An adaptive transformer fault detection device

Abstract

The present invention discloses an adaptive transformer fault detection device, which belongs to the field of transformer fault detection, is adapted to a multi-frequency sensor installed on the transformer side, and includes: a signal acquisition module, a signal processing module, and a signal diagnosis module; the signal acquisition module is electrically connected to the signal processing module, and is used to collect the broadband signal of the multi-frequency sensor installed on the transformer side; the signal processing module is electrically connected to the signal diagnosis module, and performs time-frequency calibration on the collected broadband signal and adaptively adjusts the frequency range and storage time according to the signal characteristics of the broadband signal to obtain a low-frequency diagnosis signal and a high-frequency diagnosis signal; the signal diagnosis module is electrically connected to the decision module, and is used to identify the fault mode on the corresponding frequency band according to the multi-dimensional characteristics of the low-frequency diagnosis signal and the high-frequency diagnosis signal. By performing fault mode analysis on the multi-dimensional signal characteristics of the broadband signal, comprehensive, synchronous, accurate and rapid detection of transformer faults can be achieved.

Description

Technical Field

[0001] The present invention relates to the field of transformer fault detection, and in particular to an adaptive transformer fault detection device. Background Art

[0002] In the power system, transformers are key equipment, and their safe and stable operation is of vital importance. However, during the operation of transformers, due to various factors (such as aging, overload, environmental factors, etc.), various faults may occur, such as mechanical defects (such as looseness, wear, cracks, etc.) and insulation defects (such as partial discharge, breakdown, etc.). If these faults are not discovered and handled in time, they may cause transformer performance degradation and even cause serious safety accidents.

[0003] Traditional transformer fault detection methods usually rely on a single signal acquisition and analysis method, such as only collecting low-frequency vibration signals or high-frequency ultrasonic signals for fault diagnosis. However, this method has many limitations. First, a single signal may not fully reflect the fault characteristics of the transformer, resulting in missed diagnosis or misdiagnosis. Second, due to the complexity and diversity of transformer fault signals, traditional signal processing methods may not be able to accurately extract fault characteristics, affecting the accuracy of diagnosis.

[0004] In addition, traditional methods usually require professional technicians to operate and analyze, and require high skills and experience. Secondly, although some broadband mechanical wave sensors have appeared on the market, which can solve the problem of measuring low-frequency vibration signals and high-frequency ultrasonic signals at the same location to a certain extent, the signal processing devices of these sensors are often non-integrated, which means that the processing of low-frequency vibration signals and high-frequency ultrasonic signals needs to be performed in different devices or modules, and synchronization and unified data processing cannot be achieved; not only does it increase the complexity and cost of the system, but it may also reduce the reliability of the measurement results due to inconsistency in data processing.

[0005] The above information disclosed in the Background section is only for enhancement of understanding of the background of the present application and therefore it may contain information that does not constitute the prior art that is already known to a person of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to address the problem in the prior art that transformer state detection results are unreliable due to the asynchrony in transformer state signal analysis. An adaptive transformer fault detection device is proposed, which obtains real-time data of the transformer through a multi-frequency sensor and performs multi-frequency analysis, corrects the time domain and frequency domain of the broadband signal to ensure its real-time synchronization, and performs fault mode analysis on the multi-dimensional signal characteristics of the broadband signal, thereby achieving comprehensive, synchronous, accurate and rapid detection of transformer faults.

[0007] In a first aspect, a technical solution provided in an embodiment of the present invention is an adaptive transformer fault detection device, adapted to a multi-frequency sensor installed on a transformer side, comprising: a signal acquisition module, a signal processing module and a signal diagnosis module;

[0008] The signal acquisition module is electrically connected to the signal processing module and is used to acquire broadband signals from a multi-frequency sensor installed on the transformer side;

[0009] The signal processing module is electrically connected to the signal diagnosis module, performs time-frequency calibration on the collected broadband signal and adaptively adjusts the frequency range and storage time according to the signal characteristics of the broadband signal to obtain a low-frequency diagnosis signal and a high-frequency diagnosis signal;

[0010] The signal diagnosis module is electrically connected to the decision module, and is used to identify the fault mode on the corresponding frequency band according to the multi-dimensional characteristics of the low-frequency diagnosis signal and the high-frequency diagnosis signal.

[0011] In this solution, the signal acquisition module is connected to the multi-frequency sensor installed on the side of the transformer, and is responsible for capturing the broadband signal during the operation of the transformer in real time and with high precision. The broadband signal covers a wide range from low frequency to high frequency, and can fully reflect the operating status of the transformer, including changes in various physical quantities such as vibration and sound; the signal processing module receives the original data from the signal acquisition module, and also performs complex time-frequency calibration processing to ensure the accuracy and reliability of the signal. At the same time, it can dynamically adjust the frequency monitoring range and data storage strategy according to the characteristics of the received broadband signal, effectively respond to the signal changes of the transformer under different working conditions, and improve the sensitivity and accuracy of fault detection; the signal diagnosis module performs in-depth analysis of the processed signal, extracts multi-dimensional features, such as frequency, amplitude, phase, energy distribution, etc., and then identifies the fault mode on the corresponding frequency band. Through multi-dimensional analysis, the accuracy of fault detection and early warning capabilities can be significantly improved; this application combines dynamic feature analysis with adaptive learning technology to provide strong technical support for the safe and efficient operation of transformers.

[0012] Preferably, the signal acquisition module comprises a low-frequency signal acquisition unit that responds to mechanical vibration signals and a high-frequency signal acquisition unit that responds to ultrasonic signals.

[0013] Preferably, the signal processing module includes a time-frequency calibration unit, a frequency adjustment unit and a timing storage unit;

[0014] The time-frequency calibration unit adjusts the sampling moments of the low-frequency signal acquisition unit and the high-frequency signal acquisition unit according to the time-domain adjustment factor and the frequency-domain adjustment factor to obtain a time-frequency domain synchronization signal;

[0015] The frequency adjustment unit adaptively adjusts the corresponding frequency signal in response to the time-frequency domain synchronization signal to obtain a target state signal;

[0016] The timing storage unit stores the target state signal into a corresponding storage bit according to the synchronization mark of the time-frequency domain synchronization signal;

[0017] The target state signal includes a low frequency diagnosis signal and a high frequency diagnosis signal.

[0018] Preferably, the frequency adjustment unit includes a high-frequency adjustment unit and a low-frequency adjustment unit; the high-frequency adjustment unit responds to the ultrasonic signal in the time-frequency domain synchronization signal; the low-frequency adjustment unit responds to the mechanical vibration signal in the time-frequency domain synchronization signal.

[0019] Preferably, the frequency adjustment unit further includes a signal amplification unit, and the signal amplification unit is constructed by a T-type resistor feedback network charge amplifier circuit.

[0020] Preferably, the frequency adjustment unit further includes an envelope detection unit for performing upper envelope detection on the ultrasonic signal.

[0021] Preferably, the signal diagnosis module includes a frequency domain feature extraction unit, a time domain feature extraction unit, a time-frequency domain feature extraction unit and an envelope analysis unit; the signal diagnosis module determines the corresponding unit combination mode in response to the frequency range of the low-frequency diagnostic signal and the high-frequency diagnostic signal to generate the corresponding fault mode.

[0022] Preferably, the signal diagnosis module determines the corresponding unit combination mode in response to the frequency range of the low-frequency diagnosis signal and the high-frequency diagnosis signal to generate the corresponding fault mode, comprising the following steps:

[0023] Extracting a low-frequency diagnostic signal from the target state signal, and the time domain feature extraction unit determines a first low-frequency fault mode according to a mean, a variance, a peak value, and a kurtosis in the low-frequency diagnostic signal;

[0024] The frequency domain feature extraction unit analyzes the spectrum energy, center frequency, bandwidth and harmonic components in the low-frequency diagnostic signal according to Fourier transform to determine the second low-frequency fault mode;

[0025] A target low-frequency failure mode is determined according to the first low-frequency failure mode and the second low-frequency failure mode.

[0026] Preferably, the signal diagnosis module determines the corresponding unit combination mode in response to the frequency range of the low-frequency diagnosis signal and the high-frequency diagnosis signal to generate the corresponding fault mode, and further comprises the following steps:

[0027] Extracting a high-frequency diagnostic signal from the target state signal, the envelope analysis unit extracts the envelope characteristics of the high-frequency diagnostic signal through Hilbert transform and demodulation analysis to obtain modulation information and transient impact information to determine the first high-frequency fault mode;

[0028] The frequency domain feature extraction unit performs spectrum analysis on the high frequency diagnostic signal and extracts high frequency components and harmonics to identify the high frequency vibration source and determine the second high frequency fault mode;

[0029] The time-frequency domain feature extraction unit obtains the pulse amplitude and voltage phase of the high-frequency diagnostic signal on a set time scale to obtain a third high-frequency fault mode;

[0030] A target high-frequency fault mode is acquired according to the first high-frequency fault mode, the second high-frequency fault mode, and the third high-frequency fault mode.

[0031] Preferably, the present application further comprises a decision module that responds to the failure mode and displays the failure through a visualization tool;

[0032] The decision module includes a display unit and an alarm unit, and the display unit displays the target low-frequency fault mode and / or the target high-frequency fault mode through a vibration frequency domain diagram and / or a partial discharge PRPD spectrum diagram;

[0033] The alarm unit sends the fault mode of the target transformer to the remote platform and / or the mobile terminal.

[0034] The present invention has at least the following substantial effects:

[0035] (1) In order to solve the problem of unreliable detection results caused by asynchronous analysis of transformer status signals, the present application uses a signal acquisition module to capture the broadband signal of the transformer in real time, and uses a time-frequency calibration unit in a signal processing module to accurately adjust the low-frequency signal and the high-frequency signal in the time domain and frequency domain to ensure the synchronization of the two in the time and frequency domain, which not only improves the accuracy of the signal, but also ensures the reliability of subsequent fault analysis, thereby effectively solving the problem of unreliable detection results caused by asynchronous signals;

[0036] (2) In order to solve the problem of incomplete and inaccurate transformer fault detection, the present application proposes an adaptive transformer fault detection device, which not only collects broadband signals of the transformer, but also performs in-depth multi-dimensional feature analysis of these signals through a signal diagnosis module. The multi-dimensional features include time domain features, frequency domain features, time-frequency domain features, and envelope features, which can fully reflect the operating status of the transformer and achieve comprehensive and accurate detection of transformer faults.

[0037] (3) In order to solve the problem of slow response speed of transformer fault detection, the present invention integrates functions such as signal acquisition, processing, diagnosis and decision-making into one device, realizing efficient signal processing and fault diagnosis process. In particular, the design of the decision-making module can quickly respond to fault modes, display fault information through visualization tools, and send fault alarms to remote platforms or mobile terminals. This integrated design and rapid response mechanism significantly improves the efficiency and timeliness of transformer fault detection, helps to take maintenance measures in a timely manner, and ensures the safe operation of the transformer.

[0038] The above invention content is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Other features, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings. The drawings are only for the purpose of illustrating preferred embodiments and are not to be considered as limiting the present invention. Also, the same reference symbols are used throughout the drawings to represent the same parts.

[0040] Figure 1 It is a structural schematic diagram of an adaptive transformer fault detection device of the present invention.

[0041] Figure 2 The present invention is a schematic diagram of a detection circuit of an adaptive transformer fault detection device.

[0042] Figure 3 This is a detection effect diagram of an adaptive transformer fault detection device of the present invention.

[0043] Figure 4 It is the vibration frequency domain diagram of the present invention.

[0044] Figure 5 It is the PRPD spectrum under partial discharge fault of the present invention. DETAILED DESCRIPTION

[0045] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific implementation method described herein is only an optimal embodiment of the present invention, which is only used to explain the present invention and does not limit the scope of protection of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0046] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the operations (or steps) as sequential processes, many of the operations (or steps) therein can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the drawings; the process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0047] Embodiment 1: Figure 1 As shown, an adaptive transformer fault detection device is adapted to a multi-frequency sensor installed on the transformer side. Specifically, the multi-frequency sensor is installed on the transformer box wall and is used to send a broadband signal (specifically a vibration signal and an ultrasonic signal) representing the transformer state to the transformer fault detection device, wherein the transformer fault detection device is composed of a signal acquisition module, a signal processing module and a signal diagnosis module. Specifically, the signal acquisition module is electrically connected to the signal processing module, and is used to acquire the broadband signal of the multi-frequency sensor installed on the transformer side; the signal processing module is electrically connected to the signal diagnosis module, and performs time-frequency calibration on the acquired broadband signal and adaptively adjusts the frequency range and storage time according to the signal characteristics of the broadband signal; the signal diagnosis module is electrically connected to the decision module, and is used to identify the fault mode on the corresponding frequency band according to the multi-dimensional characteristics of the broadband signal.

[0048] In this embodiment, the signal acquisition module is connected to the multi-frequency sensor installed on the side of the transformer, and is responsible for capturing the broadband signal during the operation of the transformer in real time and with high precision, wherein the broadband signal covers a wide range from low frequency to high frequency, and can comprehensively reflect the operating status of the transformer, including changes in various physical quantities such as vibration and sound; the signal processing module receives the raw data from the signal acquisition module, and also performs complex time-frequency calibration processing to ensure the accuracy and reliability of the signal. At the same time, it can dynamically adjust the frequency monitoring range and data storage strategy according to the characteristics of the received broadband signal, effectively respond to the signal changes of the transformer under different working conditions, and improve the sensitivity and accuracy of fault detection; the signal diagnosis module performs in-depth analysis of the processed signal, extracts multi-dimensional features, such as frequency, amplitude, phase, energy distribution, etc., and then identifies the fault mode on the corresponding frequency band. Through multi-dimensional analysis, the accuracy of fault detection and early warning capabilities can be significantly improved; the present application combines dynamic feature analysis with adaptive learning technology to provide strong technical support for the safe and efficient operation of the transformer.

[0049] It can be understood that, combined with the actual working conditions of the transformer, the collection range and accuracy specifications of the transformer status parameters are as follows: a. Vibration signal measurement: Frequency range: The sensor should cover a frequency range of at least 10-1000Hz to capture common low-frequency mechanical vibrations in transformer operation. Sensitivity: The sensitivity of the sensor should be greater than 50mV / g to ensure that tiny vibration signals can be detected, thereby improving the accuracy of fault detection. b. Ultrasonic signal measurement: Frequency range: The sensor should cover a frequency range of 100k-200kHz, which is the main ultrasonic signal frequency band generated by partial discharge (PD) in oil, and can effectively reflect the existence and severity of insulation defects. Sensitivity: The sensitivity of the ultrasonic signal should be greater than 80dB to ensure sufficient detection capability for weak partial discharge signals, thereby improving the sensitivity and reliability of fault identification.

[0050] Combined with the sampling standards of the above transformer state parameters, the multi-frequency sensor used in this implementation can be, for example, the HS-10A-11M2 ultrasonic emission vibration composite sensor from Fuji of Japan, which can measure vibration and ultrasonic signals at the same time. The constant current power supply provides a stable current supply for the ultrasonic emission vibration composite sensor to ensure the stable operation of the sensor; the ultrasonic sensor has a sensitivity of about 80dB and a frequency range of 50~200kHz, which includes the main frequency band of partial discharge in oil; the vibration signal sensor is an IEPE sensor, which requires a constant current power supply, a sensitivity of about 100mV / g, and a frequency range of 10Hz~20kHz, which meets the needs of vibration measurement.

[0051] As a preferred solution of this embodiment, the signal acquisition module includes a low-frequency signal acquisition unit that responds to the mechanical vibration signal and a high-frequency signal acquisition unit that responds to the ultrasonic signal.

[0052] In this embodiment, mechanical vibration is an important reflection of the internal mechanical structure state of the transformer. For example, abnormal conditions such as loose windings and deformation of the core may cause specific vibration modes. Assuming that in an actual scenario, the transformer operates at a rated frequency of 50Hz, the low-frequency signal acquisition unit can accurately capture low-frequency signals generated by mechanical vibration with a frequency range of tens of Hz to hundreds of Hz. These signals are converted into electrical signals by high-precision sensors and transmitted to the low-frequency signal acquisition unit in real time for subsequent analysis. The high-frequency signal acquisition unit mainly responds to ultrasonic signals that may be generated inside the transformer. Ultrasonic signals are usually closely related to the insulation state of the transformer, partial discharge and other phenomena. For example, when partial discharge occurs inside the transformer, high-frequency ultrasonic signals are generated. The high-frequency signal acquisition unit can detect these ultrasonic signals with frequencies usually between several thousand Hz and hundreds of kHz. In actual applications, if partial discharge occurs somewhere inside the transformer, the high-frequency signal acquisition unit will quickly capture this abnormal signal and provide key information for subsequent fault diagnosis.

[0053] As a preferred solution of this embodiment, the signal processing module includes a time-frequency calibration unit, a frequency adjustment unit and a timing storage unit;

[0054] The time-frequency calibration unit adjusts the sampling moments of the low-frequency signal acquisition unit and the high-frequency signal acquisition unit according to the time-domain adjustment factor and the frequency-domain adjustment factor to obtain a time-frequency domain synchronization signal;

[0055] The frequency adjustment unit adaptively adjusts the corresponding frequency signal in response to the time-frequency domain synchronization signal to obtain a target state signal;

[0056] The timing storage unit stores the target state signal into a corresponding storage bit according to the synchronization mark of the time-frequency domain synchronization signal;

[0057] The target state signal includes a low-frequency diagnosis signal and a high-frequency diagnosis signal.

[0058] In this embodiment, in order to optimize the efficiency and accuracy of signal processing, the sampling moments of the low-frequency signal acquisition unit and the high-frequency signal acquisition unit are finely adjusted according to the preset time domain adjustment factor and frequency domain adjustment factor. Among them, the time domain adjustment factor focuses on time synchronization to ensure that signals of different frequencies are consistent in time; while the frequency domain adjustment factor is responsible for adjusting the frequency characteristics of the signal to eliminate the frequency deviation that may be caused by differences in acquisition equipment or environmental factors. The time-frequency calibration unit can generate a time-frequency domain synchronization signal so that the low-frequency and high-frequency signals are highly synchronized in time and frequency, laying a solid foundation for subsequent processing. After receiving the time-frequency domain synchronization signal, the frequency adjustment unit adaptively adjusts the frequency of these synchronized signals based on specific characteristics of the signal, such as amplitude, phase or spectrum characteristics, to meet specific analysis or diagnosis requirements, so that the original signal is converted into a target state signal, and the target state signal includes a low-frequency diagnostic signal and a high-frequency diagnostic signal, which correspond to different frequency ranges and analysis focuses respectively. Finally, the timing storage unit is responsible for storing the processed target state signal according to the synchronization mark in the time-frequency domain synchronization signal. The key to ensuring that the signal maintains the correct time sequence when stored is through synchronization marks, allowing subsequent data analysis or playback to accurately reconstruct the time series of the original signal. By storing the low-frequency diagnostic signal and the high-frequency diagnostic signal in corresponding storage locations, the timing storage unit not only ensures the organization of the data, but also facilitates subsequent data retrieval and processing. The signal processing module in this embodiment achieves efficient and accurate processing of low-frequency and high-frequency signals through the close cooperation of the three links of time-frequency calibration, frequency adjustment, and timing storage, which not only improves the accuracy and reliability of signal processing, but also provides a high-quality data foundation for subsequent data analysis and fault diagnosis.

[0059] As a further explanation of the above embodiment; for the low-frequency signal acquisition unit, its original sampling time is , the sampling time after adjustment by the time-frequency calibration unit can be expressed as: ;in, is the time domain adjustment factor of the low-frequency signal. For the high-frequency signal acquisition unit, its original sampling time is , the sampling time after adjustment by the time-frequency calibration unit can be expressed as: ;in, It is the time domain adjustment factor of the high-frequency signal. The frequency adjustment unit receives the time-frequency domain synchronization signal and adaptively adjusts the signal of the corresponding frequency according to the signal to obtain the target state signal (including the low-frequency diagnostic signal and the high-frequency diagnostic signal). The timing storage unit accurately stores the target state signal to the corresponding storage location according to the synchronization mark in the time-frequency domain synchronization signal.

[0060] As a preferred solution of this embodiment, the frequency adjustment unit includes a high-frequency adjustment unit and a low-frequency adjustment unit; the high-frequency adjustment unit responds to the ultrasonic signal in the time-frequency domain synchronization signal; the low-frequency adjustment unit responds to the mechanical vibration signal in the time-frequency domain synchronization signal. Further, the frequency adjustment unit also includes a signal amplification unit, specifically, a first signal amplification unit is configured for the output end of the low-frequency adjustment unit, and a second signal amplification unit is configured for the output end of the high-frequency adjustment unit, and the first signal amplification unit and the second signal amplification unit are both constructed using a T-type resistor feedback network charge amplifier circuit.

[0061] In this embodiment, the vibration signal and ultrasonic signal obtained by the signal processing module have a lower signal-to-noise ratio and stronger reliability. Since the signal amount is small, they are connected to a charge amplifier, and the charge signal enters the amplification circuit through the input end of the charge amplifier; since the lower limit cutoff frequency of the broadband vibration signal is low, in order to avoid signal loss after amplification, the signal amplification unit in the present invention introduces a T-type resistor feedback network charge conversion scheme; The T-type resistor feedback network charge amplifier circuit belongs to the conventional technology in this field and will not be elaborated here.

[0062] As a preferred solution of this embodiment, the frequency adjustment unit also includes an envelope detection unit for performing upper envelope detection on the ultrasonic signal.

[0063] In this embodiment, during the ultrasonic signal measurement process, although direct measurement can obtain rich signal information, in the partial discharge analysis, only the amplitude and time of the ultrasonic signal generated by the partial discharge need to be extracted. In order to simplify the signal processing, the present invention uses an envelope detection circuit to perform upper envelope detection on the ultrasonic signal to ensure that the peak value and corresponding time of the ultrasonic signal remain unchanged, and at the same time effectively reduce the sampling rate requirement to about 10 kHz, thereby reducing the power consumption and memory usage of the detection device, and realizing the accurate collection of the peak value and time of the ultrasonic signal under low sampling rate conditions. The schematic diagram of the detection circuit is shown in Figure 2 ; Specifically, the negative input terminals of the two comparison elements of the detection circuit are connected, and the positive input terminal of the first comparison element obtains an ultrasonic signal, and the frequency range of the ultrasonic signal is: 50KHz~200KHz; the negative input terminal of the first comparison element is electrically connected to the anode terminal of the diode D2, and the output terminal of the first comparison element is electrically connected to the cathode terminal of the diode D2. The output terminal of the first comparison element is electrically connected to the anode terminal of the diode D1 through the resistor R1, and the cathode terminal of the diode D1 is grounded through the capacitor C, the variable resistor R2 is grounded, and connected to the positive input terminal of the second comparison element; the output terminal of the second comparison element outputs an ultrasonic detection signal, and the effect diagram of the detection signal is shown in Figure 3 described.

[0064] As a preferred solution of this embodiment, the signal diagnosis module includes a frequency domain feature extraction unit, a time domain feature extraction unit, a time-frequency domain feature extraction unit and an envelope analysis unit; the signal diagnosis module determines the corresponding unit combination mode in response to the frequency range of the target state signal to generate the corresponding fault mode.

[0065] Further, the signal diagnosis module determines the corresponding unit combination mode in response to the frequency range of the target state signal to generate the corresponding fault mode, including the following steps:

[0066] Extracting a low-frequency signal from the target state signal, and the time domain feature extraction unit determines a first low-frequency fault mode according to the mean, variance, peak value, and kurtosis of the low-frequency signal;

[0067] The frequency domain feature extraction unit analyzes the spectrum energy, center frequency, bandwidth and harmonic components in the low frequency signal according to Fourier transform to determine the second low frequency fault mode;

[0068] A target low-frequency failure mode is determined according to the first low-frequency failure mode and the second low-frequency failure mode.

[0069] In this embodiment, the loose or displaced winding will cause periodic vibration or impact, especially under the action of electromagnetic force when the transformer is running, which is manifested as periodic pulses in the time domain signal. As the fault worsens, the amplitude of the pulse will increase, and the frequency of the pulse is usually related to the power frequency of the power grid; when the core is loose, low-frequency vibration will appear in the time domain signal. The loose core causes mechanical shock and electromagnetic instability, resulting in obvious vibration in the low frequency band. The time domain feature extraction unit is used to analyze the time domain characteristics of the vibration signal and extract statistical features such as the mean, variance, peak value, and kurtosis of the signal to identify the fluctuation trend and abnormal impact of the signal. In the frequency domain analysis, the main feature of the loose or displaced winding is that it shows significant vibration components at the power frequency and its multiples, especially harmonic signals such as 2 times and 3 times. As the fault worsens, the energy of the higher harmonics increases, and the high-frequency components in the spectrum are more obvious; the loose core or vibration fault usually shows obvious vibration peaks in the low frequency band (<100Hz), and the power frequency and multiple frequency components will also become stronger as the fault increases. The frequency domain feature extraction unit performs a fast Fourier transform (FFT) on the signal to extract frequency domain features such as spectrum energy, center frequency, bandwidth, and harmonic components, thereby identifying the fault mode at a specific frequency. Specifically, based on the calculated time domain features, combined with the time domain performance of faults such as winding looseness or displacement, loose core, etc., it is determined whether there is a corresponding low-frequency fault mode. For example, if periodic pulses appear in the signal, and the pulse amplitude increases as the fault intensifies, for example, the slope of the amplitude increase exceeds the set threshold, there may be a winding looseness or displacement fault; if low-frequency vibration appears in the signal and exceeds the set threshold, there may be a core looseness fault. Based on the extracted frequency domain features, combined with the frequency domain performance of faults such as winding looseness or displacement, loose core, etc., it is determined whether there is a corresponding low-frequency fault mode. For example, if the spectrum shows significant vibration components at the power frequency and its multiples, especially harmonic signals such as 2 times and 3 times, there may be a winding looseness or displacement fault; if the low frequency band (<100Hz) shows an obvious vibration peak, there may be a core looseness fault. According to the judgment results of the first low-frequency fault mode and the second low-frequency fault mode, the target low-frequency fault mode is determined by comprehensively considering the time domain and frequency domain characteristics. If the judgment results of the two fault modes are consistent, the target low-frequency fault mode is the fault; if the judgment results of the two fault modes are inconsistent, further analysis or confirmation with other diagnostic methods is required.

[0070] Furthermore, the signal diagnosis module determines the corresponding unit combination mode in response to the frequency range of the target state signal to generate the corresponding fault mode, and further includes the following steps:

[0071] Extracting a high-frequency signal from the target state signal, the envelope analysis unit extracts the envelope characteristics of the high-frequency signal through Hilbert transform and demodulation analysis to obtain modulation information and transient impact information to determine the first high-frequency fault mode;

[0072] The frequency domain feature extraction unit performs spectrum analysis on the high frequency signal and extracts high frequency components and harmonics to identify the high frequency vibration source and determine the second high frequency fault mode;

[0073] The time-frequency domain feature extraction unit obtains the pulse amplitude and voltage phase of the high-frequency signal on a set time scale to obtain a third high-frequency fault mode;

[0074] A target high-frequency fault mode is acquired according to the first high-frequency fault mode, the second high-frequency fault mode, and the third high-frequency fault mode.

[0075] In this embodiment, for ultrasonic signals, partial discharge faults are manifested as short-time spike pulse signals in the time domain, with large amplitude and unclear periodicity, often accompanied by randomness and non-periodicity. The duration of such spike pulses is very short, but the amplitude is very high, and it is manifested as a transient impact signal. The envelope analysis unit extracts the envelope characteristics of the signal through Hilbert transform and demodulation analysis to capture modulation information and transient impacts. In the frequency domain, partial discharge faults are manifested as significant enhancement of high-frequency band (>10kHz) signals. Since the discharge event lasts for a very short time, it will generate broadband noise signals in the spectrum. The high-frequency harmonics and broadband noise spectrum caused by the discharge usually appear in the high-frequency band. The frequency domain feature extraction unit performs spectrum analysis on the ultrasonic signal, extracts high-frequency components and harmonics, and identifies high-frequency vibration sources. The voltage phase analysis unit analyzes and integrates the amplitude and corresponding voltage phase of each ultrasonic signal pulse within a measurement time of the ultrasonic signal. Specifically, the amplitude of the high-frequency signal envelope extracted by Hilbert transform and demodulation analysis can reflect the intensity of the transient impact. In specific implementation, when the envelope amplitude exceeds the set threshold, it is considered that there is a significant transient impact; the modulation frequency extracted from the envelope signal can reflect the periodicity of the modulation information; if the modulation frequency is close to the power grid frequency or its multiples, it may indicate a specific fault mode. The width of the short-time spike pulse signal is also an important parameter for quantifying transient impacts. The narrower the pulse width, the stronger the impact. The amplitude of the high-frequency component extracted after the spectrum analysis of the high-frequency signal. A frequency range (such as >10kHz) can be set, and the total amplitude or average amplitude of all frequency components in the range can be calculated. The amplitude of the high-frequency harmonic component caused by the discharge can reflect the intensity and frequency distribution of the harmonics. The amplitude of each harmonic can be calculated separately, or the total harmonic distortion (THD) can be calculated. The spectrum width of the high-frequency signal can reflect the frequency range of the broadband noise signal. The wider the spectrum width, the more intense the discharge event. The amplitude of each high-frequency signal pulse on a set time scale. Statistics such as the average amplitude and maximum amplitude of all pulses can be calculated. The voltage phase corresponding to each high-frequency signal pulse amplitude. A scatter plot or histogram of the pulse amplitude and the voltage phase can be drawn to analyze the relationship and distribution pattern between them. If the relationship between the pulse amplitude and the voltage phase under normal conditions is known, calculate the phase difference in the current state, that is, the degree of deviation between the actual phase and the normal phase. The greater the phase difference, the more serious the fault is. Perform weighted scoring on the quantitative indicators of the first high-frequency fault mode, the second high-frequency fault mode, and the third high-frequency fault mode, and comprehensively judge the target high-frequency fault mode based on the scoring results.

[0076] As a further explanation of the above embodiments; exemplarily:

[0077] ;

[0078] in, , , are the weights of the first, second, and third high frequency failure modes, respectively, and The weight of the fault mode can be set according to the importance and frequency of the actual fault mode.

[0079] Among them, the first high frequency failure mode score:

[0080] ;

[0081] in, is the number of quantitative indicators of the first high-frequency fault mode, , , They are the weight coefficients of envelope amplitude, modulation frequency and pulse width respectively.

[0082] Second Highest Failure Mode Score:

[0083] ;

[0084] in, is the number of quantitative indicators of the second high-frequency failure mode, , , They are the weight coefficients of high-frequency component amplitude, harmonic component amplitude, and spectrum width respectively.

[0085] Third Highest Failure Mode Score:

[0086] ;

[0087] in, is the number of quantitative indicators of the third high-frequency fault mode, , are the weight coefficients of pulse amplitude and voltage phase difference, respectively. Note that it is assumed here that the voltage phase difference is already an example quantitative indicator. If calculation is required in practice, it can be quantified according to the specific situation of the phase difference.

[0088] In this embodiment, the target high-frequency fault mode is determined based on the calculated total score and a threshold or range. For example, if the total score exceeds a certain threshold, it is considered that there is a certain serious high-frequency fault; if the total score is within a certain range, it is considered that there is a certain degree of high-frequency fault; if the total score is lower than a certain threshold, it is considered that the high-frequency signal is normal or the fault is not obvious.

[0089] Embodiment 2:

[0090] The structure is the same as that of embodiment 1, except that the present application also includes a decision module that responds to the fault mode and displays the fault through a visualization tool; the decision module includes a display unit and an alarm unit, and the display unit displays the target low-frequency fault mode and / or the target high-frequency fault mode through a vibration frequency domain diagram and / or a partial discharge PRPD spectrum diagram; the alarm unit sends the fault mode of the target transformer to a remote platform and / or a mobile terminal.

[0091] In this embodiment, the decision module performs a visual analysis of the fault mode. The carrier of the display unit can be a remote platform or a mobile terminal. The mobile terminal can be a computer, a mobile phone, a pad, or other terminal with display and data storage functions, which is not limited by the present invention. The mobile terminal performs a visual spectrum analysis on the vibration signal, and performs a visual analysis of the partial discharge PRPD spectrum on the ultrasonic envelope signal and the voltage phase signal. With regard to vibration spectrum analysis, since the transformer vibration is usually a multiple and division of 50Hz and 100Hz, when there is a mechanical fault, the vibration shape of the collected vibration signal will change to a certain extent, indicating that the content of different frequency components in the frequency domain has changed to a certain extent. Therefore, by performing a spectrum analysis on the vibration signal under different fault conditions, such as Figure 4 As shown in the figure, observe the change pattern of each frequency component. If a certain component content changes significantly, it usually indicates that there is a mechanical fault inside the transformer. For partial discharge PRPD spectrum analysis, the discharge point in the PRPD spectrum of partial discharge is drawn by the amplitude of each ultrasonic signal pulse and the corresponding voltage phase at this time, as shown in the figure below. Figure 5 As shown in the figure, by analyzing the amplitude of each ultrasonic signal pulse and the corresponding voltage phase within a measurement period, the results are summarized as the partial discharge PRPD spectrum during this period. By analyzing the characteristics between the discharge point distribution and the voltage phase in the PRPD spectrum, the type of partial discharge occurring inside the transformer can be analyzed, such as: floating potential discharge, needle plate discharge, surface discharge, etc.

[0092] The specific implementation described above is a preferred implementation of an adaptive transformer fault detection device of the present invention, and is not intended to limit the specific implementation scope of the present invention. The scope of the present invention includes but is not limited to this specific implementation. All equivalent changes made in accordance with the shape and structure of the present invention are within the protection scope of the present invention.

Claims

1. An adaptive transformer fault detection device, adapted for a multi-frequency sensor installed on the transformer side, characterized in that: It includes: signal acquisition module, signal processing module and signal diagnosis module; The signal acquisition module is electrically connected to the signal processing module and is used to acquire broadband signals from a multi-frequency sensor installed on the transformer side; The signal processing module is electrically connected to the signal diagnosis module, performs time-frequency calibration on the collected broadband signal and adaptively adjusts the frequency range and storage time according to the signal characteristics of the broadband signal to obtain a low-frequency diagnosis signal and a high-frequency diagnosis signal; The signal diagnosis module is electrically connected to the decision module and is used to identify the fault mode on the corresponding frequency band according to the multi-dimensional characteristics of the low-frequency diagnosis signal and the high-frequency diagnosis signal; The signal diagnosis module includes a frequency domain feature extraction unit, a time domain feature extraction unit, a time-frequency domain feature extraction unit and an envelope analysis unit; the signal diagnosis module determines the corresponding unit combination mode in response to the frequency range of the low-frequency diagnosis signal and the high-frequency diagnosis signal to generate the corresponding fault mode; The signal diagnosis module determines a corresponding unit combination mode in response to the frequency range of the low-frequency diagnosis signal and the high-frequency diagnosis signal to generate a corresponding fault mode, including the following steps: Extracting a low-frequency diagnostic signal from the target state signal, and the time domain feature extraction unit determines a first low-frequency fault mode according to a mean, a variance, a peak value, and a kurtosis in the low-frequency diagnostic signal; The frequency domain feature extraction unit analyzes the spectrum energy, center frequency, bandwidth and harmonic components in the low-frequency diagnostic signal according to Fourier transform to determine the second low-frequency fault mode; determining a target low-frequency fault mode according to the first low-frequency fault mode and the second low-frequency fault mode; The signal diagnosis module determines the corresponding unit combination mode in response to the frequency range of the low-frequency diagnosis signal and the high-frequency diagnosis signal to generate the corresponding fault mode, and further includes the following steps: Extracting a high-frequency diagnostic signal from the target state signal, the envelope analysis unit extracts the envelope characteristics of the high-frequency diagnostic signal through Hilbert transform and demodulation analysis to obtain modulation information and transient impact information to determine the first high-frequency fault mode; The frequency domain feature extraction unit performs spectrum analysis on the high frequency diagnostic signal and extracts high frequency components and harmonics to identify the high frequency vibration source and determine the second high frequency fault mode; The time-frequency domain feature extraction unit obtains the pulse amplitude and voltage phase of the high-frequency diagnostic signal on a set time scale to obtain a third high-frequency fault mode; A target high-frequency fault mode is acquired according to the first high-frequency fault mode, the second high-frequency fault mode, and the third high-frequency fault mode.

2. The adaptive transformer fault detection device according to claim 1, characterized in that: The signal acquisition module includes a low-frequency signal acquisition unit that responds to mechanical vibration signals and a high-frequency signal acquisition unit that responds to ultrasonic signals.

3. The adaptive transformer fault detection device according to claim 1, characterized in that: The signal processing module includes a time-frequency calibration unit, a frequency adjustment unit and a timing storage unit; The time-frequency calibration unit adjusts the sampling moments of the low-frequency signal acquisition unit and the high-frequency signal acquisition unit according to the time-domain adjustment factor and the frequency-domain adjustment factor to obtain a time-frequency domain synchronization signal; The frequency adjustment unit adaptively adjusts the corresponding frequency signal in response to the time-frequency domain synchronization signal to obtain a target state signal; The timing storage unit stores the target state signal into a corresponding storage bit according to the synchronization mark of the time-frequency domain synchronization signal; The target state signal includes a low frequency diagnosis signal and a high frequency diagnosis signal.

4. The adaptive transformer fault detection device according to claim 3, characterized in that: The frequency adjustment unit includes a high frequency adjustment unit and a low frequency adjustment unit; the high frequency adjustment unit responds to the ultrasonic signal in the time-frequency domain synchronization signal; the low frequency adjustment unit responds to the mechanical vibration signal in the time-frequency domain synchronization signal.

5. The adaptive transformer fault detection device according to claim 4, characterized in that: The frequency adjustment unit also includes a signal amplification unit, and the signal amplification unit is constructed by using a T-type resistor feedback network charge amplification circuit.

6. The adaptive transformer fault detection device according to claim 4, characterized in that: The frequency adjustment unit further includes an envelope detection unit for performing upper envelope detection on the ultrasonic signal.

7. The adaptive transformer fault detection device according to claim 1, characterized in that: Also included is a decision module that responds to the failure mode and displays the failure through a visualization tool; The decision module includes a display unit and an alarm unit, and the display unit displays the target low-frequency fault mode and / or the target high-frequency fault mode through a vibration frequency domain diagram and / or a partial discharge PRPD spectrum diagram; The alarm unit sends the fault mode of the target transformer to the remote platform and / or the mobile terminal.

Images