Transformer fault detection device and method

Abstract

The invention provides a transformer fault detection device and method, and the method comprises the steps: constructing a fault standard atlas and a fault feature library, injecting a broadband excitation signal into a transformer, collecting time domain response signals of a primary winding and a secondary winding in the transformer, and carrying out the Fourier transform to obtain a frequency domain transmission function; obtaining an amplitude-frequency characteristic curve and a phase-frequency characteristic curve based on a frequency domain transmission function, obtaining resonance point offset, a nonlinearity index and a harmonic distortion degree based on the amplitude-frequency characteristic curve and the phase-frequency characteristic curve, combining the resonance point offset, the nonlinearity index and the harmonic distortion degree into a transformer detection vector, constructing a three-dimensional coordinate system, obtaining detection data points, constructing a three-dimensional health area, and obtaining the three-dimensional health area. If the detection data point is not in the three-dimensional healthy area, it is indicated that the transformer has a fault, real-time fault type judgment is carried out based on the transformer detection vector and the fault standard map, a fault feature value is obtained, the fault feature value is compared with a fault feature library, a fault diagnosis result is output, and the accuracy of fault judgment of the transformer is improved.

Description

Technical Field

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

[0002] In the context of the intelligent transformation of power systems, with the expansion of power grids and the increasing complexity of transformers, transformer condition monitoring and fault diagnosis face significant technical challenges. Traditional transformer fault diagnosis methods rely primarily on regular manual inspections or single-stage diagnosis. These technical systems have exposed significant technical bottlenecks in long-term practice.

[0003] On the one hand, the manually-led fault diagnosis model is experience-dependent and time-lag. The formation mechanism and characterization characteristics of insulation aging and partial discharge faults of transformers under different voltage levels, load conditions and environmental conditions are significantly different. The periodic inspection plan formulated by maintenance personnel based on historical experience is difficult to accurately capture the early fault signs in the dynamic evolution process of the transformer. There are latent faults in the transformer. The traditional method is easily affected by subjective judgment bias, resulting in an increased fault misjudgment rate and the possibility of missing the best time for maintenance intervention.

[0004] On the other hand, the single-stage judgment system has the problem of misjudgment risk. The evolution of transformer faults often presents the characteristics of multi-physical field coupling. For example, the interaction of thermal, electrical and mechanical stresses can lead to chain reactions such as winding deformation and insulation degradation. Single-stage judgment is difficult to fully capture the spatiotemporal correlation characteristics of the overall health status of the transformer through only one fault detection. When encountering complex interference factors such as overload operation and sudden changes in ambient temperature, the single-stage diagnostic model is prone to misjudging fluctuations in normal operating conditions.

[0005] Therefore, traditional transformer fault diagnosis technology exhibits obvious technical limitations when dealing with complex and changeable operating conditions and early transformer defects, and it is difficult to meet the smart grid's needs for accurate perception and staged diagnosis of transformer status. Summary of the Invention

[0006] In view of the above-mentioned problems, in combination with the first aspect of the present invention, an embodiment of the present invention provides a transformer fault detection method, the method comprising: Constructing a fault standard map based on existing transformer fault types and a fault feature library based on the degree of transformer faults; Inject a broadband excitation signal into the transformer, collect the time domain response signals of the primary and secondary windings in the transformer based on the data acquisition module, perform Fourier transform on the time domain response signals, and obtain the frequency domain transfer function; Based on the frequency domain transfer function, the amplitude-frequency characteristic curve and the phase-frequency characteristic curve are obtained, the resonance point offset in the amplitude-frequency characteristic curve is extracted, the nonlinearity index and harmonic distortion in the phase-frequency characteristic curve are simultaneously extracted, and the resonance point offset, nonlinearity index and harmonic distortion are combined into a transformer detection vector; Establishing a three-dimensional coordinate system based on the transformer detection vector, obtaining detection data points within the three-dimensional coordinate system, obtaining the transformer detection vector in a healthy state, constructing a three-dimensional healthy region based on the transformer detection vector in a healthy state, and performing health detection based on the detection data points and the three-dimensional healthy region; If the detection data point is not within the three-dimensional healthy area, it means that the transformer is faulty. The real-time fault type is determined based on the transformer detection vector and the fault standard map. The weight is then matched based on the result of the real-time fault type determination. The fault feature value is obtained and compared with the fault feature library to output the fault diagnosis result. If the detection data point is within the three-dimensional healthy area, it means that there is no fault in the transformer, and the process ends.

[0007] In another aspect, an embodiment of the present invention further provides a transformer fault detection device, comprising: A construction module, wherein the construction module is used to construct a fault standard map and a fault feature library, and to construct a three-dimensional coordinate system of a transformer detection vector; A data acquisition module, wherein the data acquisition module is used to acquire time domain response signals of the primary and secondary windings in the transformer, and perform Fourier transform on the time domain response signals to obtain a frequency domain transfer function; An acquisition module, the acquisition module is used to acquire an amplitude-frequency characteristic curve and a phase-frequency characteristic curve, acquire a resonance point offset in the amplitude-frequency characteristic curve, and acquire a nonlinearity index and harmonic distortion in the phase-frequency characteristic curve; A judgment module, which is used to perform real-time health detection to determine whether the transformer is faulty; The diagnostic module can perform real-time fault type judgment based on the transformer detection vector and the fault standard map, perform weight matching according to the result of the real-time fault type judgment, obtain the fault characteristic value, compare the fault characteristic value with the fault characteristic library, and output the fault diagnosis result.

[0008] Based on the above aspects, the embodiment of the present application performs Fourier transform on the input voltage and input current of the primary winding and the output voltage and output current of the secondary winding to obtain a frequency domain transfer function, obtains detection data points based on the frequency domain transfer function, judges whether there is a fault through the constructed three-dimensional coordinate system, and judges the fault type based on the constructed standard fault map. Finally, the fault diagnosis result is output according to the constructed fault feature library, and the transformer state is accurately perceived through a multi-level judgment method. At the same time, the fault diagnosis result is output including the level of the fault type, which can accurately capture the early faults in the dynamic evolution process of the transformer, and also analyze the latent faults in the transformer to achieve the purpose of advance fault warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 1 is a schematic diagram of an execution flow of a transformer fault detection method provided by an embodiment of the present invention; Figure 2 This is a schematic diagram of an execution flow of performing graded judgment on transformer faults in a transformer fault detection method provided by an embodiment of the present invention; Figure 3 Schematic diagram of a transformer fault detection device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0010] The present invention will be described in detail below with reference to the accompanying drawings. Figure 1 FIG1 is a schematic diagram of an execution flow of a transformer fault detection method provided by an embodiment of the present invention. Figure 2 This is a schematic diagram of an execution flow of performing hierarchical judgment on transformer faults in a transformer fault detection method provided by an embodiment of the present invention. The transformer fault detection method is introduced in detail below.

[0011] Step S1, constructing a fault standard map, wherein the fault standard map is constructed based on existing transformer fault types, and simultaneously constructing a fault feature library, wherein the fault feature library is constructed based on the fault degree of the transformer.

[0012] In this embodiment, step S1 includes: Step S11, based on the existing transformer fault type, the resonance point offset, the nonlinearity index, and the harmonic distortion are divided into data areas, the transformer detection vector range of different existing transformer fault types is obtained, and a fault standard map is constructed based on the existing transformer fault type and the transformer detection vector range.

[0013] Specifically, existing transformer fault types include winding short circuit, insulation moisture, and core aging. Under normal operating conditions, the parameter of the resonance point offset is less than 1%, the parameter of the nonlinearity index is less than 0.3, and the parameter of the harmonic distortion is less than 40%. If the transformer has a winding short circuit fault, the parameter range of the transformer resonance point offset is 1%-5%. If the transformer has a moisture insulation fault, the parameter range of the transformer nonlinearity index is 0.3-0.8. If the transformer has an aging core fault, the parameter range of the transformer harmonic distortion is 40%-80%. A fault standard map is constructed based on the existing transformer fault types and the transformer detection vector range.

[0014] Step S12: classify the fault into three fault levels: minor fault, common fault, and severe fault, and divide the data area for each fault level to obtain the transformer fault level range of each fault, and build a fault feature library based on the real-time fault type, transformer fault level, and fault level range.

[0015] Specifically, winding short circuit is divided into three fault levels: slight short circuit, normal short circuit and severe short circuit, and each fault level is divided into data areas. When the transformer has a slight short circuit, the resonance point offset is 1%-2%, when the transformer has a normal short circuit, the resonance point offset is 2%-3.5%, and when the transformer has a severe short circuit, the resonance point offset is 3.5%-5%.

[0016] Similarly, insulation moisture is divided into three fault levels: slight moisture, normal moisture, and severe moisture, and the data area is divided for each fault level. When the transformer is slightly damp, the nonlinearity index is 0.3-0.4, when the transformer is normally damp, the nonlinearity index is 0.4-0.6, and when the transformer is severely damp, the nonlinearity index is 0.6-0.8.

[0017] Furthermore, core aging is divided into three fault levels: mild aging, normal aging, and severe aging. Data areas are divided for each fault level. When the transformer is slightly aged, the harmonic distortion is 40%-50%, when the transformer is normally aged, the harmonic distortion is 50%-65%, and when the transformer is severely aged, the harmonic distortion is 65%-80%.

[0018] Step S2: injecting a broadband excitation signal into the transformer, collecting time domain response signals of the primary and secondary windings in the transformer based on a data acquisition module, performing Fourier transform on the time domain response signals, and obtaining a frequency domain transfer function.

[0019] In this embodiment, step S2 includes: In step S21 , the low-frequency, medium-frequency, and high-frequency electromagnetic behaviors of the transformer are simultaneously excited based on the broadband excitation signal, covering the entire operating frequency band of the transformer.

[0020] In this embodiment, a wide-band excitation signal is used to cover the entire operating frequency band of the transformer. The wide-band excitation signal uses a swept-frequency signal. The swept-frequency signal can simultaneously excite low-frequency, medium-frequency, and high-frequency electromagnetic behaviors, thereby achieving the purpose of comprehensively capturing multiple types of fault characteristics. Among them, the low-frequency band can reflect the magnetization characteristics of the core and the overall state of the main insulation, the medium-frequency band can reveal winding deformation and inter-turn insulation degradation, and the high-frequency band can capture partial discharge and poor lead contact. In a single test, the three core components of the magnetic core, winding, and insulation can be covered simultaneously, which improves the detection range compared to traditional single-frequency detection.

[0021] In step S22 , the time domain response signal includes the input voltage and input current of the primary winding and the output voltage and output current of the secondary winding.

[0022] Specifically, the input voltage and input current of the primary winding and the output voltage and output current of the secondary winding can be collected by, for example, a voltage probe and a current sensor, and the collected voltage and current data can be converted into a discrete sequence.

[0023] Step S23 , performing Fourier transform on the input voltage and input current of the primary winding and the output voltage and output current of the secondary winding to obtain the frequency domain input voltage and current of the primary winding and the frequency domain output voltage and current of the secondary winding after Fourier transform.

[0024] Specifically, the discrete sequence of voltage or current is Fourier transformed: ; in, represents a discrete sequence of voltages or currents, Represented as a discrete frequency domain sequence after Fourier transform, is the length of the Fourier transform, Represents the frequency component of [0, N-1].

[0025] Step S24, calculating the frequency domain voltage transfer ratio based on the frequency domain input voltage and the frequency domain output voltage, calculating the frequency domain current transfer ratio based on the frequency domain input current and the frequency domain output current, and obtaining the frequency domain transfer function according to the frequency domain voltage transfer ratio and the frequency domain current transfer ratio.

[0026] Specifically, the ratio of the frequency domain input voltage to the frequency domain output voltage is the frequency domain voltage transfer ratio, and the ratio of the frequency domain input current to the frequency domain output current is the frequency domain current transfer ratio. The frequency domain voltage transfer ratio and the frequency domain current transfer ratio are used to form a frequency domain transfer function. The frequency domain transfer function provides the frequency component information of the signal, which can better identify and analyze the specific frequencies and fault characteristics in the signal.

[0027] Step S3, obtaining the amplitude-frequency characteristic curve and the phase-frequency characteristic curve based on the frequency domain transfer function, extracting the resonance point offset in the amplitude-frequency characteristic curve, and synchronously extracting the nonlinearity index and harmonic distortion in the phase-frequency characteristic curve, and combining the resonance point offset, nonlinearity index, and harmonic distortion into a transformer detection vector.

[0028] In this embodiment, step S3 includes: Step S31 , obtaining frequency, amplitude and phase according to the frequency domain transfer function, drawing an amplitude-frequency characteristic curve based on the amplitude and frequency, and drawing a phase-frequency characteristic curve based on the phase and frequency.

[0029] Specifically, the frequency is obtained according to the sampling point and the sampling time, the amplitude of the frequency domain transfer function is obtained by calculating the modulus of the frequency domain transfer function, and the amplitude-frequency characteristic curve is drawn with the frequency and amplitude as the X and Y axes respectively. The phase of the frequency domain transfer function is obtained according to the inverse tangent value of the real number and the imaginary number of the frequency domain transfer function, and the phase-frequency characteristic curve is drawn with the frequency and phase as the X and Y axes respectively.

[0030] Step S32 , obtaining the resonant frequency of the amplitude-frequency characteristic curve, comparing the resonant frequency with a preset standard value, and calculating the offset of the resonant point.

[0031] Specifically, the derivative of the amplitude-frequency characteristic curve function is taken, and the first-order derivative is equal to 0 and the second-order derivative is less than 0. This point is obtained as the maximum value of the amplitude-frequency characteristic curve, that is, the resonant frequency of the amplitude-frequency characteristic curve, and a standard value of the resonant frequency is preset. The preset standard value represents the resonant frequency of the transformer in a healthy state. The resonant frequency of the real-time detected amplitude-frequency characteristic curve is subtracted from the resonant frequency of the amplitude-frequency characteristic curve in a healthy state of the transformer, and the difference is the resonance point offset.

[0032] Step S33 : converting the phase-frequency characteristic curve into a second-order polynomial, wherein the quadratic term coefficient of the second-order polynomial is a nonlinearity index.

[0033] Specifically, three data points are extracted from the phase-frequency characteristic curve, and the phase-frequency characteristic curve is converted into a second-order polynomial according to the extracted three data points. The quadratic term coefficient of the second-order polynomial is extracted, and the quadratic term coefficient is used as a nonlinearity index. For example, the phase-frequency characteristic curve is converted into a second-order polynomial as follows: , extract the quadratic coefficients of the second-order polynomial , the quadratic coefficient as an indicator of nonlinearity.

[0034] Step S34: perform harmonic analysis on the time domain response signal of the secondary winding after Fourier transformation to obtain harmonic distortion.

[0035] Specifically, harmonic distortion represents the percentage of total harmonic energy relative to fundamental energy: ; in, To express harmonic distortion, first obtain the fundamental frequency, which is the system power frequency of the transformer, usually 50 or 60 Hz. By obtaining the highest peak in the amplitude spectrum, the highest peak is determined as the fundamental amplitude. , the harmonic frequency is an integer multiple of the fundamental wave, and the amplitude of each harmonic is extracted from the amplitude spectrum , for example, if the third harmonic frequency is three times the fundamental frequency, then the amplitude of the third harmonic is .

[0036] In step S35 , the resonance point offset, the nonlinearity index, and the harmonic distortion are combined into a transformer detection vector.

[0037] Specifically, for example, at a certain time point, the real-time resonance point offset of the transformer is 3%, the real-time nonlinearity index is 0.6, and the real-time harmonic distortion is 0.5. At this time, the transformer detection vector is (3%, 0.6, 50%).

[0038] Step S4: establish a three-dimensional coordinate system based on the transformer detection vector, obtain detection data points in the three-dimensional coordinate system, and perform real-time health detection on the transformer according to the distance from the detection data points to the origin of the three-dimensional coordinate system.

[0039] In this embodiment, step S4 includes: Step S41: The resonance point offset, nonlinearity index, and harmonic distortion are used as the X-axis, Y-axis, and Z-axis of the three-dimensional coordinate system, respectively. The resonance point offset, nonlinearity index, and harmonic distortion acquired in real time are mapped into the three-dimensional coordinate system, and detection data points are acquired. The detection data points are represented as coordinate points of the resonance point offset, nonlinearity index, and harmonic distortion data in the three-dimensional coordinate system. A transformer detection vector in a healthy state is acquired, and a three-dimensional healthy region is constructed based on the transformer detection vector in a healthy state. The three-dimensional healthy region is represented as a cuboid constructed according to the transformer vector in a healthy state. Health detection is performed based on the detection data points and the three-dimensional healthy region.

[0040] Specifically, the resonance point offset is used as the X-axis of the three-dimensional coordinate system, the nonlinearity index is used as the Y-axis of the three-dimensional coordinate system, and the harmonic distortion is used as the Z-axis of the three-dimensional coordinate system. The resonance point offset, nonlinearity index, and harmonic distortion obtained in real time are mapped to the three-dimensional coordinate system to obtain detection data points. For example, the detection data points obtained at a certain moment are (3%, 0.6, 50%). The obtained detection data points are preprocessed and converted into (0.03, 0.6, 0.5). A transformer detection vector in a healthy state is obtained and similarly preprocessed. In this embodiment, the transformer detection vector in a healthy state is (0.01, 0.3, 0.4). A cuboid is constructed based on the transformer detection vector in the healthy state. The cuboid represents a three-dimensional healthy area. At this time, it is determined whether the obtained detection data point (0.03, 0.6, 0.5) is within the three-dimensional healthy area to perform health detection.

[0041] In step S411 , if the detected data point is not within the three-dimensional healthy area, it indicates that the transformer is faulty.

[0042] For example, assuming that the detection data point is (0.03, 0.65, 0.53), the detection data point is not within the three-dimensional healthy area, indicating that the transformer is faulty.

[0043] In this embodiment, whether the transformer is faulty is determined by judging whether the acquired detection data points are within the three-dimensional healthy area. The resonance point offset reflects mechanical deformation, the nonlinearity index reflects electromagnetic nonlinearity, and the harmonic distortion reflects insulation degradation or load abnormality. The three correspond to the states of the three core systems of mechanical, electromagnetic, and insulation, respectively, forming a complementary diagnostic basis. At the same time, the parameters are converted into points in three-dimensional space. The normal state is concentrated in the three-dimensional healthy area, and the fault state diffuses outward. Through the three-dimensional coordinate system, it is possible to more intuitively judge whether the transformer is faulty, simplifying the complex multi-parameter analysis into an intuitive spatial image problem, while also improving the accuracy of transformer fault judgment.

[0044] Step S412: If there is no fault, the process ends.

[0045] In this embodiment, by determining whether a fault exists, if no fault exists, the process is terminated directly, and no subsequent calculations are required for the detection data points where no fault exists, thereby saving computing power.

[0046] Furthermore, if there is no fault in the transformer, normal transformer detection vectors are regularly extracted for analysis. If the extracted transformer detection vector is at the edge of the three-dimensional healthy area, it indicates that the transformer is at risk of failure. The fault type at the edge of the three-dimensional healthy area is obtained, and an early warning is issued. For example, if the normal transformer detection vector regularly extracted is (0.005, 0.29, 0.25), it indicates that the transformer is on the verge of moisture and an early warning is issued.

[0047] Step S5: If a fault exists, perform real-time fault type judgment based on the transformer detection vector and the fault standard map, perform weight matching according to the result of the real-time fault type judgment, obtain the fault characteristic value, compare the fault characteristic value with the fault characteristic library, and output the fault diagnosis result.

[0048] In this embodiment, step S5 includes: Step S51 , performing calculation based on the real-time resonance point offset, nonlinearity index and harmonic distortion and the standard resonance point offset, nonlinearity index and harmonic distortion in the fault standard map, obtaining the data with the largest deviation ratio and confirming the real-time fault type.

[0049] Specifically, the real-time resonance point offset, nonlinearity index and harmonic distortion are obtained as (3%, 0.4, 50%). The parameter range of the resonance point offset in the fault standard map is 0%-5%, the parameter range of the nonlinearity index is 0.3-0.8, and the parameter range of the harmonic distortion is 40%-80%. The deviation ratios of the resonance point offset, nonlinearity index and harmonic distortion are calculated to be 60%, 20% and 25% respectively. It is found that the data with the largest deviation ratio is the resonance point offset, and the real-time fault type of the transformer is judged to be a winding short circuit. In the fault standard map, if the resonance point offset is the data with the largest deviation ratio, it reflects that the real-time fault type of the transformer is a winding short circuit. If the nonlinearity index is the data with the largest deviation ratio, it reflects that the real-time fault type of the transformer is insulation moisture. If the harmonic distortion is the data with the largest deviation ratio, it reflects that the real-time fault type of the transformer is core aging.

[0050] In this embodiment, whether a fault exists is first determined, and then the real-time fault type is determined. The judgment of the transformer fault type is made more accurate through the hierarchical judgment method.

[0051] Furthermore, the fault standard map can be flexibly expanded by simply adding a new feature vector row, which includes the fault type and the parameter range corresponding to the fault type. For example, a new fault type of "casing crack" and the parameter range corresponding to "casing crack" can be added without reconstructing the algorithm framework.

[0052] Step S52: weight the real-time resonance point offset, nonlinearity index, and harmonic distortion according to the real-time fault type, obtain the fault characteristic value, compare the fault characteristic value with the data in the fault characteristic library, and output the transformer fault diagnosis result. The transformer fault diagnosis result includes the real-time fault type and fault level.

[0053] Specifically, if the real-time fault type is judged to be a winding short circuit, the weight ratio of the real-time resonance point offset, nonlinearity index and harmonic distortion is 6:3:1; if the real-time fault type is judged to be insulation moisture, the weight ratio of the real-time resonance point offset, nonlinearity index and harmonic distortion is 3:5:2; if the real-time fault type is judged to be core aging, the weight ratio of the real-time resonance point offset, nonlinearity index and harmonic distortion is 3:2:5. For example, the real-time resonance point offset, nonlinearity index and harmonic distortion are (3%, 0.4, 50%), and the real-time fault type is judged to be a winding short circuit, and the weight ratio is 6:3:1. According to the weighted summation, the fault characteristic value is 1.88, and it is compared with the data in the fault feature library. The fault diagnosis result of the output transformer is an ordinary short circuit of the winding short circuit.

[0054] Furthermore, according to the data area division of the resonance point offset, nonlinearity index and harmonic distortion and their weight ratio, the fault characteristic values ​​at different levels are obtained, and the fault characteristic library is constructed through the fault characteristic values ​​at different levels. If the real-time fault type is judged to be a winding short circuit, the fault characteristic value ranges of the three fault levels of slight short circuit, ordinary short circuit and severe short circuit are calculated and obtained respectively (1.36, 1.82], (1.82, 2.66], (2.66, 3.5). If the real-time fault type is insulation moisture, the fault characteristic value ranges of slight moisture are calculated and obtained respectively. The fault characteristic value ranges for the three fault levels of core aging, mild aging, and severe aging are (2.33, 3.06], (3.06, 5.35], and (5.35, 7.1), respectively. If the real-time fault type is core aging, the fault characteristic value ranges for the three fault levels of slight aging, normal aging, and severe aging are (2.36, 3.36], (3.36, 5.5], and (5.5, 7.1), respectively. This allows for early detection and accurate judgment of the fault severity, thus minimizing the higher repair costs and losses caused by fault expansion.

[0055] Figure 3 A schematic diagram of a transformer fault detection device provided by some embodiments of the present application that can implement the concept of the present application is shown.

[0056] Specifically, a transformer fault detection device includes: A construction module is used to construct a fault standard map and a fault feature library, and to construct a three-dimensional coordinate system for transformer detection vectors.

[0057] The data acquisition module is used to acquire the time domain response signals of the primary and secondary windings in the transformer, and perform Fourier transform on the time domain response signals to obtain the frequency domain transfer function.

[0058] The acquisition module is used to acquire the amplitude-frequency characteristic curve and the phase-frequency characteristic curve, acquire the resonance point offset in the amplitude-frequency characteristic curve, and acquire the nonlinearity index and harmonic distortion in the phase-frequency characteristic curve.

[0059] The judgment module is used to perform real-time health detection to determine whether the transformer is faulty.

[0060] The diagnostic module can perform real-time fault type judgment based on the transformer detection vector and the fault standard map, perform weight matching according to the result of the real-time fault type judgment, obtain the fault characteristic value, compare the fault characteristic value with the fault characteristic library, and output the fault diagnosis result.

[0061] The specific usage and function of this embodiment are described below: First, a fault standard map is constructed, and the fault standard map is constructed based on the existing transformer fault type. At the same time, a fault feature library is constructed, and the fault feature library is constructed based on the fault degree of the transformer. Then, a wide-band excitation signal is injected into the transformer, and the time domain response signals of the primary and secondary windings in the transformer are collected based on the data acquisition module. The time domain response signal is Fourier transformed to obtain the frequency domain transfer function. The amplitude-frequency characteristic curve and the phase-frequency characteristic curve are obtained based on the frequency domain transfer function. The resonance point offset in the amplitude-frequency characteristic curve is extracted, and the nonlinearity index and harmonic distortion in the phase-frequency characteristic curve are simultaneously extracted. The resonance point offset, the nonlinearity index and the harmonic distortion are combined into a transformer detection vector. Then, a three-dimensional coordinate system is established based on the transformer detection vector, and the detection data points in the three-dimensional coordinate system are obtained. Obtain the transformer detection vector in a healthy state, construct a three-dimensional healthy area based on the transformer detection vector in a healthy state, and perform health detection based on the detection data points and the three-dimensional healthy area. Save computing power through hierarchical judgment. If the detection data point is not within the three-dimensional healthy area, it means that the transformer is faulty. Then, perform real-time fault type judgment based on the transformer detection vector and the fault standard map. The hierarchical judgment method makes the transformer fault type judgment more accurate. Perform weight matching based on the result of the real-time fault type judgment, and obtain the fault characteristic value. Finally, compare the fault characteristic value with the fault characteristic library, and output the fault diagnosis result. If the detection data point is within the three-dimensional healthy area, it means that the transformer is not faulty, and the process ends. The hierarchical judgment method meets the needs of staged diagnosis.

[0062] In addition, an embodiment of the present invention further provides an electronic device, including: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method proposed in the first embodiment of the present invention.

[0063] The following is a detailed introduction to the various components of electronic equipment: The term "processor" is the control center of an electronic device and can be a single processor or a collective term for multiple processing elements. For example, the processor can be one or more central processing units (CPUs), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the first embodiment of the present invention, such as one or more digital signal processors (DSPs) or one or more field programmable gate arrays (FPGAs).

[0064] The processor can execute various functions of the electronic device by running or executing software programs stored in the memory and calling data stored in the memory.

[0065] The memory is used to store the software program for executing the solution of the present invention, and the execution is controlled by the processor. The specific implementation method can refer to the above method embodiment and will not be repeated here.

[0066] The memory may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, a random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only (CD-ROM), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto. The memory may be integrated with the processor or exist independently and be coupled to the processor via an interface circuit of the electronic device, and this is not specifically limited in the embodiments of the present invention.

[0067] The above embodiments can be implemented in whole or in part via software, hardware (e.g., circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in accordance with the embodiments of the present invention are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another. For example, the computer instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wireless communication (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer, or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0068] It should be understood that the term "and / or" as used herein simply describes an association between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent the existence of A alone, the existence of both A and B, or the existence of B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the related objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0069] It should be understood that in the embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0070] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A transformer fault detection method, characterized in that: The method comprises: Constructing a fault standard map based on existing transformer fault types and a fault feature library based on the degree of transformer faults; Inject a broadband excitation signal into the transformer, collect the time domain response signals of the primary and secondary windings in the transformer based on the data acquisition module, perform Fourier transform on the time domain response signals, and obtain the frequency domain transfer function; Based on the frequency domain transfer function, the amplitude-frequency characteristic curve and the phase-frequency characteristic curve are obtained, the resonance point offset in the amplitude-frequency characteristic curve is extracted, the nonlinearity index and harmonic distortion in the phase-frequency characteristic curve are simultaneously extracted, and the resonance point offset, nonlinearity index and harmonic distortion are combined into a transformer detection vector; Establishing a three-dimensional coordinate system based on the transformer detection vector, obtaining detection data points within the three-dimensional coordinate system, obtaining the transformer detection vector in a healthy state, constructing a three-dimensional healthy region based on the transformer detection vector in a healthy state, and performing health detection based on the detection data points and the three-dimensional healthy region; If the detection data point is not within the three-dimensional healthy area, it means that the transformer is faulty. The real-time fault type is determined based on the transformer detection vector and the fault standard map. The weight is then matched based on the result of the real-time fault type determination. The fault feature value is obtained and compared with the fault feature library to output the fault diagnosis result. If the detection data point is within the three-dimensional healthy area, it means that there is no fault in the transformer, and the process ends.

2. A transformer fault detection method according to claim 1, characterized in that: The step of injecting a broadband excitation signal into the transformer comprises: Based on the broadband excitation signal, the low-frequency, medium-frequency and high-frequency electromagnetic behaviors of the transformer are simultaneously stimulated, covering the full working frequency band of the transformer.

3. A transformer fault detection device and method according to claim 1, characterized in that: The data acquisition module collects time domain response signals of the primary and secondary windings in the transformer, performs Fourier transform on the time domain response signals, and obtains a frequency domain transfer function, including: The time domain response signal includes the input voltage and input current of the primary winding and the output voltage and output current of the secondary winding; Performing Fourier transform on the input voltage and input current of the primary winding and the output voltage and output current of the secondary winding to obtain the frequency domain input voltage and current of the primary winding and the frequency domain output voltage and current of the secondary winding after Fourier transform; The frequency domain voltage transfer ratio is calculated based on the frequency domain input voltage and the frequency domain output voltage, the frequency domain current transfer ratio is calculated based on the frequency domain input current and the frequency domain output current, and the frequency domain transfer function is obtained according to the frequency domain voltage transfer ratio and the frequency domain current transfer ratio.

4. A transformer fault detection method according to claim 1, characterized in that: The obtaining of the amplitude-frequency characteristic curve and the phase-frequency characteristic curve based on the frequency domain transfer function includes: The frequency, amplitude and phase are obtained according to the frequency domain transfer function. The amplitude-frequency characteristic curve is drawn based on the amplitude and frequency, and the phase-frequency characteristic curve is drawn based on the phase and frequency.

5. A transformer fault detection method according to claim 1, characterized in that: The method of extracting the resonance point offset in the amplitude-frequency characteristic curve, synchronously extracting the nonlinearity index and harmonic distortion in the phase-frequency characteristic curve, and combining the resonance point offset, the nonlinearity index, and the harmonic distortion into a transformer detection vector includes: Obtain the resonant frequency of the amplitude-frequency characteristic curve, compare the resonant frequency with a preset standard value, and calculate the offset of the resonance point; The phase-frequency characteristic curve is converted into a second-order polynomial, wherein the quadratic term coefficient of the second-order polynomial is the nonlinearity index; Perform harmonic analysis on the time domain response signal of the secondary winding after Fourier transform to obtain the harmonic distortion; The resonance point offset, nonlinearity index and harmonic distortion are combined into a transformer detection vector.

6. A transformer fault detection method according to claim 1, characterized in that: The method includes establishing a three-dimensional coordinate system based on the transformer detection vector, obtaining detection data points within the three-dimensional coordinate system, obtaining a transformer detection vector in a healthy state, constructing a three-dimensional healthy area based on the transformer detection vector in the healthy state, and performing health detection based on the detection data points and the three-dimensional healthy area, including: The resonance point offset, nonlinearity index, and harmonic distortion are respectively used as the X-axis, Y-axis, and Z-axis of the three-dimensional coordinate system, and the real-time acquired resonance point offset, nonlinearity index, and harmonic distortion are mapped into the three-dimensional coordinate system, and detection data points are obtained. The detection data points are represented as coordinate points of the resonance point offset, nonlinearity index, and harmonic distortion data in the three-dimensional coordinate system. A transformer detection vector in a healthy state is obtained, and a three-dimensional healthy region is constructed based on the transformer detection vector in a healthy state. The three-dimensional healthy region is represented as a cuboid constructed according to the transformer vector in a healthy state, and health detection is performed based on the detection data points and the three-dimensional healthy region: If the detection data point is not within the three-dimensional healthy area, it means that the transformer is faulty; If the detection data point is within the three-dimensional healthy area, it means that there is no fault in the transformer.

7. A transformer fault detection method according to claim 1, characterized in that: The method of performing real-time fault type judgment based on the transformer detection vector and the fault standard map, performing weight matching according to the result of the real-time fault type judgment, obtaining the fault characteristic value, comparing the fault characteristic value with the fault characteristic library, and outputting the fault diagnosis result includes: Calculate the real-time resonance point offset, nonlinearity index, and harmonic distortion against the standard resonance point offset, nonlinearity index, and harmonic distortion in the fault standard map to obtain the data with the largest deviation and confirm the real-time fault type. According to the real-time fault type, the real-time resonance point offset, nonlinearity index and harmonic distortion are weighted and matched to obtain the fault characteristic value. The fault characteristic value is compared with the data in the fault characteristic library, and the fault diagnosis result of the transformer is output. The fault diagnosis result of the transformer includes the real-time fault type and fault level.

8. A transformer fault detection method according to claim 7, characterized in that: The fault standard map includes: Based on the existing transformer fault types, the data areas of resonance point offset, nonlinearity index and harmonic distortion are divided to obtain the transformer detection vector range of different existing transformer fault types. Based on the existing transformer fault types and transformer detection vector range, a fault standard map is constructed.

9. A transformer fault detection method according to claim 7, characterized in that: The fault feature library includes: The fault severity is divided into three fault levels: minor fault, common fault and severe fault. The data area is divided for each fault level to obtain the transformer fault level range of each fault. A fault feature library is constructed based on the real-time fault type, transformer fault level and fault level range.

10. A transformer fault detection device, characterized in that: include: A construction module, wherein the construction module is used to construct a fault standard map and a fault feature library, and to construct a three-dimensional coordinate system of a transformer detection vector; A data acquisition module, wherein the data acquisition module is used to acquire time domain response signals of the primary and secondary windings in the transformer, and perform Fourier transform on the time domain response signals to obtain a frequency domain transfer function; An acquisition module, the acquisition module is used to acquire an amplitude-frequency characteristic curve and a phase-frequency characteristic curve, acquire a resonance point offset in the amplitude-frequency characteristic curve, and acquire a nonlinearity index and harmonic distortion in the phase-frequency characteristic curve; A judgment module, which is used to perform real-time health detection to determine whether the transformer is faulty; The diagnostic module can perform real-time fault type judgment based on the transformer detection vector and the fault standard map, perform weight matching according to the result of the real-time fault type judgment, obtain the fault characteristic value, compare the fault characteristic value with the fault characteristic library, and output the fault diagnosis result.

Images