases are selected as the fitting curves of the gas, which can reduce the uncertainty caused by the error of a single measurement circuit and improve the measurement accuracy.

[0056] Step S2: Figure 7 , 8 As shown, the ambient temperature and humidity are obtained, and combined with the ambient temperature and humidity, the PSO-BP neural network algorithm combining the BP neural network algorithm and the particle swarm algorithm is used to optimize the temperature compensation of the preliminary gas calibration formula;

[0057] The constructed PSO-BP neural network adopts a three-layer neural network model with 1 hidden layer. The number of input layer neurons is set to 4, which are the actually measured NO2 gas concentration value, SO2 gas concentration value, CO gas concentration value and the corresponding temperature and humidity T. The number of output neurons is 3, which are the NO2 gas concentration value, SO2 gas concentration value, and CO gas concentration value after temperature and humidity compensation. The number of hidden layer neurons is 5. In the PSO-BP neural network model, the particles of the particle swarm algorithm represent the weights and thresholds in the network. Through iterative search, the optimal weights and thresholds are found, and then used as the initial weights and thresholds of the BP neural network, and substituted into the neural network for training and prediction.

[0058] Step S3: optimizing the gas cross sensitivity of the gas calibration formula optimized in step S2 by using a least squares support vector machine optimized by an adaptive particle swarm algorithm to obtain a final gas calibration formula;

[0059] The process of optimizing the least squares support vector machine by adaptive particle swarm algorithm: the mean square error of the output result of the preliminary regression model of LSSVM is used as the fitness value of the adaptive particle swarm algorithm, and the adaptive particle swarm algorithm is executed, and the optimal solution is recorded and updated until the optimization result meets the set mean square error or reaches the maximum number of iterations. The optimal solution is output and substituted into the initially constructed LSSVM model to reconstruct the LSSVM model and obtain the LSSVM optimized by the adaptive particle swarm algorithm.

[0060] Temperature changes have a significant impact on gas sensors. First, temperature changes the catalytic activity, thereby affecting the adsorption, desorption and surface reaction rates of gas molecules on the sensor surface. Second, temperature fluctuations may affect the stability of metal oxide semiconductor materials, resulting in a decrease in their performance. In addition, temperature will affect the distribution and mobility of carriers in semiconductor materials, thereby affecting their electrical properties. These changes will ultimately affect the sensitivity and selectivity of the sensor to different gases. By optimizing the temperature compensation first and then the gas cross-sensitivity optimization, the direct impact of temperature changes on the sensor output signal can be eliminated or reduced, making the subsequent gas cross-sensitivity compensation more accurate and effective, ensuring the accuracy and consistency of the entire measurement process, thereby improving the accuracy and reliability of gas measurement.

[0061] Step S4: The voltage data corresponding to each gas in the environment is collected in real time by the gas sensor, and the concentration of each gas is calculated according to the voltage data by the final gas calibration formula.

[0062] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A gas detection system, characterized in that: include: A gas sensor and a data processor; the output end of the gas sensor is connected to the input end of the data processor; The gas sensor collects three gases through three groups of circuits including gas-sensitive field effect tubes; the data processor calculates the concentration of the three gases based on the output signals of the gas sensors. Among them, the circuit in the gas sensor includes: six acquisition units, the six acquisition units form three groups of circuits including gas-sensitive field-effect tubes; the circuit of the acquisition unit specifically includes: four gas-sensitive field-effect tubes, four non-gas-sensitive field-effect tubes and two triodes, the base of the first triode is respectively connected to the drain of the first gas-sensitive field-effect tube and the source of the first non-gas-sensitive field-effect tube; the collector of the first triode is connected to the source of the second gas-sensitive field-effect tube and serves as an output end of the acquisition unit; the emitter of the first triode is connected to the drain of the second non-gas-sensitive field-effect tube; the base of the second triode is respectively connected to the drain of the third non-gas-sensitive field-effect tube and the source of the third gas-sensitive field-effect tube The source of the field effect tube is connected; the collector of the second triode is connected to the source of the fourth non-gas-sensitive field effect tube and serves as another output end of the acquisition unit; the emitter of the second triode is connected to the drain of the fourth gas-sensitive field effect tube; the drain of the first non-gas-sensitive field effect tube, the drain of the second gas-sensitive field effect tube, the drain of the fourth non-gas-sensitive field effect tube and the drain of the third gas-sensitive field effect tube are connected to an external power supply; the source of the first gas-sensitive field effect tube, the source of the second non-gas-sensitive field effect tube, the source of the fourth gas-sensitive field effect tube and the source of the third non-gas-sensitive field effect tube are grounded; the gates of the four gas-sensitive field effect tubes and the gates of the four non-gas-sensitive field effect tubes are all connected to an external power supply; The six collection units are divided into three groups to collect three gas concentrations. Among them, two collection units form a group and use the same gas-sensitive field-effect tube to collect the corresponding gas concentrations. The six collection units are arranged in an array on the PCB circuit board, and a pair of collection units collecting the same gas are relatively set.

2. The gas detection system according to claim 1, characterized in that: The gas sensor also includes a temperature and humidity sensor, the output end of the temperature and humidity sensor is connected to the input end of the data processor, and the data processor optimizes the calculated gas concentration according to the temperature data collected by the temperature and humidity sensor.

3. A detection method for a gas detection system, applicable to the gas detection system as claimed in any one of claims 1 to 2, characterized in that: The detection method of the gas detection system comprises the following steps: Step 1: Obtain the preliminary gas calibration formula of the gas sensor for each gas; Step 2: Obtain the ambient temperature and humidity, and optimize the temperature compensation of the preliminary gas calibration formula through a neural network based on the ambient temperature and humidity; Step 3: Optimize the gas cross sensitivity of the gas calibration formula optimized in step 2 by using the least squares support vector machine to obtain the final gas calibration formula; Step 4: Use the gas sensor to collect the voltage data corresponding to each gas in the environment in real time, and calculate the concentration of each gas based on the voltage data using the final gas calibration formula.

4. The detection method of the gas detection system according to claim 3, characterized in that: The output voltages of the collection units that collect the same gas are averaged, and the voltage after the average processing is used as the voltage data corresponding to the gas concentration.

5. The detection method of the gas detection system according to claim 3, characterized in that: The neural network in step 2 is: PSO-BP neural network.

6. The detection method of the gas detection system according to claim 3, characterized in that: The step 3 also includes: optimizing the penalty parameter and kernel parameter of the least squares support vector machine by using an adaptive particle swarm algorithm.

Images