Gas sensor and detection system for detecting partial discharge of AIS (Automatic Identification System) equipment

Abstract

The utility model provides a gas sensitive sensor and a detection system for detecting partial discharge of AIS (Automatic Identification System) equipment, belonging to the technical field of partial discharge detection. The gas-sensitive sensor for detecting the partial discharge of the AIS equipment comprises a PET flexible substrate and n sections of interdigital electrodes which are positioned on the PET flexible substrate, wherein the surfaces of the interdigital electrodes are coated with doped modified graphene gas-sensitive films; each section of interdigital electrode shares a wire inlet tap and is provided with a wire outlet tap; the type of a thin film used by each section of interdigital electrode is one of a metal doping type and a non-metal doping type; the gas-sensitive sensor outputs n kinds of gas-sensitive signals and supports the acquisition device to acquire 2n-1 kinds of gas-sensitive signals in a self-selection doubling mode. The AIS partial discharge detection system is composed of a gas sensor, an n-in m-out multichannel acquisition module and an upper computer, and m belongs to [1, 2n-1]. And the upper computer is used for identifying the component type and the concentration interval of the partial discharge characteristic decomposition component mixed gas in the switch equipment according to the gas sensitive signal transmitted by the multi-channel acquisition module.

Description

Technical Field

[0001] The utility model relates to the technical field of partial discharge detection, in particular to a gas sensor and a detection system for detecting partial discharge of AIS equipment. Background Art

[0002] Abnormal partial discharge is an important factor that reduces the operating life of power equipment, and is also one of the common abnormal operating conditions of various power equipment. Partial discharge detection is a key engineering technology and an important basis for judging whether the operating status of many power equipment is normal. Implementing online partial discharge monitoring and timely detection of abnormal discharge can formulate more reasonable operation plans and maintenance strategies based on a full understanding of the actual condition of the equipment, extend its service life and service life, and reduce the cost of the equipment throughout its life cycle.

[0003] When partial discharge occurs inside the air-insulated AIS equipment, high-energy particles are generated. The energy generated causes the air insulation medium to decompose and ionize, generating a variety of new gas components. At present, NO2, O3 and CO are generally identified as the characteristic decomposition components of AIS discharge. These components will corrode metal electrodes, shells and organic insulating materials, further deteriorate the insulation strength of the equipment, and endanger the safety of equipment operation. On the other hand, the type and concentration of the components are closely related to AIS failures, so the insulation failures and severity inside the equipment can be diagnosed by decomposing the gas.

[0004] At present, there is still a lack of convenient and effective detection means and detection technology for the decomposition components of AIS discharge characteristics, especially the accurate detection of each discharge characteristic component in a mixed gas environment, which limits the promotion and application of miniaturized online monitoring devices. At present, the gas detection and monitoring technologies that are studied more include gas chromatography and spectral detection. Chromatography is to use a carrier gas to send a small amount of sample gas into a porous body called a chromatographic column, and use the different diffusion rates of different component gases in the chromatographic column to achieve the separation of mixed gases and ultimately achieve the purpose of gas analysis. However, this method has a long analysis cycle and a slow reaction speed during online measurement; Spectrometry is to use the intensity of the absorption of the gas to be measured at a specific characteristic absorption frequency to determine the concentration of the gas to be measured. It has the advantage of fast analysis speed, but due to the limitation of laser monochromaticity, the components detected online are often relatively single, and when the mixed gas is relatively complex, there is a problem of cross-influence of different gases on the measurement results, and the accuracy is difficult to achieve the same as chromatography. The online monitoring of AIS equipment requires the detection equipment to quickly and accurately feedback data in order to formulate more reasonable operation plans and maintenance strategies in a timely manner. Existing technologies are difficult to meet the needs of high-precision, efficient monitoring, and fast feedback. Utility Model Content

[0005] In view of this, the utility model provides a gas sensor and a detection system for detecting partial discharge of AIS equipment, providing a variety of gas-sensitive signal rapid acquisition solutions and high-precision rapid data feedback solutions, thereby improving the efficiency of online monitoring of AIS equipment.

[0006] The technical solution adopted by the embodiment of the utility model to solve the technical problem is:

[0007] A gas sensor for detecting partial discharge of AIS equipment, comprising a PET flexible substrate, and n sections of interdigital electrodes on the PET flexible substrate, the surfaces of which are sprayed with doped modified graphene gas-sensitive films, n≥2; each section of the interdigital electrodes shares a common input tap and each has an output tap, the input tap and the n output taps are used to connect to a collection device; the film type used in each section of the interdigital electrodes is one of a metal-doped type and a non-metal-doped type; the gas sensor has n types of gas-sensitive signal outputs, and supports the collection device to perform 2 n -1 type of gas sensitive signal acquisition.

[0008] Preferably, the gas sensor is used to sense three characteristic decomposition components of NO2, O3 and CO produced by partial discharge.

[0009] Preferably, the interdigitated electrodes are made of Au-plated material, have a thickness of 30 μm, and an electrode spacing of 0.10 mm; and the thickness of the doped modified graphene gas-sensitive film is 250 to 500 nm.

[0010] Preferably, the metal doping type is: using graphene oxide and metal salt solution as raw materials to prepare metal doped modified graphene, using graphene dimethylformamide DMF solution to disperse the metal doped modified graphene to obtain a metal doped graphene dispersion, and spraying the upper clear liquid of the metal doped graphene dispersion on the doped modified graphene gas sensitive film formed by the interdigitated electrodes;

[0011] The metal salt solution is one of palladium chloride PdCl2 solution, ammonium hexachloroplatinate (NH4)2PtCl6 solution, and gold chloride AuCl3 solution;

[0012] According to the difference in raw materials used, the metal doping type is further divided into Pd doping type, Pt doping type and Au doping type.

[0013] Preferably, the non-metallic doping type is further divided into B-doping type and N-doping type, wherein:

[0014] The B-doped type is: using boron oxide B2O3 powder and graphene oxide powder as raw materials to prepare B-doped modified graphene, using graphene dimethylformamide DMF solution to disperse the B-doped modified graphene to obtain a B-doped graphene dispersion, and spraying the upper clear liquid of the B-doped graphene dispersion on the doped modified graphene gas-sensitive film formed by the interdigitated electrodes;

[0015] The N-doped type is as follows: N-doped modified graphene is prepared by using graphene oxide powder and NH3, Ar mixed gas as raw materials, the N-doped modified graphene is dispersed by using graphene dimethylformamide DMF solution to obtain an N-doped graphene dispersion, and the upper clear liquid of the N-doped graphene dispersion is sprayed on the doped modified graphene gas-sensitive film formed by the interdigitated electrodes.

[0016] Preferably, the doped modified graphene gas-sensitive film on the surface of the interdigitated electrode is one of Pd-doped, Pt-doped, Au-doped, B-doped and N-doped, and the film types used in adjacent segments are the same or different.

[0017] Preferably, let n=3, and the size of the three-segment doped electrode is set to 35 mm×10 mm×50 μm; or, let n=5, and the size of the five-segment doped electrode is set to 55 mm×10 mm×50 μm.

[0018] An AIS partial discharge detection system includes a gas sensor, a multi-channel acquisition module, and a host computer that are electrically connected in sequence, wherein:

[0019] The gas sensor is the aforementioned gas sensor for detecting partial discharge of the AIS device, and the gas sensor is placed inside the AIS device;

[0020] The multi-channel acquisition module is an n-input and m-output signal acquisition device, wherein the n input ports are connected to the n output taps of the gas sensor, and the m output ports are connected to the host computer, m∈[1,2 n -1];

[0021] The host computer is used to identify the component attributes of each partial discharge characteristic decomposition component in the AIS device according to the m gas-sensitive signals collected by the multi-channel acquisition module, and further derive the partial discharge fault type according to the component combination state of the mixed gas, wherein the component attributes include none, low concentration, and high concentration, and the component combination state refers to the combination of the attributes corresponding to each component.

[0022] Preferably, in the multi-channel acquisition module, a series-parallel acquisition circuit is provided between the input port and the output port, wherein the output port is connected in series with one of the input ports, or in parallel with at least two of the input ports.

[0023] Preferably, the host computer is used for:

[0024] By establishing an analysis model based on a BP neural network, the model is trained using the component concentration properties of the sample gas and the gas-sensitive signal of the gas sensor;

[0025] The components in the mixed gas are qualitatively analyzed by using the trained analysis model, the component attributes of each partial discharge characteristic decomposition component in the AIS device are identified, and the component combination state of the mixed gas is further obtained; 27 component combination states are obtained according to the number of attributes 3 and the number of components 3, and the host computer is used to identify the 27 partial discharge states;

[0026] Based on the partial discharge fault types corresponding to the preset 27 partial discharge states, the host computer derives the partial discharge fault type according to the component combination state.

[0027] It can be seen from the above technical scheme that the gas sensor and detection system for detecting partial discharge of AIS equipment provided by the embodiment of the utility model. The gas sensor for detecting partial discharge of AIS equipment includes a PET flexible substrate, and n segments of interdigital electrodes on the PET flexible substrate with doped modified graphene gas-sensitive films sprayed on the surface; each segment of the interdigital electrode shares a common input tap and each has an output tap; the film type used in each segment of the interdigital electrode is one of metal-doped type and non-metal-doped type; the gas sensor has n types of gas-sensitive signal outputs, and supports the acquisition device to perform 2 n -1 type of gas-sensitive signal acquisition. The AIS partial discharge detection system consists of a gas-sensitive sensor, an n-input and m-output multi-channel acquisition module and a host computer, m∈[1,2 n -1], the host computer is used to identify the component attributes of each partial discharge characteristic decomposition component in the AIS device according to the m gas-sensitive signals collected by the multi-channel acquisition module, and further obtain the partial discharge fault type according to the component combination state of the mixed gas. The solution of the utility model can provide a variety of gas-sensitive signal fast acquisition solutions and high-precision fast data feedback solutions, thereby improving the efficiency of online monitoring of AIS equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 Schematic diagram of the three-section structure of the gas sensor used to detect partial discharge of AIS equipment.

[0029] Figure 2 Schematic diagram of the five-segment structure of the gas sensor used to detect partial discharge of AIS equipment.

[0030] Figure 3 Schematic diagram of the AIS partial discharge detection system.

[0031] Figure 4Schematic diagram of the preparation process of Pd, Pt and Au doped modified graphene respectively.

[0032] Figure 5 Schematic diagram of the preparation process of B and N doped modified graphene.

[0033] In the figure: 1-PET flexible substrate, 2-interdigitated electrode, 3-doped modified graphene gas sensitive film, A-inlet tap, B-outlet tap, C-outlet tap, D-outlet tap, E-outlet tap. DETAILED DESCRIPTION

[0034] The technical solution and technical effects of the present invention are further described in detail below in conjunction with the accompanying drawings of the present invention.

[0035] The utility model provides a gas sensor for detecting partial discharge of AIS equipment, which consists of a PET flexible substrate and n-segment interdigital electrodes on the PET flexible substrate with surfaces sprayed with doped modified graphene gas sensitive films (n≥2, indicating that multiple-segment electrodes are used for collection), and is specifically used for sensing three characteristic decomposition components of NO2, O3 and CO generated by partial discharge.

[0036] Reference Figure 1 The three-segment gas sensor (n=3) example shown above uses PET flexible interdigital electrode segments to prepare nano-micro sensors. The size of the sensor module is set to 35mm×10mm×50μm, which meets the requirements of sensor miniaturization. The distance between adjacent electrodes is 0.10mm, which has high precision. The interdigital electrode pieces are made of Au-plated material, which provides a guarantee for the sensitivity of the sensor. The flexible characteristics of PET material improve the plasticity and installation convenience of the sensor. Among them, each segment of the interdigital electrode shares a common input tap and each has an output tap. The input tap and n output taps are used to connect to the acquisition device. The gas sensor has n types of gas-sensitive signal outputs, which can support the acquisition device to conduct 2 n -1 type of gas sensitive signal acquisition.

[0037] The doped modified graphene gas-sensitive film used in each segment of the interdigital electrode is one of metal-doped (Pd, Pt or Au) or non-metal-doped (B or N), and is prepared by segmented spraying. Figure 1 1 is a PET substrate, 2 is a forked electrode, and 3 is a doped graphene gas-sensitive film. The doped graphene gas-sensitive film is evenly dispersed on the surface of the forked electrode by spraying, which is conducive to the adsorption and desorption of gas molecules. Different types of doped modified graphene gas-sensitive films can be sprayed on each section of the gas sensor according to needs to form different combinations. Regarding the preparation of segmented gas sensors, the utility model provides a feasible implementation method, and the process is as follows:

[0038] (1) First, prepare graphene oxide powder, refer to Figure 4 , as shown in (a)-(b), the preparation process of graphene oxide powder is as follows: 100 mg of graphene is added to a mixture of sulfuric acid (H2SO4: 98%, 87.5 mL) and nitric acid (HNO3: 68%, 27 mL), and then 500 mg of potassium chlorate (KClO3) powder is slowly added to the above mixture, cooled in an ice water bath, and stirred continuously for 96 hours; after the reaction is completed, the mixture is poured into 3 L of deionized water, and the filtrate is added with 5% hydrochloric acid (HCl) to be thoroughly washed to remove residual impurities, and then filtered and dried to obtain graphene oxide.

[0039] (2) Using dopants to prepare doped modified graphene from graphene oxide powder.

[0040] Reference Figure 4 Process (b)-(d), for metal-doped type (divided into Pd-doped type, Pt-doped type, Au-doped type), 10 mg of graphene oxide is dispersed in metal salt solution (1 mg of metal salt: 10 mL of water), and the metal salts are palladium chloride (PdCl2), ammonium hexachloroplatinate ((NH4)2PtCl6)) and gold chloride (AuCl3); secondly, the mixed solution is dried at 60°C in a vacuum furnace for 48 hours to produce a metal-doped modified graphene precursor; finally, the precursor is calcined at 500°C for 12 minutes in a high-purity N2 protective gas to obtain the final product of metal-doped modified graphene;

[0041] Reference Figure 5 For non-metallic doping types (divided into B-doping type and N-doping type), the utility model provides treatment solutions respectively. Figure 5 As shown in (c1)-(d1), boron oxide (B2O3) powder is covered on graphene oxide powder and calcined at 600°C for 4 h in an Ar flowing gas protection atmosphere to form B-doped modified graphene; Figure 5 As shown in (c2)-(d2), the graphene oxide powder was calcined in a NH3 / Ar (NH3:Ar=5%:95%) mixed gas atmosphere at 500°C for 1 h to form N-doped modified graphene.

[0042] (3) Dispersion of doped modified graphene.

[0043] Due to the inherent hydrophobicity of graphene, it does not have good dispersibility in a variety of organic solvents. At the same time, graphene prepared by the method of reducing graphene oxide is easy to agglomerate, affecting the excellent electrical properties of graphene itself. Therefore, the dispersion of graphene is a key step in the production of graphene gas sensors. After repeated experiments, it was found that dimethylformamide (DMF) organic solvent has a good effect on the dispersion of graphene. The dispersion steps are briefly described as follows: prepare a graphene dimethylformamide (DMF) solution with a concentration of 0.025 mg / mL, soak it for 8 hours, use an ultrasonic oscillator (power 100W) for 2 hours, and then centrifuge at a speed of 3000r / min to obtain a graphene dispersion with good effect.

[0044] (4) Preparation of segmented doped gas sensitive sensors.

[0045] In this embodiment, PET flexible interdigital electrodes are used to prepare segmented doped nano-micro sensors, such as Figure 1 and Figure 2 As shown. The size of the three-segment doped electrode is set to 35mm×10mm×50μm, and the size of the five-segment doped electrode is set to 55mm×10mm×50μm to meet the requirements of sensor miniaturization; each segment leads to the corresponding wiring tap, and the distance between adjacent electrodes is 0.10mm, which has high precision; the Au-plated material provides a guarantee for the sensitivity of the sensor; the flexible characteristics of the PET material improve the plasticity of the sensor and the convenience of installation.

[0046] Specifically, the gas sensor film can be prepared by spraying, with the interdigital electrode as the substrate for depositing the doped modified graphene, the substrate thickness is 30 μm, and the spacing between the electrodes is 0.10 mm. The graphene dispersions with different doping modifications are sprayed on the surface of the interdigital electrode at 80°C in a segmented windowing manner to accelerate the evaporation of DMF. By controlling the solution concentration, spraying speed and time, the different doping modified graphenes are evenly dispersed in segments on the surface of the corresponding interdigital electrodes. Thus, the gas sensor of the utility model is formed.

[0047] Therefore, for the n-segment gas sensor, the doped modified graphene gas sensitive film on the surface of each segment of the interdigitated electrode is one of Pd-doped, Pt-doped, Au-doped, B-doped and N-doped types. The types of gas sensitive films used in each segment of the same gas sensor can be different or repeated. Repeated use of the same type can also serve as a comparison.

[0048] The gas sensor of this embodiment has n outgoing line taps, corresponding to each gas sensing section, and the outgoing line taps are used to connect to the collection device, which can support the collection device to perform 2 n -1 type of gas-sensing signal acquisition, for example, Figure 1The three-stage doping and Figure 2 The five-stage doping shown in the figure can be even more doping forms. Each stage is connected in parallel to share an incoming tap (A), and each stage is independently set with an outgoing tap ( Figure 1 B, C, D), ( Figure 2 The output taps of n-segment metal-doped graphene gas-sensitive films can selectively generate m sensor signals through a multi-channel selection module. Taking the three-segment type as an example, the three-segment taps B, C, and D can independently output three gas-sensitive signals B, C, and D, or can output three mixed-doped gas-sensitive signals through BC parallel connection, CD parallel connection, and BCD parallel connection. Therefore, the three-segment doped gas sensor can output 6 different gas-sensitive signals, which effectively enhances the data features, provides more abundant data samples for subsequent intelligent identification, and improves the detection sensitivity and response speed of the system.

[0049] The power supply of the gas sensor can be provided by an external power supply or through the port of the acquisition device.

[0050] Further, refer to Figure 3 As shown, the utility model also provides an AIS partial discharge detection system for detecting partial discharge of AIS equipment, comprising a gas sensor, a multi-channel acquisition module, and a host computer electrically connected in sequence, wherein:

[0051] The gas sensor is the aforementioned gas sensor for detecting partial discharge of the AIS device, and the gas sensor is placed inside the AIS device (it can be fixedly installed);

[0052] The multi-channel acquisition module is an n-input and m-output signal acquisition device, where the n input ports are connected to the n output taps of the gas sensor, and the m output ports are connected to the host computer, m∈[1,2 n -1]; In the multi-channel acquisition module, a series-parallel acquisition circuit is provided between the input port and the output port, wherein the output port is connected in series with one of the input ports, or in parallel with at least two of the input ports.

[0053] The host computer is used to identify the component attributes of each partial discharge characteristic decomposition component in the AIS device according to the m gas-sensitive signals collected by the multi-channel acquisition module, and further derive the partial discharge fault type according to the component combination state of the mixed gas. Here, the component attributes include none, low concentration, and high concentration, and the component combination state refers to the combination of the attributes corresponding to each component.

[0054] The upper computer establishes a DBN analysis model, which is a fully connected network, including an input layer, three hidden layers and an output layer, with three layers of RBM and one layer of BP neural network, where: the number of neurons in the input layer is m, and the input data of the input layer is the normalized value of m signals output by the multi-channel acquisition module; the three RBMs are stacked in sequence, so that the input data of the input layer is output to the BP neural network after three times of RBM learning in the three hidden layers; the output layer has three neurons, which correspond to the properties of the three components of NO2, O3 and CO in the mixed gas respectively; the analysis model is used to perform qualitative analysis on the components, and according to the number of attributes 3 and the number of components 3, the analysis model obtains 27 combination states for the identification of 27 partial discharge states;

[0055] According to the three components and the three concentration ranges corresponding to each component, 81 groups of sample gas concentration ratio data were designed using a three-factor nine-level orthogonal test table, and a sample database was further established; a single sample data in the sample database includes at least 6 types of data, which are the actual concentration values ​​of the three components of NO2, O3 and CO in the sample gas, and the attribute label values ​​corresponding to the three components of NO2, O3 and CO obtained based on the correspondence between the concentration range and the attribute;

[0056] The mixed gas is modulated according to the sample data in the sample database. The initial gas-sensitive signal is collected by the gas sensor, and the multi-channel acquisition module performs secondary acquisition and outputs it to the host computer. The host computer uses the sample data to train the DBN analysis model until the training is completed;

[0057] Use the trained DBN analysis model to perform qualitative analysis on the components of the signals collected in real time by the gas sensor and the multi-channel acquisition module, and identify the component attributes of each partial discharge characteristic decomposition component in the current AIS equipment;

[0058] Based on the partial discharge fault types corresponding to the pre-set 27 partial discharge states, the partial discharge fault type is further derived according to the component combination state of the mixed gas. Specifically, a partial discharge state set is configured for each partial discharge fault type, and each partial discharge state set may have an intersection. A single partial discharge state may correspond to at least one partial discharge fault type. At this time, all partial discharge fault types corresponding to the state are proposed and represented in the recognition result.

[0059] Taking a three-stage doped gas sensor module as an example, by configuring a multi-channel selection module, 6 gas-sensitive signals are correspondingly generated from 3 sensor modules. Through experimental tests, training sample data is established in advance, and the concentrations of three characteristic components can be set to: 0 ppm, 0.2 ppm, 0.5 ppm, 1 ppm, 2 ppm, 5 ppm, 10 ppm, 20 ppm, 50 ppm. Since the combination method reaches 9×9×9 = 729 kinds, in order to improve the test efficiency, a three-factor nine-level high-strength orthogonal test table is adopted to obtain 81 groups of test concentrations, as shown in Table 1. The entire test process can automatically perform gas distribution and sensing data sampling through a program, thereby forming a training sample.

[0060] Table 1 Concentration levels of each gas sample (ppm)

[0061]

[0062]

[0063] To facilitate the analysis and comparison of the characteristic values of each sensing unit and reduce the neuron saturation, it is necessary to perform (01) normalization processing on the characteristic values of each sensing data sample to make them at the same order of magnitude:

[0064]

[0065] In the formula, N and n i are the normalized and non-normalized values respectively, min[n i and max[n i are the minimum and maximum values of the non-normalized array.

[0066] The CNN adopts the Adam optimizer, and the convolutional layer adopts the Relu activation function. The CNN flattens and extracts features from the dynamic response curve, and adds environmental condition variables such as humidity, air pressure, and long-term use time to correct the drift features, which serves as the input layer of the DBN to further suppress the drift. Then it enters the DBN stacked by three RBMs. The DBN is trained through unsupervised learning, and then the structure in the DBN is fine-tuned using supervised learning. Specifically, it is divided into two stages. One is to train the intermediate hidden parameters using the Contrastive Divergence method, and the other is to fine-tune the network weights and biases using stochastic gradient descent. The DBN is finally connected to a fully connected layer to realize the qualitative analysis of gas mixture classification through the softmax function. The classification model divides each component into three categories, namely none (0 ppm), low concentration (0 ppm < C < 10 ppm), and high concentration (C ≥ 10 ppm), a total of 27 categories, and labels the input samples. The loss function of classification prediction adopts Cross Entropy.

[0067] For regression prediction, samples are labeled by linear normalized concentrations of 0 to 50 ppm. CNN is trained as a regression network, and the structure and parameter settings are the same as the classification algorithm. The number of RBM stacks in DBN remains unchanged, and only the last layer output is a regression output with three neurons, which provide normalized concentration values ​​of three components, NO2, CO and O3. For each component, the loss function of concentration prediction is the root mean square error (RMSE). All training sets and test sets in classification and regression are divided into 80% and 20% and a 5-fold cross validation is performed to prevent overfitting.

[0068] Through data analysis, the component types of each local discharge characteristic decomposition component in the AIS equipment can be quickly obtained, and the local discharge fault type can be further obtained according to the component combination state of the mixed gas. It can also be judged in real time whether there is local discharge phenomenon and the degree of local discharge in the current AIS equipment, which provides assistance for the maintenance of AIS equipment and the stable operation of the power grid.

[0069] The above disclosure is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of the present invention. A person skilled in the art can understand that all or part of the processes of the above embodiments and equivalent changes made according to the claims of the present invention still fall within the scope of the present invention.

Claims

1. A gas sensor for detecting partial discharge of AIS equipment, characterized in that: It comprises a PET flexible substrate, and n segments of interdigital electrodes on the PET flexible substrate, the surfaces of which are sprayed with doped modified graphene gas-sensitive films, where n≥2; each segment of the interdigital electrodes shares an incoming line tap and each has an outgoing line tap, and the incoming line tap and the n outgoing line taps are used to connect to a collection device; The film type used for the interdigital electrodes in each section is one of metal-doped type and non-metal-doped type; the gas sensor has n types of gas-sensitive signal outputs, and supports the acquisition device to perform 2 n -1 type of gas sensitive signal acquisition.

2. The gas sensor for detecting partial discharge of AIS equipment according to claim 1, characterized in that: The gas sensor is used to sense three characteristic decomposition components of NO2, O3 and CO generated by partial discharge.

3. The gas sensor for detecting partial discharge of AIS equipment according to claim 1, characterized in that: The interdigital electrodes are Au-plated, 30 µm thick, and the electrode spacing is 0.10 mm; the thickness of the doped modified graphene gas-sensitive film is 250-500 nm.

4. The gas sensor for detecting partial discharge of AIS equipment according to claim 1, characterized in that: The metal doping type is further divided into Pd doping type, Pt doping type and Au doping type.

5. The gas sensor for detecting partial discharge of AIS equipment according to claim 2, characterized in that: The non-metal doping type is further divided into a B doping type and an N doping type.

6. The gas sensor for detecting partial discharge of AIS equipment according to claim 5, characterized in that: The doped modified graphene gas-sensitive film on the surface of the interdigital electrode is one of Pd-doped, Pt-doped, Au-doped, B-doped and N-doped types, and the film types used in adjacent sections are the same or different.

7. The gas sensor for detecting partial discharge of AIS equipment according to claim 1, characterized in that: Let n=3, the size of the three-segment doped electrode is set to 35 mm×10 mm×50 mm; or let n=5, the size of the five-segment doped electrode is set to 55 mm×10 mm×50 mm.

8. An AIS partial discharge detection system, characterized in that: It includes a gas sensor, a multi-channel acquisition module, and a host computer which are electrically connected in sequence, wherein: The gas sensor is a gas sensor for detecting partial discharge of an AIS device according to any one of claims 1 to 7, and the gas sensor is placed inside the AIS device; The multi-channel acquisition module is an n-input and m-output signal acquisition device, wherein the n input ports are connected to the n output taps of the gas sensor, and the m output ports are connected to the host computer, m∈[1,2 n -1].

9. The AIS partial discharge detection system according to claim 8, characterized in that: In the multi-channel acquisition module, a series-parallel acquisition circuit is provided between the input port and the output port, wherein the output port is connected in series with one of the input ports, or in parallel with at least two of the input ports.

Images