Transformer winding fault diagnosis test platform and diagnosis method

Abstract

The invention discloses a transformer winding fault diagnosis test platform and a diagnosis method, and relates to the technical field of transformer winding fault diagnosis. The method comprises the following steps: firstly, controlling a digital signal generator to emit a standard excitation signal by using a computer, obtaining response signals at different tap positions, and then performing Fourier transform on an original signal to obtain frequency domain information; then calculating amplitude-frequency response curves at different tap positions, extracting amplitude-frequency data at different tap positions, and further calculating accumulated distances from different tap positions to the reference position; and finally, calculating weights, fusing the weights to calculate a winding state factor to judge the winding state, and if the winding state is abnormal, calculating a response error index F to judge the winding fault type. According to the transformer winding state diagnosis method provided by the invention, the fault of the transformer winding can be accurately and effectively diagnosed.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer winding fault diagnosis, and in particular to a transformer winding fault diagnosis test platform and a diagnosis method. Background Art

[0002] The health of transformer windings plays a vital role in maintaining the stability and reliability of power systems. As the core component of the transformer, the winding undertakes the key task of power conversion, and its performance is directly related to the efficiency of power transmission and the reliability of power supply. Considering the complex structural characteristics of transformer windings and long-term operation in high-voltage environments, they are susceptible to factors such as overload, adverse climatic conditions, and material degradation, which may cause failures. Once a transformer winding fails, the consequences are multifaceted: it may not only cause the transformer to stop working and power supply to be interrupted, thereby affecting the normal power consumption of users and causing economic losses, but also cause short circuits or partial discharges inside the transformer, increase the risk of fire or explosion, and pose a threat to the safety of personnel and equipment. At the same time, winding failures may also lead to a decrease in transformer efficiency, increase the burden on the power system, and affect the stability of the entire power system. Therefore, regular monitoring and fault diagnosis of transformer windings are key steps to ensure the safe operation of the power system.

[0003] In the field of transformer winding fault diagnosis, the application of multi-point response signal fusion technology is particularly important. By injecting test signals from different locations of the transformer and collecting response data, this method can provide more comprehensive information on the winding status. The response signals at each point contain unique electrical and physical characteristics. Fusion of these signals can reveal local defects and overall health status inside the winding. The integrated analysis of multi-point data enhances the accuracy and robustness of fault detection, making the diagnosis results more reliable. In addition, signal fusion technology can improve the accuracy of fault location, reduce false alarms and missed alarms, thereby optimizing maintenance strategies, reducing maintenance costs, and ultimately ensuring the continuity and safety of the power system. Therefore, multi-point response signal fusion is not only a technological advancement for transformer winding fault diagnosis, but also an indispensable part of power system maintenance. Summary of the invention

[0004] In order to solve the above problems, the present invention proposes a transformer winding fault diagnosis test platform and a diagnosis method. According to the proposed method, transformer winding faults can be diagnosed accurately and effectively.

[0005] The technical solution adopted by the present invention is:

[0006] A transformer winding fault diagnosis test platform comprises a transformer box, an iron core installed in the transformer box, a high-voltage winding, a medium-voltage winding, a low-voltage winding, and a tap switch control module, and a high-voltage bushing, a medium-voltage bushing, and a low-voltage bushing installed on the transformer and electrically connected to the high-voltage winding, the medium-voltage winding, and the low-voltage winding; the transformer winding fault diagnosis test platform also comprises a resistor a, a resistor b, a computer, a digital signal generator, a noise reduction filter module, and a signal acquisition module; the computer is connected to the digital signal generator, the digital signal generator is connected to the resistor a, and the resistor a is connected to the high-voltage bushing; the low-voltage bushing is connected to the resistor b, the resistor b is connected to the noise reduction filter module, the noise reduction filter module is connected to the signal acquisition module, and the signal acquisition module is connected to the computer.

[0007] A transformer winding fault diagnosis method is provided. The transformer winding fault diagnosis method is based on the above-mentioned transformer winding fault diagnosis test platform. The transformer winding fault diagnosis method combines amplitude-frequency data at different tap positions to calculate feature weights and fuses them to determine the winding state. The method comprises the following steps:

[0008] Step 1: Generate a standard excitation signal by controlling a digital signal generator through a computer, inject it into the transformer winding, and measure the input and output voltage response signals at different tap positions to provide a data basis for subsequent analysis;

[0009] Step 2: Calculate the amplitude-frequency response curve A by performing Fourier transform on the input and output voltages at different tap positions. Z , and intercept the amplitude data X in the frequency range of 1kHz-1MHz Z , providing frequency domain features for subsequent analysis;

[0010] Step 3: Calculate the ratio of the sum of the absolute differences between the non-reference position and the reference position in the amplitude-frequency response data to the combination of the amplitude extreme values ​​to obtain the cumulative distance D Z , used to quantify the difference between the two;

[0011] Step 4: By comprehensively considering the variance of the amplitude-frequency data and the difference in frequency domain characteristics, the distance collaborative optimization weight w of the non-reference position Z≠1 is calculated. z ,provides more accurate weight distribution for fault diagnosis;

[0012] Step 5, by calculating the winding state coefficient Y i The winding state factor W is combined with the threshold K to determine whether the transformer winding state is abnormal, thus achieving comprehensive evaluation and diagnosis of the winding state;

[0013] Step 6, by calculating the response error index F and combining it with the threshold G, it is determined whether the transformer winding has a turn-to-turn short circuit fault or a bulging warping fault, so as to achieve accurate diagnosis of the fault type.

[0014] Further, step 1 includes the following operation process:

[0015] Excitation signal generation: Use a computer-controlled digital signal generator to generate standard excitation signals;

[0016] Signal injection: The excitation signal is injected into the transformer winding through the high-voltage bushing, and flows through the high-voltage winding, medium-voltage winding and low-voltage winding in sequence;

[0017] Signal output: The signal flows out of the transformer through the low-voltage bushing, and after being processed by the noise reduction filter module and the signal acquisition module, the computer measures the output response;

[0018] The measurement result of step 1 is expressed as:

[0019] Input Voltage: In the formula, It represents the input voltage signal at the zth tap position, which is a time function, indicating the change of input voltage over time; Indicates the amplitude of the input voltage at a certain moment; It represents the amplitude of the input voltage after stabilization under the z-th position tap;

[0020] Output voltage: In the formula, It represents the output voltage signal at the zth tap position, which is also a time function, indicating the change of output voltage over time; Indicates the amplitude of the output voltage at a certain moment; It represents the amplitude of the output voltage after stabilization at the z-th position tap;

[0021] The tap position is expressed as:

[0022] Z=[12···m]; where z represents the number of the tap position; z=1 is the reference position, and the value of m depends on the voltage regulation method of the transformer.

[0023] Further, step 2 includes the following operation process:

[0024] Fourier transform: Input voltage at different tap positions Output voltage Perform Fourier transform to obtain the frequency domain information of input voltage and output voltage

[0025] Amplitude-frequency response calculation: Calculate the amplitude-frequency response curve based on frequency domain information:

[0026]

[0027] In the formula, A Z It is represented as the amplitude-frequency response curve when the tap position is z; z represents the number of the tap position;

[0028] Data interception: intercept A Z The amplitude data of the intermediate frequency in the range of 1kHz-1MHz is recorded as:

[0029] In the formula, each Indicates the amplitude corresponding to a certain frequency point in the frequency range of 1kHz-1MHz;

[0030] The corresponding frequency is:

[0031] [f1 f2···f n ], indicating that the amplitude data X Z The corresponding frequency points are in the range of 1kHz-1MHz;

[0032] The output of step 2 is:

[0033] Amplitude-frequency response curve A at each tap position z Z ;

[0034] The clipped amplitude data X Z and its corresponding frequency points [f1 f2···f n ].

[0035] Further, in step 3, the cumulative distance D is calculated Z The calculation formula is:

[0036]

[0037] Where D Z It is expressed as the cumulative distance when the tap position is Z and when the tap position is the reference; is the sum of the absolute differences between the amplitude data of the zth tap position and the reference position at each frequency point; is the amplitude data of the ith frequency point at the zth tap position; is the amplitude data of the ith frequency point at the reference position z=1; n is the total number of frequency points; is a combination of the maximum difference and the minimum sum of the amplitude data of the z-th tap position and the reference position at the last frequency point i=n; Indicates the maximum difference between the amplitude data of the zth tap position and the reference position at the last frequency point i=n; It represents the minimum sum of the amplitude data of the zth tap position and the reference position at the last frequency point i=n; z represents the number of the tap position; Z≠1 means the cumulative distance DZ The calculation is performed only for non-reference positions Z=2, 3, ..., m, and the reference position Z=1 is used as a reference point and does not participate in the calculation of its own difference.

[0038] Further, in step 4, the distance collaborative optimization weight w is calculated z The calculation formula is:

[0039]

[0040] In the formula, Var z It is expressed as the amplitude-frequency data X when the tap position is Z z The variance of z It is expressed as the distance collaborative optimization weight when the tap position is Z; It represents the sum of the variance of amplitude-frequency data at all tap positions; is the amplitude data of the ith frequency point at the zth tap position; is the amplitude data of the ith frequency point at the reference position z=1; f i represents the frequency value of the ith frequency point; D i represents the cumulative distance corresponding to the i-th frequency point; e () represents an exponential function, which is used to amplify the difference of frequency domain features; z represents the number of the tap position; Z≠1 represents the distance collaborative optimization weight w z The calculation is performed only for non-reference positions Z=2, 3, ..., m, and the reference position Z=1 is used as a reference point and does not participate in the weight calculation.

[0041] Further, in step 5, the winding state coefficient Y i The calculation formula is as follows:

[0042]

[0043] Where Y i The winding state coefficient of the i-th tap position; D Z It is expressed as the cumulative distance between the tap position Z and the tap position as the reference; max(D 1 ··· D Z ) represents the maximum value of the cumulative distance of all tap positions; min(D 1 ··· D Z ) represents the minimum value of the cumulative distance of all tap positions; It represents the sum of the square differences between the cumulative distances of all tap positions and the cumulative distances of the reference position; i represents the number of the tap position; n is the total number of taps; i=1 represents the reference position, and its winding state coefficient is fixed to 1, that is, Y1=1; i≠1 represents the non-reference position, and its winding state coefficient is calculated by the formula.

[0044] Further, in step 5, the calculation formula of the winding state factor W is as follows:

[0045]

[0046] Where W represents the winding condition factor, which is used to comprehensively evaluate the overall condition of the transformer winding; Y i represents the winding state coefficient of the i-th tap position; w i represents the distance collaborative optimization weight of the i-th tap position; Indicates the winding state coefficient Y i Co-optimize weight w with distance i The weighted sum of i 、sinY i Respectively represent the winding state coefficient Y i The cosine and sine of ; represents the sum of the cosine and sine values ​​of all tap positions; represents the sum of the distance co-optimization weights of all tap positions; ln() represents the natural logarithm function, which is used to adjust the scale and distribution of the data; e () represents an exponential function used to amplify the difference characteristics; n is the total number of taps.

[0047] Further, in step 5, the condition for determining the winding state is:

[0048] If W ≥ K, the transformer winding state is judged to be abnormal;

[0049] If W<K, it is judged that the transformer winding state is normal;

[0050] Where W represents the winding state factor, which is used to comprehensively evaluate the overall state of the transformer winding; K is a constant related to the transformer model, winding signal injection method, and external equipment.

[0051] Further, in step 6, the calculation formula of the response error index F is as follows:

[0052]

[0053] Where F is the response error index, which is used to comprehensively evaluate the fault type and severity of the winding; Represents the amplitude data of the i-th frequency point at the z-th tap position; is the amplitude data of the ith frequency point at the reference position z=1; f i Represents the frequency value of the i-th frequency point; It represents the square difference of the amplitude data of the zth tap position and the reference position at the ith frequency point; represents the frequency weight factor, which is used to weight the amplitude difference according to the frequency value; W represents the winding state factor; ln(1+W) represents the natural logarithm of the winding state factor W plus 1, which is used to adjust the proportion and distribution of the data; w z represents the distance collaborative optimization weight of the zth tap position; w1 represents the distance collaborative optimization weight of the 1st tap position; |w z -w1| represents the absolute difference of the distance co-optimization weight between the zth tap position and the 1st tap position; represents a double summation of all tap positions and frequency points;

[0054] The conditions for determining the fault type are:

[0055] If F≥G, it is judged that the transformer winding has a turn-to-turn short circuit fault;

[0056] If F<G, it is judged that the transformer winding has a bulging and warping fault;

[0057] Where G is a constant related to the transformer model, winding signal injection method, and external equipment.

[0058] The beneficial effects of the present invention are:

[0059] The transformer winding fault diagnosis test platform and diagnosis method have the following significant beneficial effects:

[0060] Comprehensiveness: By measuring the response signals at different tap positions and obtaining multi-point amplitude and frequency data, the overall state of the transformer winding can be fully reflected, avoiding misjudgment that may be caused by single-position measurement.

[0061] High precision: The time domain signal is converted into frequency domain information using Fourier transform, and combined with the amplitude-frequency response curve and feature weight calculation, the fault characteristics of the winding can be extracted more accurately, improving the accuracy of diagnosis.

[0062] Strong anti-interference ability: The signal is processed through the noise reduction filter module and the signal acquisition module, which effectively reduces the impact of noise and external interference on the measurement results and improves the reliability of the data.

[0063] Dynamic weight optimization: By calculating the distance collaborative optimization weight, comprehensively considering the variance of the amplitude-frequency data and the difference in frequency domain characteristics, the weight of each tap position is dynamically adjusted to optimize the weight distribution of fault diagnosis.

[0064] Multi-dimensional evaluation: By accumulating multi-dimensional indicators such as distance, winding condition coefficient, winding condition factor, etc., the overall condition of the winding is comprehensively evaluated, which can more comprehensively reflect the health status of the winding.

[0065] Fault type identification: By calculating the response error index and combining the threshold to determine whether the winding has a turn-to-turn short circuit fault or a bulging and warping fault, accurate identification of the fault type is achieved.

[0066] Strong adaptability: The threshold in the method can be adjusted according to the transformer model, winding signal injection method and external equipment, and has strong adaptability and versatility.

[0067] High degree of automation: Through computer control of digital signal generator, signal acquisition module and data processing module, a fully automated process from signal injection, data acquisition to fault diagnosis is realized, which improves the diagnostic efficiency.

[0068] Early warning: Through multi-point response signal fusion and frequency domain analysis, potential faults of windings can be discovered early to avoid further deterioration of the faults and improve the operational reliability of the transformer.

[0069] Scalability: This method can be combined with other fault diagnosis methods to further expand its application scope and diagnostic capabilities.

[0070] In summary, the transformer winding fault diagnosis test platform and diagnosis method realize a comprehensive and accurate evaluation of the transformer winding status through multi-point response signal fusion, frequency domain analysis and dynamic weight optimization. It has strong anti-interference ability, adaptability and degree of automation, can detect and identify winding fault types at an early stage, and improve the reliability and safety of transformer operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0072] Figure 1 It is a system diagram of the transformer winding fault diagnosis test platform of the present invention;

[0073] Figure 2 is a flow chart of a transformer winding fault diagnosis method of the present invention;

[0074] In the figure, 1-transformer box, 2-iron core, 3-low voltage winding, 4-medium voltage winding, 5-high voltage winding, 6-high voltage bushing, 7-low voltage bushing, 8-computer, 9-digital signal generator, 10-tap switch control module, 11-noise reduction filter module, 12-medium voltage bushing, 13-resistor a, 14-resistor b, 15-signal acquisition module. DETAILED DESCRIPTION

[0075] To make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be described clearly and completely in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them.

[0076] Therefore, the following detailed description of the embodiments of the present invention is not intended to limit the scope of the claimed invention, but merely represents some embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. It should be noted that the embodiments of the present invention and the features and technical solutions in the embodiments can be combined with each other without conflict. It should be noted that similar numbers and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0077] For transformer winding fault diagnosis, this embodiment provides a transformer winding fault diagnosis test platform; Figure 1 As shown, the transformer winding fault diagnosis test platform includes a transformer box 1, an iron core 2, a high-voltage winding 5, a medium-voltage winding 4, a low-voltage winding 3, and a tap switch control module 10 installed in the transformer box 1, and a high-voltage bushing 6, a medium-voltage bushing 12, and a low-voltage bushing 7 installed on the transformer and electrically connected to the high-voltage winding 5, the medium-voltage winding 4, and the low-voltage winding 3. In addition, the transformer winding fault diagnosis test platform also includes a resistor a13, a resistor b14, a computer 8, a digital signal generator 9, a noise reduction filter module 11, and a signal acquisition module 15; the computer 8 is connected to the digital signal generator 9, the digital signal generator 9 is connected to the resistor a13, and the resistor a13 is connected to the high-voltage bushing 6; the low-voltage bushing 7 is connected to the resistor b14, the resistor b14 is connected to the noise reduction filter module 11, the noise reduction filter module 11 is connected to the signal acquisition module 15, and the signal acquisition module 15 is connected to the computer 8.

[0078] Based on the above-mentioned transformer winding fault diagnosis test platform, this embodiment also provides a transformer winding fault diagnosis method. The transformer winding fault diagnosis method first uses a computer to control a digital signal generator to send a standard excitation signal to obtain a response signal at different tap positions, and then performs Fourier transform on the original signal to obtain frequency domain information; then calculates the amplitude-frequency response curve at different tap positions, extracts the amplitude-frequency data at different tap positions, and further calculates the cumulative distance of different tap positions to the reference position; finally, calculates the weight, and calculates the winding state factor by fusing the weight to judge the winding state. If the winding state is abnormal, then calculates the response error index F to judge the winding fault type, so that the transformer winding fault can be accurately and effectively diagnosed according to the proposed method.

[0079] The transformer winding fault diagnosis method combines the amplitude-frequency data at different tap positions to calculate the feature weights and fuses them to determine the winding status, such as Figure 2 As shown, the following steps are included:

[0080] Step 1: Generate a standard excitation signal through a computer-controlled digital signal generator, inject it into the transformer winding, and measure the input and output voltage response signals at different tap positions to provide a data basis for subsequent analysis.

[0081] Step 1 includes the following steps:

[0082] Excitation signal generation: Use a computer-controlled digital signal generator to generate standard excitation signals;

[0083] Signal injection: The excitation signal is injected into the transformer winding through the high-voltage bushing, and flows through the high-voltage winding, medium-voltage winding and low-voltage winding in sequence;

[0084] Signal output: The signal flows out of the transformer through the low-voltage bushing, and after being processed by the noise reduction filter module and the signal acquisition module, the computer measures the output response;

[0085] The measurement result of step 1 is expressed as:

[0086] Input Voltage: In the formula, It represents the input voltage signal at the zth tap position, which is a time function, indicating the change of input voltage over time; Indicates the amplitude of the input voltage at a certain moment; It represents the amplitude of the input voltage after stabilization under the z-th position tap;

[0087] Output voltage: In the formula, It represents the output voltage signal at the zth tap position, which is also a time function, indicating the change of output voltage over time; Indicates the amplitude of the output voltage at a certain moment; It represents the amplitude of the output voltage after stabilization at the z-th position tap;

[0088] The tap position is expressed as:

[0089] Z=[12···m]; where z represents the number of the tap position; z=1 is the reference position, and the value of m depends on the voltage regulation method of the transformer.

[0090] Core output: The input and output voltage signals at different tap positions are obtained by measuring, providing data support for subsequent frequency response analysis and fault diagnosis.

[0091] The core function of step 1 is to inject the standard excitation signal into the transformer winding and measure the input and output voltage response signals under different tap positions, laying the foundation for subsequent frequency response data calculation and fault diagnosis.

[0092] Step 2: Calculate the amplitude-frequency response curve A by performing Fourier transform on the input and output voltages at different tap positions. Z , and intercept the amplitude data X in the frequency range of 1kHz-1MHz Z , providing frequency domain features for subsequent analysis.

[0093] Step 2 includes the following steps:

[0094] Fourier transform: Input voltage at different tap positions Output voltage Perform Fourier transform to obtain the frequency domain information of input voltage and output voltage

[0095] Amplitude-frequency response calculation: Calculate the amplitude-frequency response curve based on frequency domain information:

[0096]

[0097] In the formula, A Z It is represented as the amplitude-frequency response curve when the tap position is z; z represents the number of the tap position;

[0098] Data interception: intercept A Z The amplitude data of the intermediate frequency in the range of 1kHz-1MHz is recorded as:

[0099] In the formula, each Indicates the amplitude corresponding to a certain frequency point in the frequency range of 1kHz-1MHz;

[0100] The corresponding frequency is:

[0101] [f1 f2···f n ], indicating that the amplitude data X Z The corresponding frequency points are in the range of 1kHz-1MHz;

[0102] The output of step 2 is:

[0103] Amplitude-frequency response curve A at each tap position z Z ;

[0104] The clipped amplitude data X Z and its corresponding frequency points [f1 f2···f n ].

[0105] The core function of step 2 is to convert the time domain signal into frequency domain information through Fourier transform, calculate the amplitude-frequency response curve, and extract the amplitude data within the key frequency range to provide frequency domain features for subsequent fault diagnosis.

[0106] Step 3, by calculating the ratio of the sum of the absolute differences between the non-reference position and the reference position in the amplitude-frequency response data to the combination of the amplitude extreme values, the cumulative distance is obtained to quantify the difference between the two.

[0107] Calculate the cumulative distance D Z The specific process is:

[0108] First, calculate the sum of the absolute differences between the amplitude data of the zth tap position and the reference position at each frequency point, that is,

[0109] Secondly, calculate the combination of the maximum difference and the minimum sum of the amplitude data of the z-th tap position and the reference position at the last frequency point i=n; that is,

[0110] Finally, calculate the cumulative distance D Z , expressed as:

[0111]

[0112] Where D Z It is expressed as the cumulative distance when the tap position is Z and when the tap position is the reference; is the amplitude data of the ith frequency point at the reference position z=1; n is the total number of frequency points; Indicates the maximum difference between the amplitude data of the zth tap position and the reference position at the last frequency point i=n; represents the minimum sum of the amplitude data of the zth tap position and the reference position at the last frequency point i=n;

[0113] Where z represents the number of the tap position; Z≠1 means the cumulative distance D Z The calculation is performed only for non-reference positions Z=2, 3, ..., m, and the reference position Z=1 is used as a reference point and does not participate in the calculation of its own difference.

[0114] D Z It reflects the overall difference between the zth tap position and the reference position in the amplitude-frequency response data; the larger its value, the greater the difference between the two; the smaller its value, the smaller the difference between the two. Through the above calculation, step 3 can output the cumulative distance D corresponding to each non-reference position z Z .

[0115] Step 3: Calculate the cumulative distance D Z , quantifies the difference in amplitude-frequency response data between the non-reference position and the reference position, and provides a difference measurement index for subsequent fault diagnosis. The reference position Z = 1 is used as a reference point and does not participate in the calculation of its own difference.

[0116] Step 4: By comprehensively considering the variance of the amplitude-frequency data and the difference in frequency domain characteristics, the distance collaborative optimization weight w of the non-reference position Z≠1 is calculated. z , providing more accurate weight distribution for fault diagnosis.

[0117] Calculate the distance collaborative optimization weight w z The specific process is:

[0118] First, calculate the amplitude-frequency data X Z Variance Var z , and normalize it, that is,

[0119] Then, the difference weights of the frequency domain features are calculated, including a combination of sine function, cosine function and exponential function, which are used to describe the difference characteristics of the amplitude-frequency data in the frequency domain;

[0120] Finally, the variance normalization result is multiplied by the frequency domain difference weight to obtain the distance collaborative optimization weight w of the zth tap position: z , expressed as:

[0121]

[0122] In the formula, Var z It is expressed as the amplitude-frequency data X when the tap position is Z z The variance of z It is expressed as the distance collaborative optimization weight when the tap position is Z; It represents the sum of the variance of amplitude-frequency data at all tap positions; is the amplitude data of the ith frequency point at the zth tap position; is the amplitude data of the ith frequency point at the reference position z=1; f i represents the frequency value of the ith frequency point; D i represents the cumulative distance corresponding to the i-th frequency point; e () represents an exponential function, which is used to amplify the difference of frequency domain features; z represents the number of the tap position; Z≠1 represents the distance collaborative optimization weight w z The calculation is performed only for non-reference positions Z=2, 3, ..., m, and the reference position Z=1 is used as a reference point and does not participate in the weight calculation.

[0123] w z Taking into account the variance of the amplitude-frequency data and the difference in frequency domain characteristics, it can more accurately reflect the degree of difference between the zth tap position and the reference position; the variance part reflects the discreteness of the data, and the frequency domain difference part reflects the variation characteristics of the data in the frequency domain. The output of step 4 is the distance collaborative optimization weight w corresponding to each non-reference position z. z .

[0124] Step 4: Calculate the distance collaborative optimization weight w of the non-reference position by comprehensively considering the variance of the amplitude-frequency data and the difference in frequency domain characteristics. z , providing more accurate weight distribution for subsequent fault diagnosis. The reference position Z = 1 is used as a reference point and does not participate in the weight calculation.

[0125] Step 5: By calculating the winding state coefficient and the winding state factor W and combining the threshold K, it is determined whether the transformer winding state is abnormal, thereby achieving a comprehensive evaluation and diagnosis of the winding state.

[0126] Specifically, first calculate the winding state coefficient Y i , the calculation formula is as follows:

[0127]

[0128] Where Y i The winding state coefficient of the i-th tap position; D Z It is expressed as the cumulative distance between the tap position Z and the tap position as the reference; max(D 1 ··· D Z ) represents the maximum value of the cumulative distance of all tap positions; min(D 1 ··· D Z ) represents the minimum value of the cumulative distance of all tap positions; It represents the sum of the square differences between the cumulative distances of all tap positions and the cumulative distances of the reference position; i represents the number of the tap position; n is the total number of taps; i=1 represents the reference position, and its winding state coefficient is fixed to 1, that is, Y1=1; i≠1 represents the non-reference position, and its winding state coefficient is calculated by the formula.

[0129] Then, calculate the winding state factor W, the calculation formula is as follows:

[0130]

[0131] Where W represents the winding condition factor, which is used to comprehensively evaluate the overall condition of the transformer winding; Y i represents the winding state coefficient of the i-th tap position; w i represents the distance collaborative optimization weight of the i-th tap position; Indicates the winding state coefficient Y i Co-optimize weight w with distance i The weighted sum of i 、sinY i Respectively represent the winding state coefficient Y i The cosine and sine of ; represents the sum of the cosine and sine values ​​of all tap positions; represents the sum of the distance co-optimization weights of all tap positions; ln() represents the natural logarithm function, which is used to adjust the scale and distribution of the data; e () represents an exponential function used to amplify the difference characteristics; n is the total number of taps.

[0132] Finally, the judgment is made based on the judgment conditions of the winding state:

[0133] If W ≥ K, the transformer winding state is judged to be abnormal;

[0134] If W<K, it is judged that the transformer winding state is normal;

[0135] Where W represents the winding state factor, which is used to comprehensively evaluate the overall state of the transformer winding; K is a constant related to the transformer model, winding signal injection method, and external equipment.

[0136] Step 5: Calculate the winding state factor Y i The transformer winding status factor W is used to comprehensively evaluate the status of the transformer winding and determine whether the winding is abnormal based on the threshold value K. The status factor of the reference position i=1 is fixed to 1, and the status factor of the non-reference position is calculated by accumulating the distance.

[0137] Step 6, by calculating the response error index F and combining it with the threshold G, it is determined whether the transformer winding has a turn-to-turn short circuit fault or a bulging warping fault, so as to achieve accurate diagnosis of the fault type.

[0138] Specifically, first calculate the response error index F, the calculation formula is as follows:

[0139]

[0140] Where F is the response error index, which is used to comprehensively evaluate the fault type and severity of the winding; Represents the amplitude data of the i-th frequency point at the z-th tap position; is the amplitude data of the ith frequency point at the reference position z=1; f i Represents the frequency value of the i-th frequency point; It represents the square difference of the amplitude data of the zth tap position and the reference position at the ith frequency point; represents the frequency weight factor, which is used to weight the amplitude difference according to the frequency value; W represents the winding state factor; ln(1+W) represents the natural logarithm of the winding state factor W plus 1, which is used to adjust the proportion and distribution of the data; w z represents the distance collaborative optimization weight of the zth tap position; w1 represents the distance collaborative optimization weight of the 1st tap position; |w z -w1| represents the absolute difference of the distance co-optimization weight between the zth tap position and the 1st tap position; represents a double summation over all tap positions and frequency points.

[0141] Then, the fault type is determined based on the following criteria:

[0142] If F≥G, it is judged that the transformer winding has a turn-to-turn short circuit fault;

[0143] If F<G, it is judged that the transformer winding has a bulging and warping fault;

[0144] Where G is a constant related to the transformer model, winding signal injection method, and external equipment.

[0145] Step 6 determines the fault type of the transformer winding, i.e., turn-to-turn short circuit or bulging warping, by calculating the response error index F, combining the amplitude data and the weight difference, and gives the final diagnosis result according to the threshold G.

[0146] At the same time, it should be pointed out that the above description is only a typical embodiment of the present invention and does not limit the present invention in any form. Any changes made to the above embodiments based on the ideas and technical essence of the present invention are within the protection scope of the present invention.

Claims

1. A transformer winding fault diagnosis test platform, comprising a transformer box, an iron core installed in the transformer box, a high-voltage winding, a medium-voltage winding, a low-voltage winding, and a tap switch control module, a high-voltage bushing, a medium-voltage bushing, and a low-voltage bushing installed on the transformer and electrically connected to the high-voltage winding, the medium-voltage winding, and the low-voltage winding; characterized in that: The transformer winding fault diagnosis test platform also includes a resistor a, a resistor b, a computer, a digital signal generator, a noise reduction filter module and a signal acquisition module; the computer is connected to the digital signal generator, the digital signal generator is connected to the resistor a, and the resistor a is connected to the high-voltage bushing; the low-voltage bushing is connected to the resistor b, the resistor b is connected to the noise reduction filter module, the noise reduction filter module is connected to the signal acquisition module, and the signal acquisition module is connected to the computer.

2. A transformer winding fault diagnosis method, the transformer winding fault diagnosis method is based on the transformer winding fault diagnosis test platform according to claim 1, characterized in that: The transformer winding fault diagnosis method combines the amplitude-frequency data at different tap positions to calculate the feature weights and fuses them to determine the winding state, including the following steps: Step 1: Generate a standard excitation signal by controlling a digital signal generator through a computer, inject it into the transformer winding, and measure the input and output voltage response signals at different tap positions to provide a data basis for subsequent analysis; Step 2: Calculate the amplitude-frequency response curve A by performing Fourier transform on the input and output voltages at different tap positions. Z , and intercept the amplitude data X in the frequency range of 1kHz-1MHz Z , providing frequency domain features for subsequent analysis; Step 3: Calculate the ratio of the sum of the absolute differences between the non-reference position and the reference position in the amplitude-frequency response data to the combination of the amplitude extreme values ​​to obtain the cumulative distance D Z , used to quantify the difference between the two; Step 4: By comprehensively considering the variance of the amplitude-frequency data and the difference in frequency domain characteristics, the distance collaborative optimization weight w of the non-reference position Z≠1 is calculated. z ,provides more accurate weight distribution for fault diagnosis; Step 5, by calculating the winding state coefficient Y i The winding state factor W is combined with the threshold K to determine whether the transformer winding state is abnormal, thus achieving comprehensive evaluation and diagnosis of the winding state; Step 6, by calculating the response error index F and combining it with the threshold G, it is determined whether the transformer winding has a turn-to-turn short circuit fault or a bulging warping fault, so as to achieve accurate diagnosis of the fault type.

3. The transformer winding fault diagnosis method according to claim 2, characterized in that: Step 1 includes the following steps: Excitation signal generation: Use a computer-controlled digital signal generator to generate standard excitation signals; Signal injection: The excitation signal is injected into the transformer winding through the high-voltage bushing, and flows through the high-voltage winding, medium-voltage winding and low-voltage winding in sequence; Signal output: The signal flows out of the transformer through the low-voltage bushing, and after being processed by the noise reduction filter module and the signal acquisition module, the computer measures the output response; The measurement result of step 1 is expressed as: Input Voltage: In the formula, It represents the input voltage signal at the zth tap position, which is a time function, indicating the change of input voltage over time; Indicates the amplitude of the input voltage at a certain moment; It represents the amplitude of the input voltage after stabilization under the z-th position tap; Output voltage: In the formula, It represents the output voltage signal at the zth tap position, which is also a time function, indicating the change of output voltage over time; Indicates the amplitude of the output voltage at a certain moment; It represents the amplitude of the output voltage after stabilization at the z-th position tap; The tap position is expressed as: Z=[12···m]; where z represents the number of the tap position; z=1 is the reference position, and the value of m depends on the voltage regulation method of the transformer.

4. The transformer winding fault diagnosis method according to claim 2, characterized in that: Step 2 includes the following steps: Fourier transform: Input voltage at different tap positions Output voltage Perform Fourier transform to obtain the frequency domain information of input voltage and output voltage Amplitude-frequency response calculation: Calculate the amplitude-frequency response curve based on frequency domain information: In the formula, A Z It is represented as the amplitude-frequency response curve when the tap position is z; z represents the number of the tap position; Data interception: intercept A Z The amplitude data of the intermediate frequency in the range of 1kHz-1MHz is recorded as: In the formula, each Indicates the amplitude corresponding to a certain frequency point in the frequency range of 1kHz-1MHz; The corresponding frequency is: [f1 f2···f n ], indicating that the amplitude data X Z The corresponding frequency points are in the range of 1kHz-1MHz; The output of step 2 is: Amplitude-frequency response curve A at each tap position z Z ; The clipped amplitude data X Z and its corresponding frequency points [f1 f2···f n ].

5. The transformer winding fault diagnosis method according to claim 2, characterized in that: In step 3, calculate the cumulative distance D Z The calculation formula is: Where D Z It is expressed as the cumulative distance when the tap position is Z and when the tap position is the reference; is the sum of the absolute differences between the amplitude data of the zth tap position and the reference position at each frequency point; is the amplitude data of the ith frequency point at the zth tap position; is the amplitude data of the ith frequency point at the reference position z=1; n is the total number of frequency points; is a combination of the maximum difference and the minimum sum of the amplitude data of the z-th tap position and the reference position at the last frequency point i=n; Indicates the maximum difference between the amplitude data of the zth tap position and the reference position at the last frequency point i=n; It represents the minimum sum of the amplitude data of the zth tap position and the reference position at the last frequency point i=n; z represents the number of the tap position; Z≠1 means the cumulative distance D Z The calculation is performed only for non-reference positions Z=2, 3, ..., m, and the reference position Z=1 is used as a reference point and does not participate in the calculation of its own difference.

6. The transformer winding fault diagnosis method according to claim 2, characterized in that: In step 4, the distance collaborative optimization weight w is calculated z The calculation formula is: In the formula, Var z It is expressed as the amplitude-frequency data X when the tap position is Z z The variance of z It is expressed as the distance collaborative optimization weight when the tap position is Z; It represents the sum of the variance of amplitude-frequency data at all tap positions; is the amplitude data of the ith frequency point at the zth tap position; is the amplitude data of the ith frequency point at the reference position z=1; f i represents the frequency value of the ith frequency point; D i represents the cumulative distance corresponding to the i-th frequency point; e () represents an exponential function, which is used to amplify the difference of frequency domain features; z represents the number of the tap position; Z≠1 represents the distance collaborative optimization weight w z The calculation is performed only for non-reference positions Z=2, 3, ..., m, and the reference position Z=1 is used as a reference point and does not participate in the weight calculation.

7. The transformer winding fault diagnosis method according to claim 2, characterized in that: In step 5, the winding state coefficient Y i The calculation formula is as follows: Where Y i The winding state coefficient of the i-th tap position; D Z It is expressed as the cumulative distance between the tap position Z and the tap position as the reference; max(D 1 ··· D Z ) represents the maximum value of the cumulative distance of all tap positions; min(D 1 ··· D Z ) represents the minimum value of the cumulative distance of all tap positions; It represents the sum of the square differences between the cumulative distances of all tap positions and the cumulative distances of the reference position; i represents the number of the tap position; n is the total number of taps; i=1 represents the reference position, and its winding state coefficient is fixed to 1, that is, Y1=1; i≠1 represents the non-reference position, and its winding state coefficient is calculated by the formula.

8. The transformer winding fault diagnosis method according to claim 2, characterized in that: In step 5, the calculation formula of the winding state factor W is as follows: Where W represents the winding condition factor, which is used to comprehensively evaluate the overall condition of the transformer winding; Y i represents the winding state coefficient of the i-th tap position; w i represents the distance collaborative optimization weight of the i-th tap position; Indicates the winding state coefficient Y i Co-optimize weight w with distance i The weighted sum of i 、sinY i Respectively represent the winding state coefficient Y i The cosine and sine of ; represents the sum of the cosine and sine values ​​of all tap positions; represents the sum of the distance co-optimization weights of all tap positions; ln() represents the natural logarithm function, which is used to adjust the scale and distribution of the data; e () represents an exponential function used to amplify the difference characteristics; n is the total number of taps.

9. The transformer winding fault diagnosis method according to claim 2, characterized in that: In step 5, the conditions for determining the winding state are: If W ≥ K, the transformer winding state is judged to be abnormal; If W<K, it is judged that the transformer winding state is normal; Where W represents the winding state factor, which is used to comprehensively evaluate the overall state of the transformer winding; K is a constant related to the transformer model, winding signal injection method, and external equipment.

10. The transformer winding fault diagnosis method according to claim 2, characterized in that: In step 6, the calculation formula of the response error index F is as follows: Where, F is the response error index, which is used to comprehensively evaluate the fault type and severity of the winding; Represents the amplitude data of the i-th frequency point at the z-th tap position; is the amplitude data of the ith frequency point at the reference position z=1; f i Represents the frequency value of the i-th frequency point; It represents the square difference of the amplitude data of the zth tap position and the reference position at the ith frequency point; represents the frequency weight factor, which is used to weight the amplitude difference according to the frequency value; W represents the winding state factor; ln(1+W) represents the natural logarithm of the winding state factor W plus 1, which is used to adjust the proportion and distribution of the data; w z represents the distance collaborative optimization weight of the z-th tap position; w1 represents the distance collaborative optimization weight of the first tap position; |w z -w1| represents the absolute difference of the distance co-optimization weight between the zth tap position and the 1st tap position; represents a double summation of all tap positions and frequency points; The conditions for determining the fault type are: If F≥G, it is judged that the transformer winding has a turn-to-turn short circuit fault; If F<G, it is judged that the transformer winding has a bulging and warping fault; Where G is a constant related to the transformer model, winding signal injection method, and external equipment.

Images