True transformer partial discharge simulation system and discharge type identification method

Abstract

Compared with an existing transformer partial discharge simulation experiment platform, the real transformer partial discharge simulation system and the discharge type identification method provided by the invention have the advantages that four partial discharge defect models are arranged, a transformer box body does not need to be frequently opened and closed during partial discharge simulation, the operation process is simplified, and the operation efficiency is improved. The experiment efficiency is improved; meanwhile, built-in acoustic-electric integrated sensors are installed at different positions of a real-model test transformer box body, and compared with installation positions of sensors in other test platforms, the installation positions of the sensors of the test platform can receive signals with wider incoming wave ranges, the coverage range of the sensors is enlarged, and detection blind areas are avoided. According to the discharge type identification method of the system, by extracting the equivalent bandwidth of the ultrasonic signal and the signal length of the ultrahigh frequency signal, more partial discharge characteristic parameters are mastered, a partial discharge type identification program is trained accordingly, and the identification accuracy can be continuously improved.

Description

Technical Field

[0001] The present invention relates to the technical field of partial discharge detection of power transformers, and in particular to a real transformer partial discharge simulation system and a discharge type identification method. Background Art

[0002] Transformers are core equipment for power transmission and distribution in power systems, and their construction costs are high. Insulation failure in transformers poses a serious threat to the safe operation of power grids. Therefore, ensuring the safe and stable operation of transformers is crucial to the safety of power grids. Transformer insulation materials operate in high voltage and high temperature environments for a long time, and are prone to insulation degradation such as aging. Partial discharge is an early manifestation of transformer insulation degradation, so it is of great significance to effectively detect, diagnose and identify transformer partial discharge. For the diagnosis of transformer partial discharge defect types, existing methods usually construct and optimize diagnostic algorithms based on field measured data.

[0003] The detection, diagnosis and identification of partial discharge of transformers need to be based on a large amount of measured data, and accurate identification of defect types requires the combination of multiple sensing methods. The data detected on site are usually a single sensing method, lacking data from multi-sensor joint synchronous detection. Moreover, before the transformer is dismantled, it is usually impossible to accurately judge the defect type from the field data, making it difficult to diagnose and identify partial discharges. In addition, there are few samples of local discharge data on site, which makes it difficult to improve the accuracy of the constructed diagnostic algorithm model. Since it is difficult to obtain data suitable for field tests, existing methods also conduct experimental research based on the transformer partial discharge simulation platform, but the transformer models used are mostly simplified models, ignoring the structure of the transformer, or the actual transformer is scaled down, which cannot accurately reflect the propagation characteristics of the partial discharge signal inside the transformer. Summary of the invention

[0004] In view of this, the present invention proposes, on one hand, a real transformer partial discharge simulation system, comprising a real test transformer, a transformer partial discharge defect model and a partial discharge multi-parameter diagnosis system; the transformer partial discharge defect model comprises a plurality of discharge models, which are respectively arranged at different positions in the real experimental transformer box, and can perform different types of partial discharge defect simulations; the partial discharge multi-parameter diagnosis system comprises a data acquisition system, a data extraction system and a discharge type recognition system, the data acquisition system can collect electrical signals and acoustic signals generated by different types of partial discharge defects, the data extraction system performs feature extraction on the collected data, and then sends the data to the discharge type recognition system for recognition and judgment.

[0005] Furthermore, the box of the real experimental transformer is a rectangular parallelepiped, and the box is filled with transformer oil; there are iron cores, high-voltage and low-voltage windings inside the box, and the windings are simplified to a hollow cylindrical structure; there are 6 groups of bushings on the top of the box, 3 groups of high-voltage and low-voltage bushings each, and they are connected to the rising seat and flange structure.

[0006] Preferably, the box body is a cuboid with a size of 0.96m*0.54m*0.712m, and the wall thickness of the box body is 0.03m.

[0007] Furthermore, the transformer partial discharge defect model includes four discharge models: a suspended discharge model, a surface discharge model, a tip discharge model and an oil gap discharge model. Each discharge model is connected to a grounding rod, and the grounding rod is installed on the inner top of the real test transformer box. Each discharge model is placed in the real test transformer box.

[0008] Furthermore, the data acquisition system includes a plurality of built-in acoustic-electric integrated sensors, which are respectively arranged at different positions of the real experimental transformer box and can collect electrical signals and acoustic signals generated by different types of partial discharge defects.

[0009] Furthermore, the four discharge models of the transformer local discharge defect model are respectively placed in a closed placement tube, and a plurality of circular holes are arranged on the side wall of the placement tube. The local defect discharge signal is transmitted through the circular holes, and the transformer oil in the real test transformer box enters the placement tube through the circular holes.

[0010] Preferably, the placement tube is a plastic tube made of Teflon material, and both ends of the tube are made of metal.

[0011] Preferably, the dimensions of the placement tube are designed to have a bottom radius of 2 cm and a length of 5 cm; at the same time, four circular holes with a diameter of 5 mm are arranged on the side wall of the placement tube to facilitate the propagation of partial discharge signals and the entry of transformer oil into the placement tube.

[0012] Furthermore, the suspension discharge model includes a nylon screw arranged at one end of the box of the real experimental transformer and a metal screw arranged at the other end, which can simulate the suspension potential.

[0013] Furthermore, the creeping discharge model includes two metal rods, and a piece of quartz glass of preset specifications is placed between the two metal rods, which can simulate the creeping discharge.

[0014] Preferably, the quartz glass has a diameter of 11 mm and a thickness of 1.5 mm.

[0015] Furthermore, the tip discharge model includes a tungsten needle of a preset length connected to one pole of the box of the real experimental transformer, which can simulate needle tip discharge.

[0016] Preferably, the tungsten needle is a 2 cm tungsten needle.

[0017] Furthermore, the oil gap discharge model includes a gap between the positive electrode and the negative electrode, and the gap has a preset width. After the discharge model is placed in the placement tube, an oil gap is formed between the positive electrode and the negative electrode.

[0018] Preferably, the gap between the positive electrode and the negative electrode has a width of 4-6 mm.

[0019] A real transformer partial discharge simulation system provided by the present invention has the following advantages: compared with the existing transformer partial discharge simulation experimental platform, the present invention designs four different types of partial discharge defect models, and there is no need to frequently open and close the transformer box when simulating partial discharge, which simplifies the operation process and improves the experimental efficiency; at the same time, the present invention installs built-in acoustic and electrical integrated sensors at different positions of the real test transformer box. Compared with the installation position of the sensor in other experimental platforms, the sensor installation position of the experimental platform can receive signals with a wider range of incoming waves, thereby improving the coverage of the sensor and avoiding detection blind spots.

[0020] Another aspect of the present invention further provides a method for identifying a discharge type of a real transformer partial discharge simulation system as described in any one of the above items, comprising: Obtaining the original data of partial discharge ultrasonic signals and ultra-high frequency signals and storing them; Extract the original data and use feature extraction to obtain the equivalent bandwidth of the ultrasonic signal and the length of the UHF signal; The extracted equivalent bandwidth of the ultrasonic signal and the length of the ultra-high frequency signal are sent to the discharge type training module based on the convolutional neural network for accumulation training. When the discharge type training module accumulates a preset number of samples, the sample collection is stopped, the discharge type classification and recognition are performed, and the recognition result is given.

[0021] Preferably, when the number of samples extracted from the ultrasonic signal equivalent bandwidth data reaches 1000, sample collection is stopped.

[0022] Preferably, when the number of samples extracted from the length data of the UHF signal reaches 1000, sample collection is stopped.

[0023] Furthermore, the data extraction includes equivalent bandwidth extraction and signal length extraction; in the equivalent bandwidth extraction process, the spectrum range from the peak frequency point to its attenuation of 3dB is extracted from the ultrasonic signal spectrum as the equivalent bandwidth of the ultrasonic data; at the same time, when extracting the signal length, the time length of 15% of the ultra-high frequency original data from the beginning to the signal peak is extracted as the signal length of the ultra-high frequency waveform data.

[0024] The present invention provides a discharge type identification method for a real transformer partial discharge simulation system. By extracting the equivalent bandwidth of an ultrasonic signal and the signal length of an ultra-high frequency signal, more partial discharge characteristic parameters are obtained, and the partial discharge type identification program is trained based on the parameters, thereby continuously improving the identification accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings: Figure 1 A schematic diagram of the structure of a real transformer partial discharge simulation system provided by an embodiment of the present invention; Figure 2 A schematic diagram of specific parameters of the insertion position of the top grounding rod of a real test transformer provided in an embodiment of the present invention; Figure 3 A process diagram of a method for identifying a discharge type in a real transformer partial discharge simulation system provided by an embodiment of the present invention; In the figure, 1. Suspended discharge model, 2. Surface discharge model, 3. Tip discharge model, 4. Oil gap discharge model; A. Built-in acoustic-electric integrated sensor (1), B. Built-in acoustic-electric integrated sensor (2), C. Built-in acoustic-electric integrated sensor (3). DETAILED DESCRIPTION

[0026] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to be able to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0027] See also Figure 1As shown, it is a real transformer partial discharge simulation and multi-parameter joint diagnosis experimental platform provided by an embodiment of the present invention, including a real test transformer, a transformer partial discharge defect model and a partial discharge multi-parameter diagnosis system; the transformer partial discharge defect model includes a suspended discharge model, a surface discharge model, a tip discharge model and an oil gap discharge model, which are respectively arranged at different positions of the real experimental transformer box. In this embodiment, the four discharge models can correspond to the attached Figure 1 In the serial number, 1 is the installation position of the suspended discharge model, 2 is the installation position of the surface discharge model, 3 is the installation position of the tip discharge model, and 4 is the installation position of the oil gap discharge model, and the four discharge models can be interchanged at any position 1-4, and the four discharge models can simulate different types of local discharge defects; the local discharge multi-parameter diagnosis system includes a data acquisition system, a data extraction system and a discharge type recognition system, the data acquisition system can collect electrical signals and acoustic signals generated by different types of local discharge defects, the data extraction system extracts features from the collected data, and then sends them to the discharge type recognition system for recognition and judgment.

[0028] Continue to refer to Figure 1 As shown, the box of the real experimental transformer is a rectangular parallelepiped of 0.96m*0.54m*0.712m, the wall thickness of the box is 0.03m, and the box is filled with transformer oil; there are iron cores, high-voltage and low-voltage windings inside the box, and the windings are simplified to a hollow cylindrical structure; there are 6 groups of bushings on the top of the box, 3 groups of high-voltage and low-voltage bushings, and they are connected to the riser and flange structure.

[0029] Reference Figure 1 As shown, in this embodiment, four partial defect discharge models, suspension discharge model, surface discharge model, tip discharge model and oil gap discharge model, each discharge model is connected to a grounding rod, and the grounding rod is installed on the inner top of the real test transformer box, refer to the attached figure Figure 1 At the positions of the middle numbers 1-4, the grounding rods are close to the top of the real test transformer box. The four grounding rods are set inside the real test transformer box, and one end is connected to a discharge model respectively. Each discharge model is placed in the real test transformer box. During the experiment, the top of the defective model is connected to the high-voltage winding, and the grounding rod is installed on the top of the real test transformer box. When pressurized, press the grounding rod of the corresponding model, and the top of the model contacts the high-voltage winding. When the applied voltage exceeds the starting discharge voltage of the defective model, the model will have partial discharge.

[0030] In this embodiment, the data acquisition system includes a plurality of built-in acoustic-electric integrated sensors, which are arranged at different positions of the real experimental transformer box. In this embodiment, three built-in acoustic-electric integrated sensors are arranged, which are respectively arranged at the adjacent Figure 1At A, B, and C, the electrical signals and acoustic signals generated by different types of local discharge defects can be collected, and the built-in acoustic-electric integrated sensors installed at A, B, and C are used to detect the electrical signals and acoustic signals generated by local discharge. In this embodiment, three built-in acoustic-electric integrated sensors are set and arranged at different positions, so that signals at different positions can be better detected.

[0031] In this embodiment, the four discharge models included in the multiple transformer partial discharge defect models are respectively placed in corresponding plastic placement tubes made of Teflon material, and the two ends of the tube are made of metal. Taking into full consideration the transformer voltage level, whether the transformer oil can enter the tube, the material insulation strength and the propagation of the partial discharge signal, the size of the tube is designed to have a bottom radius of 2 cm and a length of 5 cm. At the same time, 4 circular holes with a diameter of 5 mm are set on the side wall of the placement tube to facilitate the propagation of the partial discharge signal and facilitate the transformer oil in the real test transformer box to enter the placement tube through the circular holes.

[0032] In this embodiment, the method for making the suspended discharge model is: a nylon screw is set at one end of the model and a metal screw is set at the other end to simulate the suspended potential.

[0033] In this embodiment, the method for making the creeping discharge model is as follows: a piece of quartz glass with a diameter of 11 mm and a thickness of 1.5 mm is placed between two metal rods to simulate the creeping discharge.

[0034] In this embodiment, the method for making the tip discharge model is as follows: a tungsten needle with a length of 2 cm is connected to one pole of the model to simulate needle tip discharge.

[0035] In this embodiment, the method for making the oil gap discharge model is as follows: a gap of about 5 mm in width is left between the positive electrode and the negative electrode, and an oil gap is formed between the positive electrode and the negative electrode after the transformer oil enters the plastic tube.

[0036] In this embodiment, the specific parameters of the insertion position of the grounding rod installed on the top of the real test transformer box for the four partial defect discharge models are as follows: Figure 2 as shown in .

[0037] In the embodiment of the present invention, the process of identifying data in the partial discharge multi-parameter diagnosis system includes the following process: Figure 3 As shown: (1) The data acquisition system obtains the original data of partial discharge ultrasonic signals and ultra-high frequency signals and stores them; (2) Extract the original data. The data extraction system uses feature extraction to obtain the equivalent bandwidth of the ultrasonic signal and the length of the UHF signal; (3) The extracted ultrasonic signal equivalent bandwidth and the length of the ultra-high frequency signal are sent to a discharge type training module based on a convolutional neural network for accumulation training. When the number of samples of the ultrasonic signal equivalent bandwidth data extracted by the discharge type training module reaches 1000 or greater, sample collection is stopped; at the same time, when the number of samples of the ultra-high frequency signal length data extracted reaches 1000 or greater, sample collection is stopped; and the discharge type recognition system performs discharge type classification and recognition and gives a recognition result.

[0038] In this embodiment, the data extraction system includes an equivalent bandwidth extraction module and a signal length extraction module; the equivalent bandwidth extraction module extracts the spectrum range from the peak frequency point to its attenuation of 3dB from the ultrasonic signal spectrum as the equivalent bandwidth of the ultrasonic data; at the same time, the signal length extraction module extracts the time length of 15% of the UHF original data from the start to the signal peak as the signal length of the UHF waveform data.

[0039] A real transformer partial discharge simulation system and a discharge type identification method provided by the present invention have the following advantages: compared with the existing transformer partial discharge simulation experimental platform, the present invention does not need to frequently open and close the transformer box when simulating partial discharge, which simplifies the operation process and improves the experimental efficiency; at the same time, the present invention installs built-in acoustic and electrical integrated sensors at different positions of the real test transformer box. Compared with the installation positions of sensors in other experimental platforms, the sensor installation positions of this experimental platform can receive signals with a wider range of incoming waves, improve the coverage of the sensor, and avoid detection blind spots; the discharge point type identification method of the present invention, by extracting the equivalent bandwidth of the ultrasonic signal and the signal length of the ultra-high frequency signal, grasps more partial discharge characteristic parameters, and uses them to train the partial discharge type identification program, which can continuously improve the identification accuracy.

[0040] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0041] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0042] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A real transformer partial discharge simulation system, characterized in that: It includes a real test transformer, a transformer partial discharge defect model and a partial discharge multi-parameter diagnosis system; the transformer partial discharge defect model includes multiple discharge models, which are respectively arranged at different positions in the real test transformer box and can simulate different types of partial discharge defects; the partial discharge multi-parameter diagnosis system includes a data acquisition system, a data extraction system and a discharge type recognition system, the data acquisition system can collect electrical signals and acoustic signals generated by different types of partial discharge defects, the data extraction system extracts features from the collected data, and then sends them to the discharge type recognition system for recognition and judgment.

2. The real transformer partial discharge simulation system according to claim 1 is characterized in that: The transformer partial discharge defect model includes four discharge models: a suspended discharge model, a surface discharge model, a tip discharge model and an oil gap discharge model. Each discharge model is connected to a grounding rod, and the grounding rod is installed on the inner top of the real test transformer box. Each discharge model is placed in the real test transformer box.

3. The real transformer partial discharge simulation system according to claim 1 is characterized in that: The data acquisition system includes a plurality of built-in acoustic-electric integrated sensors, which are arranged at different positions of the real experimental transformer box and can collect electrical signals and acoustic signals generated by different types of partial discharge defects.

4. The real transformer partial discharge simulation system according to claim 2 is characterized in that: The four discharge models of the transformer partial discharge defect model are respectively placed in a closed placement tube. A plurality of circular holes are arranged on the side wall of the placement tube. The partial defect discharge signal is transmitted through the circular holes, and the transformer oil in the real test transformer box enters the placement tube through the circular holes.

5. The real transformer partial discharge simulation system according to claim 4 is characterized in that: The suspension discharge model includes a nylon screw arranged at one end of the box of the real experimental transformer and a metal screw arranged at the other end, which can simulate the suspension potential.

6. The real transformer partial discharge simulation system according to claim 4 is characterized in that: The creeping discharge model includes two metal rods, and a piece of quartz glass of preset specifications is placed between the two metal rods, which can simulate the creeping discharge.

7. The real transformer partial discharge simulation system according to claim 4, characterized in that: The tip discharge model includes a tungsten needle of a preset length connected to one pole of the box of the real experimental transformer, which can simulate needle tip discharge.

8. The real transformer partial discharge simulation system according to claim 4 is characterized in that: The oil gap discharge model comprises a gap between the positive electrode and the negative electrode, wherein the gap has a preset width. After the discharge model is placed in the placement tube, an oil gap is formed between the positive electrode and the negative electrode.

9. A method for identifying discharge type of a real transformer partial discharge simulation system according to any one of claims 1 to 8, characterized in that: include: Obtaining the original data of partial discharge ultrasonic signals and ultra-high frequency signals and storing them; Extract the original data and use feature extraction to obtain the equivalent bandwidth of the ultrasonic signal and the length of the UHF signal; The extracted equivalent bandwidth of the ultrasonic signal and the length of the ultra-high frequency signal are sent to the discharge type training module based on the convolutional neural network for accumulation training. When the discharge type training module accumulates a preset number of samples, the sample collection is stopped, the discharge type classification and recognition are performed, and the recognition result is given.

10. The method for identifying the discharge type of a real transformer partial discharge simulation system according to claim 9, characterized in that: The data extraction includes equivalent bandwidth extraction and signal length extraction; in the equivalent bandwidth extraction process, the spectrum range from the peak frequency point to its attenuation of 3dB is extracted from the ultrasonic signal spectrum as the equivalent bandwidth of the ultrasonic data; at the same time, when extracting the signal length, the time length of 15% of the ultra-high frequency original data from the beginning to the signal peak is extracted as the signal length of the ultra-high frequency waveform data.

Images