A method and system for integrated testing of transformer direct resistance and transformation ratio

Abstract

The present application belongs to the technical field of power equipment testing, and particularly relates to a transformer direct resistance and ratio integrated testing method and system, comprising the following steps: S101, injecting direct current test current to the winding of the transformer, after meeting the current stability criterion, synchronously collecting the direct current voltage at both ends of the winding and calculating the direct current resistance value, so as to realize transformer direct resistance testing; S102, connecting the low-voltage side winding of the transformer to the discharge circuit through a fixed discharge resistance, keeping the high-voltage side winding in an open circuit state, and synchronously and continuously collecting the high-voltage induced voltage of the high-voltage side and the low-voltage induced voltage of the low-voltage side in the controlled discharge process. The present application completes the ratio measurement by using the inherent characteristics of the electromagnetic induction law in the discharge process, eliminates the influence of the residual magnetism of the core on the ratio measurement from the physical principle, and synchronously outputs the direct current resistance value and the ratio result in the primary wiring and primary discharge process.

Description

Technical Field

[0001] This invention belongs to the field of power equipment testing technology, specifically relating to an integrated testing method and system for transformer DC resistance and turns ratio. Background Technology

[0002] Transformer winding DC resistance testing and turns ratio testing are mandatory inspection items for transformer acceptance, preventive testing, and fault diagnosis in power systems. The former is used to assess the contact quality of winding conductors and their connections, while the latter is used to verify the integrity of winding turns and the correctness of tap changer positions. With the continuous expansion of the power grid and the increasing demands for operation and maintenance efficiency, the need for on-site testing to shorten the testing time for a single transformer and simplify operating procedures is becoming increasingly urgent.

[0003] Currently, the commonly used testing method in engineering practice is a step-by-step independent test: first, a DC current is injected into the winding under test to complete the DC resistance test; after the test, a rapid short circuit is performed on the winding to release the stored energy; after the discharge is complete, a separate turns ratio tester is used to apply a low-amplitude AC excitation signal to the winding, and the turns ratio is obtained by measuring the ratio of the AC voltage on the high-voltage side to the low-voltage side. This step-by-step method treats the DC resistance test and the turns ratio test as two completely independent processes, requiring separate wiring, separate excitation, and separate data acquisition.

[0004] Patent application CN114487929A discloses an automatically switching transformer ratio and DC resistance measurement device and method. The device includes a storage and display unit, a PLC control and measurement unit, a transformer ratio measurement unit, a DC resistance measurement unit, and a transformer under test. The storage and display unit is connected to the PLC control and measurement unit to display and store measurement data. The PLC control and measurement unit is connected to the transformer ratio measurement unit, the DC resistance measurement unit, the low-voltage winding side contactor, and the high-voltage winding side contactor via transmission lines. The low-voltage winding side of the transformer under test is connected to the low-voltage ports of the transformer ratio measurement unit and the DC resistance measurement unit via a low-voltage winding contactor, and the high-voltage winding side of the transformer under test is connected to the high-voltage ports of the transformer ratio measurement unit and the DC resistance measurement unit via a high-voltage winding contactor.

[0005] However, the continuous injection of DC current causes the core to migrate along the magnetization curve towards the unidirectional deep magnetization region. After the rapid short-circuit discharge, a significant residual magnetic flux remains in the core. The low-amplitude AC excitation signal applied in the subsequent turns ratio test is insufficient to dissipate this residual magnetism. The core operates in an asymmetrical magnetization state, and the excitation current is distorted. This causes the ratio of the induced voltage on the high-voltage side to the low-voltage side to deviate from the true turns ratio, resulting in a systematic deviation in the turns ratio measurement results. To eliminate this deviation, operators have to insert a time-consuming demagnetization operation between the DC resistance test and the turns ratio test, or wait for the core to demagnetize naturally. This significantly increases the overall testing time for a single transformer and complicates the on-site operation process. Summary of the Invention

[0006] This invention provides an integrated testing method and system for transformer DC resistance and turns ratio, which solves the technical problems in the prior art where the residual magnetism of the iron core causes systematic deviations in the turns ratio measurement results, and the DC resistance test and turns ratio test need to be performed separately, resulting in long testing time and complex process.

[0007] In a first aspect, the present invention provides an integrated testing method for transformer DC resistance and turns ratio, comprising the following steps: S101 injects DC test current into the transformer windings. After the current stability criterion is met, the DC voltage across the windings is collected simultaneously and the DC resistance value is calculated to realize the DC resistance test of the transformer. S102 connects the low-voltage side winding of the transformer to the discharge circuit through a fixed discharge resistor, keeps the high-voltage side winding in an open circuit state, and synchronously and continuously collects the high-voltage induced voltage on the high-voltage side and the low-voltage induced voltage on the low-voltage side during the controlled discharge process. S103, calculate the real-time correlation coefficient between the high-voltage induced voltage and the low-voltage induced voltage, and calculate the real-time signal-to-noise ratio corresponding to the high-voltage induced voltage and the low-voltage induced voltage. S104, using real-time correlation coefficient and real-time signal-to-noise ratio to determine the effective start time and effective end time of the effective data set; S105, within the valid data set, calculate the ratio of high-voltage induced voltage to low-voltage induced voltage point by point to obtain the point-by-point transformation ratio sequence, and use the real-time correlation coefficient as the reliability weight to perform a weighted average calculation on the point-by-point transformation ratio sequence to obtain the final transformation ratio result.

[0008] By simultaneously calculating the real-time correlation coefficient and real-time signal-to-noise ratio during the controlled discharge process, the effective start and end times of the effective data set are determined jointly from two independent physical dimensions—signal source consistency and signal absolute strength—using dual criteria. This automatically shields the deviation data in the nonlinear section of the iron core during the initial discharge phase and the inferior data in the noise-dominated section during the final discharge phase. Within the effective data set, the real-time correlation coefficient is used as a reliability weight to perform a weighted average calculation on the point-by-point transformation ratio sequence, so that sampling times with a high degree of linear homology contribute more weight to the final transformation ratio result, further improving the transformation ratio calculation accuracy without shortening the effective measurement window.

[0009] Furthermore, the integrated testing method also includes a noise extraction step: Before injecting DC test current into the transformer windings, acquire the silence signals of the high-voltage acquisition channel and the low-voltage acquisition channel; The absolute values ​​of all sampling points of the silent signal are taken and the average value is calculated to obtain the high-voltage noise floor and the low-voltage noise floor, respectively.

[0010] Before injecting DC test current into the transformer windings, the silent signals of the high-voltage acquisition channel and the low-voltage acquisition channel are obtained in advance, and the high-voltage noise base and the low-voltage noise base are calculated. A quantitative noise reference benchmark is established at the very beginning of the test process, so that the calculation of the real-time signal-to-noise ratio during the subsequent controlled discharge process has fixed and reliable denominator data, avoiding mutual interference between noise estimation and signal acquisition, and ensuring the judgment accuracy of the signal-to-noise ratio criterion.

[0011] Furthermore, the calculation of the real-time signal-to-noise ratio values ​​corresponding to the high-voltage induced voltage and the low-voltage induced voltage includes: The short-time voltage amplitudes of the high-voltage induced voltage and the low-voltage induced voltage are calculated based on the first sliding window. Divide the short-time voltage amplitude of the high-voltage induced voltage by the high-voltage noise floor to obtain the high-voltage signal-to-noise ratio. Divide the short-time voltage amplitude of the low-voltage induced voltage by the low-voltage noise floor to obtain the low-voltage signal-to-noise ratio value; The high-voltage signal-to-noise ratio and the low-voltage signal-to-noise ratio together constitute the real-time signal-to-noise ratio.

[0012] Furthermore, calculating the real-time correlation coefficient between the high-voltage induced voltage and the low-voltage induced voltage includes: calculating the Pearson correlation between the high-voltage induced voltage and the low-voltage induced voltage using the same first sliding window, and using the Pearson correlation as the real-time correlation coefficient.

[0013] Furthermore, determining the effective start time of the valid data set includes: in response to the real-time correlation coefficient being continuously greater than the first reference threshold, determining the starting collection point of the first sliding window as the effective start time.

[0014] The real-time correlation coefficient is continuously greater than the first reference threshold as the trigger condition for the effective start time. By utilizing the physical characteristic that the real-time correlation coefficient increases from low to high when the core migrates from the nonlinear saturation region to the linear region, the starting boundary of the effective measurement window is adaptively determined without relying on preset fixed delay parameters. This makes the determination of the starting boundary directly related to the actual magnetization state of the transformer under test, and is suitable for transformers with different capacity levels and different core characteristics.

[0015] Furthermore, determining the effective termination time of the valid data set includes: after determining the effective start time, judging whether the real-time signal-to-noise ratio value drops to the next reference threshold, and judging whether the real-time correlation coefficient is continuously less than the first reference threshold; the earliest time that triggers the judgment condition is determined as the effective termination time.

[0016] Furthermore, before using the real-time correlation coefficient as a reliability weight to perform a weighted average calculation on the point-by-point variation sequence, the integrated testing method also includes a weight optimization step: performing non-negative truncation on the real-time correlation coefficient at each sampling time within the effective dataset to obtain the reliability weight corresponding to each sampling time.

[0017] The real-time correlation coefficients at each sampling time within the effective dataset are truncated to be used as reliability weights. This mathematically eliminates the negative correlation coefficient bias caused by random fluctuations in noise at the end of the discharge period, ensuring that the physical meaning of the reliability weights is always a positive contribution. This prevents the stability of the weighted average calculation results from being disturbed by occasional negative correlation data under extreme operating conditions.

[0018] Furthermore, before calculating the ratio of high-voltage induced voltage to low-voltage induced voltage point by point, the integrated testing method also includes a data verification step: Obtain the absolute value of the low-voltage induced voltage at each sampling time within the valid data set; In response to an absolute value less than the low-voltage noise floor, the corresponding sampling time is removed from the valid data set; Count the number of valid points in the valid dataset after the elimination operation; The ratio calculation process is terminated when the number of valid points is less than the preset lower limit.

[0019] Furthermore, the resistance value of the fixed discharge resistor is adjusted in stages based on the estimated inductance value of the winding, so that the discharge time constant of the controlled discharge process is within the preset duration range.

[0020] The value of the fixed discharge resistor is adjusted in stages based on the estimated inductance of the winding to keep the discharge time constant within the preset time range. This ensures that the controlled discharge process is slow enough to obtain a sufficient number of effective sampling points, but not too long to affect the efficiency of on-site testing. This solution can be adapted to transformer windings with different inductance parameters without requiring operators to manually select the discharge resistor value based on experience.

[0021] Secondly, the present invention provides an integrated testing system for transformer DC resistance and turns ratio, including a memory and a processor. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned integrated testing method for transformer DC resistance and turns ratio is realized.

[0022] The beneficial effects are as follows: This invention directly transforms the controlled discharge process after the transformer DC resistance test into a turns ratio measurement process. During the discharge, the high-voltage induced voltage and low-voltage induced voltage are collected synchronously and continuously. The turns ratio is calculated by utilizing the equivalent relationship between the ratio of the induced voltages on both sides and the turns ratio in the law of electromagnetic induction. There is no need to perform additional demagnetization operations or apply AC excitation signals. From a physical principle perspective, the influence of residual magnetism of the iron core on the systematic deviation of the turns ratio measurement is eliminated. The DC resistance value and the final turns ratio result are output synchronously within a single wiring and discharge process. The total test time is significantly shortened compared with the traditional step-by-step test method. The test process is simplified and no manual intervention by the operator is required for the data interception interval throughout the entire process. Attached Figure Description

[0023] Figure 1 This is a flowchart of an integrated testing method for transformer DC resistance and turns ratio.

[0024] Figure 2 This is a schematic diagram illustrating the changing trend of the dual criteria with controlled discharge time and the extraction of the effective data set. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] This solution is applied to on-site testing scenarios for transformer acceptance and preventative maintenance in power systems. Taking an oil-immersed power transformer with a rated capacity of 10 MVA and a rated voltage ratio of 110 kV and 10.5 kV as an example, this transformer is deployed in a 110 kV substation and requires regular winding DC resistance and turns ratio tests to assess the conductor connection quality and turn integrity. In traditional testing procedures, after the DC resistance test, a rapid short-circuit discharge is required to release the winding's stored energy. After discharge, a residual unidirectional magnetic flux remains in the core. The low-amplitude AC excitation signal applied for subsequent turns ratio tests is insufficient to dissipate this residual magnetism. The core operates in an asymmetrical magnetization state, leading to systematic deviations in the turns ratio measurement results. Operators must perform time-consuming demagnetization or wait for the core to naturally demagnetize before conducting the turns ratio test, resulting in a lengthy and complex testing process. This solution directly transforms the necessary discharge process after the DC resistance test into a turns ratio measurement process, simultaneously completing the DC resistance and turns ratio tests within a single connection and discharge operation, eliminating the need for additional demagnetization. The following describes the specific implementation process of each step in the above transformer testing scenario.

[0027] An embodiment of the integrated testing method for transformer DC resistance and turns ratio provided by this invention: like Figure 1 As shown, the integrated testing method for transformer DC resistance and turns ratio includes the following steps: S101, inject DC current, and use current stabilization and voltage measurement to calculate the DC resistance of the transformer winding.

[0028] A DC test current is injected into the transformer winding. After the current stability criterion is met, the DC voltage across the winding is collected simultaneously and the DC resistance value is calculated to realize the DC resistance test of the transformer.

[0029] In one embodiment, the tester establishes electrical connections with the high-voltage side winding terminals and the low-voltage side winding terminals of the transformer respectively through test leads. The hardware configurations of the high-voltage acquisition channel and the low-voltage acquisition channel are consistent, the sampling rate is set to 1 kHz, and the range is pre-selected based on the rated parameters of the winding under test.

[0030] Before injecting DC test current into the transformer windings, a noise extraction step is first performed to obtain the noise reference standard required for subsequent turns ratio measurements. Specifically, with the test instrument and transformer acquisition circuits connected but without any electrical excitation applied, the high-voltage and low-voltage acquisition channels simultaneously acquire a silent signal. The acquisition duration of this silent signal is set to 2 seconds, corresponding to 2000 sampling points at a 1 kHz sampling rate. The background noise of the acquisition system consists of thermal noise from the acquisition hardware, quantization errors from analog-to-digital conversion, and electromagnetic interference from the field environment. The intensity of these interference sources remains stable during a single test; therefore, the noise level reflected by the silent signal can be used as a fixed reference quantity throughout the entire test. Taking the aforementioned 110 kV substation as an example, the 2-second acquisition duration is exactly an integer multiple of the 20 millisecond power frequency cycle. The power frequency interference component is automatically canceled out during the averaging calculation, requiring no additional filtering.

[0031] The absolute values ​​of all sampling points of the acquired silent signal are taken and averaged. The high-voltage acquisition channel yields the high-voltage noise floor, and the low-voltage acquisition channel yields the low-voltage noise floor. In the actual measurement of the aforementioned 110 kV transformer, the typical value of the high-voltage noise floor is 0.23 mV, and the typical value of the low-voltage noise floor is 0.18 mV. Both are in the sub-millivolt range, reflecting the background noise level of the acquisition system in this field environment. The high-voltage and low-voltage noise floors will be used as the fixed denominator reference for signal-to-noise ratio calculation in subsequent steps, and will be multiplexed throughout the entire process after a single acquisition.

[0032] After completing the noise extraction step, the tester injects a stable DC test current into the winding under test. The current amplitude is selected based on the rated current of the winding, typically ranging from 10% to 40% of the rated current. For the aforementioned transformer with a rated capacity of 10 MVA, the rated current on the high-voltage side is approximately 52.5 amps. The DC test current is taken as 15% of the rated current, i.e., 7.88 amps. This value balances the measurement signal-to-noise ratio with winding heating control. If the current is too small, the DC voltage across the winding will drop into the noise range; if the current is too large, prolonged injection will cause the winding temperature to rise, affecting the accuracy of the resistance measurement.

[0033] After the DC test current is injected, due to the winding inductance, the current needs to undergo a rising transition process before reaching a steady state. The current stability criterion is set as the change in current reading not exceeding 0.1% of the reading value within 3 consecutive seconds. This criterion is formulated with reference to the stability requirements of transformer winding DC resistance testing in the power industry standard. After the current stability criterion is met, the tester synchronously acquires the DC voltage across the winding. According to Ohm's law, the DC voltage is divided by the DC test current to calculate the DC resistance value, thus completing the transformer DC resistance test. In the above actual test scenario, the measured value of the DC resistance of the high-voltage side winding is 1.263 ohms.

[0034] It should be noted that during the continuous injection of DC test current, the iron core migrates along the magnetization curve towards the unidirectional deep magnetization region, and by the end of the DC resistance test, the iron core has accumulated significant residual magnetic flux. This residual magnetism is an inevitable physical result of DC excitation. This scheme does not perform any active demagnetization operation, but instead utilizes the discharge process itself to complete the turns ratio measurement in subsequent steps, thus bypassing the influence of residual magnetism on the turns ratio measurement from the perspective of the law of electromagnetic induction.

[0035] Thus, the noise extraction step establishes a quantitative noise reference benchmark at the very beginning of the testing process, providing reliable denominator data for the calculation of the real-time signal-to-noise ratio during the subsequent controlled discharge process; the DC resistance test completes the acquisition and calculation after the current is fully stabilized, ensuring the measurement accuracy of the DC resistance value, and at the same time reserving the core magnetic energy for the subsequent controlled discharge process.

[0036] S102, low-voltage resistance discharge, open-circuit high-voltage side synchronously collects induced voltage on both sides.

[0037] The low-voltage side winding of the transformer is connected to the discharge circuit through a fixed bleed resistor, while the high-voltage side winding is kept in an open circuit state. During the controlled discharge process, the high-voltage induced voltage on the high-voltage side and the low-voltage induced voltage on the low-voltage side are collected synchronously and continuously.

[0038] After the DC resistance test in S101 is completed, DC test current still flows in the winding, and the iron core is in a unidirectional magnetized state, storing a considerable amount of magnetic field energy. In order to release this stored energy in a controlled manner and simultaneously obtain the induced voltage signal required for the turns ratio measurement, the tester immediately performs a discharge circuit switching operation.

[0039] In one embodiment, the tester connects the low-voltage winding of the transformer to the discharge circuit via a fixed bleed resistor, while keeping the high-voltage winding in an open-circuit state. The physical basis for this connection is that when the high-voltage winding is in an open-circuit state, its terminal voltage is the open-circuit induced electromotive force (EMF), excluding the DC resistance voltage drop and leakage reactance voltage drop, which has the highest consistency with the theoretical induced EMF given by the law of electromagnetic induction. After the low-voltage winding is connected to the discharge circuit, the discharge current flows through the low-voltage winding, and the low-voltage terminal voltage is the induced EMF minus the difference between the leakage reactance voltage drop and the winding resistance voltage drop. However, within the effective measurement range where the discharge current amplitude is small, the leakage reactance voltage drop relative to the induced EMF is typically less than 0.1%, and can be ignored. Therefore, within the effective measurement range, the ratio of the high-voltage induced voltage to the low-voltage induced voltage is approximately equal to the ratio of the number of turns on the high-voltage side to the number of turns on the low-voltage side, i.e., the true turns ratio of the transformer. This approximate relationship is independent of the absolute value of the current residual magnetism of the core, which is a direct derivation of the law of electromagnetic induction and constitutes the physical basis for this scheme to measure the turns ratio using a controlled discharge process.

[0040] The value of the fixed bleeder resistor is adjusted in stages based on the estimated inductance of the winding to ensure that the discharge time constant of the controlled discharge process is within a preset range. The discharge time constant is equal to the winding inductance divided by the value of the fixed bleeder resistor, with the preset range set from 0.5 seconds to 5 seconds. The lower limit of 0.5 seconds ensures that the controlled discharge process is slow enough to acquire at least several hundred sampling points within the effective measurement interval at a 1 kHz sampling rate, meeting the data requirements for subsequent statistical calculations; the upper limit of 5 seconds prevents the controlled discharge process from becoming too long, which would reduce the efficiency of on-site testing.

[0041] In the actual test scenario of the aforementioned 110 kV transformer, the estimated inductance of the low-voltage winding was approximately 12 Henry. The step-by-step adjustment process is as follows: Initially, a fixed bleed resistor of 100 ohms was selected. At this point, the discharge time constant was 12 Henry divided by 100 ohms, which equaled 0.12 seconds, less than the lower limit of the preset time range of 0.5 seconds, indicating excessively rapid discharge. The resistance value was then progressively reduced. At 50 ohms, the discharge time constant was 0.24 seconds, still insufficient. At 25 ohms, the discharge time constant was 0.48 seconds, still slightly below the lower limit. At 10 ohms, the discharge time constant was 1.2 seconds, falling within the preset time range of 0.5 to 5 seconds. Therefore, the final fixed bleed resistor value was determined to be 10 ohms. This step-by-step adjustment mechanism allows this solution to adapt to transformers of different capacity levels and inductance parameters, eliminating the need for operators to manually select the fixed bleed resistor based on experience.

[0042] After the discharge circuit switching is completed, the tester disconnects the DC current source, and the energy stored in the winding begins to be released in a controlled manner through the fixed discharge resistor. During the discharge, the core magnetic flux monotonically decays from the residual magnetism level to zero, and the rate of change of magnetic flux remains non-zero, causing synchronous induced voltages in both the high-voltage and low-voltage windings. During the controlled discharge process, the tester synchronously and continuously acquires the high-voltage induced voltage on the high-voltage side and the low-voltage induced voltage on the low-voltage side using the same 1 kHz sampling rate and hardware configuration as in the S101 noise extraction step. The two channels share the same sampling clock trigger to ensure that the high-voltage and low-voltage induced voltages are strictly aligned at each sampling moment. Acquisition begins at the start of the discharge and continues until the discharge current drops to less than 1% of its initial value, completely recording the timing sequence of the high-voltage and low-voltage induced voltages throughout the entire controlled discharge process without any truncation, for subsequent processing.

[0043] Under the experimental conditions of a fixed 10-ohm discharge resistor and a 1.2-second discharge time constant, the controlled discharge process lasted approximately 5.8 seconds, with a total of 5800 sampling points collected. At the initial stage of discharge, the peak value of the high-voltage induced voltage was approximately 86 volts, and the peak value of the low-voltage induced voltage was approximately 8.2 volts, both exhibiting an exponential decay trend over time. Towards the end of the discharge, the amplitudes of the induced voltages on both sides gradually decreased to the millivolt level, eventually being submerged by the background noise of the acquisition system. The timing sequences of the high-voltage and low-voltage induced voltages throughout the controlled discharge process constitute the raw data input for subsequent real-time correlation coefficient and real-time signal-to-noise ratio calculations.

[0044] By connecting the low-voltage side winding to the discharge circuit through a fixed bleed resistor and keeping the high-voltage side winding open, the high-voltage side terminal voltage directly reflects the product of the core flux change rate and the number of high-voltage side turns during controlled discharge. The low-voltage side terminal voltage approximately reflects the product of the flux change rate and the number of low-voltage side turns within the effective range. The two signals are driven by the same physical quantity and differ only by a constant multiple, providing sufficient raw data for calculating the turns ratio using dual-channel signals in subsequent steps.

[0045] S103, calculate the real-time correlation coefficient and signal-to-noise ratio of the induced voltages on both sides.

[0046] Calculate the real-time correlation coefficient between the high-voltage induced voltage and the low-voltage induced voltage, and calculate the real-time signal-to-noise ratio values ​​corresponding to the high-voltage induced voltage and the low-voltage induced voltage.

[0047] Based on the high-voltage induced voltage timing and low-voltage induced voltage timing obtained in S102, this step performs sliding window analysis on the two signals and simultaneously calculates the real-time correlation coefficient and real-time signal-to-noise ratio, providing dual quantitative indicators for the subsequent determination of the effective data set boundary.

[0048] The length of the first sliding window is set to 20 milliseconds, corresponding to 20 sampling points at a sampling rate of 1 kHz, with a window step size of 1 sampling point. Starting from the discharge initiation moment, the first sliding window slides along the time axis point by point, with each step performing a calculation on the 20 sampling points within the window, until the entire controlled discharge process is covered. The selection of the first sliding window length is based on the following criteria: if the window is too short, the number of sampling points within the window is insufficient to support stable calculation of the statistics; if the window is too long, the time resolution decreases, making it impossible to capture the inflection points of signal quality changes in a timely manner. Based on the fitting of measured data from controlled discharge processes of transformers of different capacity levels, when the discharge time constant is in the range of 0.5 seconds to 5 seconds, the signal change amplitude corresponding to a 20-millisecond window length does not exceed 4% of the mean within the window, satisfying the short-time stationarity assumption; simultaneously, the 20 sampling points provide sufficient degrees of freedom for Pearson correlation calculation. If a special condition occurs where the discharge time constant is less than 200 milliseconds, the length of the first sliding window can be shortened to 10 milliseconds to ensure a sufficient number of independent windows within the effective period.

[0049] The short-time voltage amplitudes of the high-voltage induced voltage and the low-voltage induced voltage are calculated separately based on the first sliding window. At each window position, the arithmetic mean of the absolute values ​​of all sampling points of the high-voltage induced voltage within the window is calculated to obtain the short-time voltage amplitude of the high-voltage induced voltage corresponding to that window position. The same absolute value averaging operation is performed on all sampling points of the low-voltage induced voltage within the window to obtain the short-time voltage amplitude of the low-voltage induced voltage. As the window slides point by point along the time axis, the short-time voltage amplitudes of the high-voltage induced voltage and the low-voltage induced voltage each form a time-varying amplitude envelope sequence, reflecting the instantaneous intensity evolution trend of the two induced voltage signals.

[0050] The high-voltage signal-to-noise ratio (SNR) is obtained by dividing the short-time amplitude of the high-voltage induced voltage by the high-voltage noise floor obtained in S101; the low-voltage SNR is obtained by dividing the short-time amplitude of the low-voltage induced voltage by the low-voltage noise floor obtained in S101. The high-voltage and low-voltage SNR values ​​together constitute the real-time SNR. In the above-mentioned test scenario of the 110 kV transformer, both the high-voltage and low-voltage SNR values ​​are much higher than the noise level in the early stage of discharge. As the controlled discharge process progresses, the amplitudes of the two induced voltages continuously decay, and the high-voltage and low-voltage SNR values ​​decrease synchronously. When the discharge enters the final stage, the real-time SNR gradually approaches and eventually falls below the decision threshold, indicating that the signal has been submerged by noise and is no longer suitable for transformer ratio calculation.

[0051] The physical meaning of the real-time signal-to-noise ratio (SNR) is that it represents the strength advantage of the induced voltage signal relative to the background noise of the acquisition system at the current moment. A larger value indicates a higher degree of signal dominance, and less noise contamination in the point-by-point transformation ratio calculation at that moment; a smaller value indicates a larger proportion of noise components in the signal, and more significant random errors in the transformation ratio calculation results. By dividing the short-time voltage amplitude by the pre-calibrated high-voltage and low-voltage noise bases in S101, the real-time SNR transforms the absolute voltage amplitude into a dimensionless index relative to the noise level, ensuring uniform comparability of signal quality across different ranges and field environments.

[0052] The Pearson correlation between the high-voltage induced voltage and the low-voltage induced voltage is calculated using the same first sliding window, and this Pearson correlation is used as the real-time correlation coefficient. At each window position, 20 sampling points of the high-voltage induced voltage and 20 sampling points of the low-voltage induced voltage within the window are used as one set of data, and the normalized linear correlation between the two sets of data is calculated using the standard Pearson correlation coefficient. The resulting value is the Pearson correlation corresponding to that window position, ranging from -1 to 1. This Pearson correlation is used as the real-time correlation coefficient at that moment. As the window slides point by point, the real-time correlation coefficient forms a time-varying sequence, reflecting the degree of linear homology between the high-voltage and low-voltage induced voltages at each moment.

[0053] In the above-mentioned test scenario, the real-time correlation coefficient exhibits three characteristics: In the initial stage of discharge, within approximately 0.08 seconds, the core operating point is in the nonlinear saturation region of the magnetization curve. The voltage waveforms on the high-voltage side and the low-voltage side are affected by the difference in nonlinear excitation impedance, and linear homology has not yet been fully established. The real-time correlation coefficient fluctuates between 0.6 and 0.85. Subsequently, the core operating point migrates to the approximately linear region of the magnetization curve. The two induced voltages are linearly driven by the same flux change rate, and the real-time correlation coefficient rapidly rises to above 0.95 and remains stable. In the final stage of discharge, the amplitude of the induced voltage decays to near the noise level. The two signals are dominated by independent random noise from their respective acquisition channels, and the real-time correlation coefficient gradually decreases from 0.95, eventually fluctuating randomly around 0.

[0054] The physical meaning of the real-time correlation coefficient lies in the fact that both the high-voltage and low-voltage induced voltages are driven by the same core flux change rate. Under ideal linear coupling conditions, the two signals differ only by a constant multiple, i.e., the turns ratio, and the theoretical value of the Pearson correlation is strictly equal to 1. The degree to which the real-time correlation coefficient deviates from 1 directly reflects the degree to which the two signals deviate from the ideal linear homology at the current moment. That is, the smaller the deviation, the closer the point-by-point turns ratio calculation result at that moment is to the true turns ratio; the larger the deviation, the more serious the contamination of the turns ratio calculation by nonlinear factors or noise interference. Therefore, the real-time correlation coefficient simultaneously carries the discrimination information of the effective window's start and end boundaries and can be directly used as a quantitative weight for the reliability of data at each sampling moment in subsequent steps.

[0055] The same first sliding window is used to synchronously calculate the real-time signal-to-noise ratio (SNR) and real-time correlation coefficient, ensuring strict alignment of the two indicators on the time axis. This provides time-synchronized quantitative input for the joint decision-making of the dual criteria in subsequent steps. The real-time SNR measures data quality from the perspective of absolute signal strength, while the real-time correlation coefficient measures data quality from the perspective of consistency between the two channels. Both assess the quality of the same set of sampled data from independent physical perspectives. Misjudgments caused by occasional interference in either single indicator can be constrained by the other indicator, forming a complementary dual quality evaluation system.

[0056] S104 determines the start and end times of valid data based on the coefficients and signal-to-noise ratio.

[0057] The effective start and end times of the effective dataset are determined by using the real-time correlation coefficient and the real-time signal-to-noise ratio.

[0058] Based on the real-time correlation coefficient sequence and real-time signal-to-noise ratio sequence calculated in S103, this step uses dual criteria to jointly determine the time boundary of the effective data set and select the data segments whose signal quality meets the accuracy requirements of the ratio calculation throughout the controlled discharge process.

[0059] The process for determining the effective start time is as follows: Starting from the discharge start time, the real-time correlation coefficient is checked window by window. When the real-time correlation coefficient is continuously greater than the first reference threshold, the starting point of the window that first meets this condition is determined as the effective start time. The value of the first reference threshold is determined by fitting measured data from controlled discharge processes of multiple transformers of different capacity levels: with a turns ratio calculation error not exceeding 0.5% as the constraint target, the relationship between the real-time correlation coefficient and the corresponding point-by-point turns ratio deviation is statistically analyzed. The fitting results show that when the real-time correlation coefficient is higher than 0.92, the 95% confidence interval of the point-by-point turns ratio deviation falls within 0.5%. Considering engineering margin, the first reference threshold is set to 0.92. The number of consecutive windows is set to 3, meaning the real-time correlation coefficient must be greater than 0.92 for 3 consecutive first sliding windows to confirm the effective start time. The requirement of 3 consecutive windows effectively suppresses the false triggering caused by occasional electromagnetic interference within a single window, which leads to a momentary increase in the correlation coefficient. In other words, a single momentary interference is not repeatable in adjacent windows, making it impossible to meet the threshold condition for 3 consecutive windows.

[0060] Once the effective start time is determined, it is fixed and will not be updated. The basis of this design is that the migration of the iron core from the nonlinear saturation region to the linear region during the controlled discharge process is a monotonous and irreversible physical process. Once the iron core enters the linear region, its operating point will not spontaneously retreat to the saturation region. Therefore, the overall trend of the real-time correlation coefficient after the effective start time is to remain at a high level until the signal decays at the end of the discharge. There is no physical scenario that requires resetting the start time.

[0061] In the aforementioned test scenario of the 110 kV transformer, the real-time correlation coefficient fluctuated between 0.63 and 0.87 within the first 0.06 seconds after the discharge began, reflecting that the core was still in the nonlinear region. From 0.07 seconds onwards, the real-time correlation coefficient rose to 0.93 and then remained above 0.92. The starting time of the third consecutive window that met the conditions was 0.11 seconds, therefore the effective starting time was determined to be 0.11 seconds. Although the discharge data before this time included induced voltage signals, due to the nonlinear effect of the core, the point-by-point turns ratio calculation results deviated from the true turns ratio and were not included in subsequent turns ratio calculations.

[0062] After the valid start time is determined, the logic for determining the valid end time is immediately initiated. Before the valid start time is determined, the termination criterion is suspended and not calculated to avoid false triggering when the signal is not yet stable in the early stage of discharge. The determination of the valid end time is based on the combined effect of two independent judgment conditions: judging whether the real-time signal-to-noise ratio value has dropped to the next reference threshold, and judging whether the real-time correlation coefficient is continuously less than the first reference threshold; the earliest time that triggers the judgment condition is determined as the valid end time.

[0063] The value of the second reference threshold was also determined through data fitting: In the measured data, the relationship between the real-time signal-to-noise ratio (SNR) value and the random error of the point-by-point ratio calculation at the corresponding time was statistically analyzed. When either the high-voltage or low-voltage SNR value first drops below 3, the single-point random error of the point-by-point ratio calculation is approximately 33%. Although subsequent weighted average calculation can compress the statistical error, in data segments where the real-time SNR value is below 3, the noise component already accounts for a considerable proportion of the signal energy. Continuing to include it in the calculation will significantly increase the uncertainty of the weighted average calculation result. Therefore, the second reference threshold is set to 3, that is, the moment when either the high-voltage or low-voltage SNR value first drops below 3, is recorded as the SNR criterion trigger moment.

[0064] The termination criterion for the real-time correlation coefficient adopts a logical structure symmetrical to the starting criterion: when the real-time correlation coefficient is less than the first reference threshold of 0.92 for three consecutive first sliding windows, the time corresponding to the starting acquisition point of the first window is recorded as the correlation coefficient criterion trigger time. The requirement of three consecutive windows is also used to suppress false alarms caused by occasional interference.

[0065] The effective termination time is the earlier of the signal-to-noise ratio (SNR) criterion trigger time and the correlation coefficient criterion trigger time. The logic behind choosing the earlier one is that the two criteria monitor data quality from two independent dimensions: absolute signal strength and dual-channel consistency. If either criterion triggers first, it means that the data after that moment no longer meets the accuracy requirements for the transformation ratio calculation, and the effective window should be closed immediately. If the SNR criterion is not triggered before the end of the discharge, the effective termination time is determined solely by the correlation coefficient criterion; if the correlation coefficient criterion is not triggered, the effective termination time is determined solely by the SNR criterion; if neither is triggered at the end of the discharge, the effective termination time is the end of the discharge, indicating that the signal quality meets the requirements throughout the controlled discharge process.

[0066] In the actual measurements of the aforementioned 110 kV transformer, the effective start time was determined to be 0.11 seconds after the termination criterion was initiated. During the controlled discharge process, the real-time correlation coefficient remained between 0.94 and 0.98 within the range of 0.11 seconds to 4.67 seconds, without triggering the correlation coefficient termination criterion. The high-voltage signal-to-noise ratio (SNR) dropped to 2.8 at 4.53 seconds, triggering the SNR criterion at 4.53 seconds. Therefore, the effective termination time was determined to be 4.53 seconds. The effective data set covers the time interval from 0.11 seconds to 4.53 seconds, containing a total of 4420 sampling points.

[0067] The effective start time and effective end time together define the time range of the effective data set. The high-voltage induced voltage, low-voltage induced voltage, and the corresponding real-time correlation coefficient within this set will serve as the complete input for subsequent turns ratio calculations. Dual criteria safeguard the boundaries of the effective data set from two independent physical dimensions: signal strength and signal homology. The start boundary automatically filters out deviation data from the nonlinear segment of the core during the initial discharge phase, while the end boundary automatically cuts off inferior data from the noise-dominated segment at the end of the discharge phase. This ensures that the data segment included in the turns ratio calculation meets accuracy requirements throughout the entire time range, eliminating the need for operators to manually set the truncation interval based on experience.

[0068] S105, weighted average turns ratio sequence, used to calculate the final turns ratio of the transformer.

[0069] Within the valid dataset, the ratio of high-voltage induced voltage to low-voltage induced voltage is calculated point by point to obtain a point-by-point transformation ratio sequence. The real-time correlation coefficient is then used as a reliability weight to perform a weighted average calculation on the point-by-point transformation ratio sequence to obtain the final transformation ratio result.

[0070] The valid start and end times determined in S104 define the time range of the valid data set. This step completes the final calculation of the ratio within this set. Before performing the point-by-point ratio calculation, the valid data set is first validated to remove sampling times where the denominator is too small, which may lead to abnormal amplification of the ratio calculation. Then, the correlation coefficient-weighted ratio calculation is performed on the validated data.

[0071] The data verification process is as follows: The absolute values ​​of the low-voltage induced voltage at each sampling moment within the valid data set are obtained and compared point-by-point with the low-voltage noise floor. If, at a given sampling moment, the absolute value of the low-voltage induced voltage is less than the low-voltage noise floor, it indicates that the low-voltage signal at that moment has been completely submerged by noise. If the low-voltage induced voltage at that moment is used as the divisor for ratio calculation, the result will be dominated by random noise values ​​and lose its physical meaning. Therefore, the corresponding sampling moment is removed from the valid data set. In the above-mentioned 110 kV transformer test, the valid data set covers 4420 sampling points from 0.11 seconds to 4.53 seconds. The low-voltage noise floor is 0.18 mV. After point-by-point comparison, the absolute values ​​of the low-voltage induced voltage at 7 sampling moments within the 4.51-4.53 second interval are lower than 0.18 mV and are removed, leaving 4413 valid sampling points.

[0072] The system counts the number of valid points in the effective data set after the elimination operation. If the number of valid points is less than a preset lower limit, the ratio calculation process terminates. The preset lower limit is determined based on statistical convergence requirements: the statistical standard error of the weighted average calculation is inversely proportional to the square root of the number of valid points. Based on the fit of measured data, when the number of valid points reaches 50, the standard error of the weighted average calculation drops to within 0.1% of the true value of the ratio, meeting the engineering requirement that the ratio measurement accuracy does not exceed 0.5%. Therefore, the preset lower limit is set to 50. In the above measured data, the number of valid points was 4413, far exceeding the preset lower limit. The data verification passed, and subsequent calculations continued. If the number of valid points is less than 50, the final ratio result will not be output, but the measured DC resistance value will still be output separately. It is recommended that the operator increase the fixed discharge resistor to extend the discharge time constant and then retest.

[0073] After data verification, within the valid data set after removing invalid points, the ratio of high-voltage induced voltage to low-voltage induced voltage is calculated point by point to obtain a point-by-point transformation ratio sequence. For each valid sampling time, the high-voltage induced voltage at that time is divided by the low-voltage induced voltage at the same time, and the quotient is the point-by-point transformation ratio at that time. The point-by-point transformation ratios of all valid sampling times are arranged in chronological order to form a point-by-point transformation ratio sequence. In the above actual measurement, the point-by-point transformation ratio sequence contains 4413 values, each fluctuating narrowly between 10.47 and 10.49, with the central value close to the transformer's nameplate transformation ratio of 10.476.

[0074] Before using the real-time correlation coefficient as a reliability weight to calculate the weighted average of the point-by-point ratio sequence, the real-time correlation coefficients at each sampling time within the effective dataset are truncated to obtain the reliability weights corresponding to each sampling time. The specific operation of the non-negative truncation is as follows: The real-time correlation coefficient at each sampling time within the effective dataset is checked point by point. If the real-time correlation coefficient at a certain time is less than 0, the reliability weight at that time is reset to 0; if the real-time correlation coefficient is greater than or equal to 0, the reliability weight is directly taken from the real-time correlation coefficient value at that time. The basis for performing the non-negative truncation is that the physical meaning of the real-time correlation coefficient is the degree of linear homology between the high-voltage induced voltage and the low-voltage induced voltage at that time. A positive value, and the closer it is to 1, the stronger the homology, and the more reliable the point-by-point ratio at that time; a negative value indicates that the two signals exhibit an inverse relationship, which should not occur during normal controlled discharge, but only occurs occasionally due to random fluctuations when noise completely dominates. The point-by-point ratio at such times has no physical meaning, and its weight should be 0 rather than negative. Allowing negative weights to participate in the weighted average calculation would cause the weighted average result to deviate from the true ratio. Non-negative truncation mathematically eliminates this risk. In the above-mentioned experiments, the real-time correlation coefficients at all 4413 sampling times within the effective dataset were between 0.93 and 0.98, with no negative values. Non-negative truncation did not change the weight values ​​at any time, but this step was always executed as a protective mechanism to deal with abnormal situations under extreme conditions.

[0075] After non-negative truncation, the reliability weights corresponding to each sampling time are applied to the point-by-point ratio sequence, and a weighted average is calculated to obtain the final ratio result. The weighted average is calculated as follows: the point-by-point ratio of each valid sampling time is multiplied by the reliability weight for that time, the sum is calculated over all valid sampling times, and then divided by the sum of the reliability weights for all valid sampling times. The quotient is the final ratio result. This calculation method ensures that sampling times with higher real-time correlation coefficients have a greater impact on the final ratio result, while sampling times with lower real-time correlation coefficients have a smaller impact, thus making the final ratio result more concentrated in reflecting the measurements of high-quality data segments.

[0076] The physical basis for using reliability weight as a weighting factor is that the higher the real-time correlation coefficient at a certain sampling moment, the stronger the linear homology between the high-voltage induced voltage and the low-voltage induced voltage at that moment, that is, the more fully the core flux is coupled to the windings on both sides and the smaller the external interference component. The point-by-point transformation ratio at that moment is more representative of the true turns ratio and should occupy a larger weight share in the final transformation ratio result.

[0077] In contrast, if the turns ratio is calculated using the equal-weighted average method, sampling points near the effective termination time and sampling points near the effective start time contribute the same influence, and the contamination of the mean by low-quality data is unavoidable. In the above-mentioned actual test, the final turns ratio calculated using the reliability-weighted average was 10.477, while the result calculated using the equal-weighted average was 10.479. Compared with the equal-weighted average, the weighted average calculation can reduce the turns ratio calculation error by 0.1% to 0.2% under conditions where the signal quality gradient within the effective window is steep. This has practical engineering value for transformer testing with strict accuracy requirements.

[0078] In the complete test procedure of the 110 kV transformer described above, steps S101 to S105 were all completed within one wiring and one discharge cycle. The measured DC resistance value was 1.263 ohms, and the final turns ratio result was 10.477, with a relative deviation of 0.01% from the nameplate turns ratio of 10.476, meeting the accuracy requirement of a turns ratio error not exceeding 0.5%. The overall test took approximately 53 seconds. Compared to the traditional step-by-step test procedure, which involves additional demagnetization after DC resistance testing and then a separate turns ratio test, the total test time was significantly shortened. Furthermore, no manual intervention from operators was required to extract data or assess signal quality throughout the entire process. The dual criteria and reliability weighting mechanism automatically ensured the accuracy and reliability of the final turns ratio result.

[0079] like Figure 2 As shown, the real-time correlation coefficient rises above the first reference threshold after experiencing nonlinear fluctuations in the initial stage of discharge, triggering an effective start time. As the controlled discharge process progresses, the real-time signal-to-noise ratio continuously decays exponentially, and when it drops to the second reference threshold, an effective termination time is triggered. The dual criteria jointly define the effective data set interval for subsequent ratio calculations, automatically shielding the nonlinear segment in the initial stage of discharge and the noise-dominated segment in the final stage of discharge.

[0080] An embodiment of the integrated testing system for transformer DC resistance and turns ratio provided by this invention: The integrated testing system for transformer DC resistance and turns ratio includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the aforementioned integrated testing method for transformer DC resistance and turns ratio.

[0081] The integrated testing system for transformer DC resistance and turns ratio also includes other components well known to those skilled in the art, such as communication interfaces. Their setup and functions are known in the art and will not be described in detail here.

[0082] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. An integrated testing method for transformer DC resistance and turns ratio, characterized in that, Includes the following steps: S101 injects DC test current into the transformer windings. After the current stability criterion is met, the DC voltage across the windings is collected simultaneously and the DC resistance value is calculated to realize the DC resistance test of the transformer. S102 connects the low-voltage side winding of the transformer to the discharge circuit through a fixed discharge resistor, keeps the high-voltage side winding in an open circuit state, and synchronously and continuously collects the high-voltage induced voltage on the high-voltage side and the low-voltage induced voltage on the low-voltage side during the controlled discharge process. S103, calculate the real-time correlation coefficient between the high-voltage induced voltage and the low-voltage induced voltage, and calculate the real-time signal-to-noise ratio corresponding to the high-voltage induced voltage and the low-voltage induced voltage. S104, using real-time correlation coefficient and real-time signal-to-noise ratio to determine the effective start time and effective end time of the effective data set; S105, within the valid data set, calculate the ratio of high-voltage induced voltage to low-voltage induced voltage point by point to obtain the point-by-point transformation ratio sequence, and use the real-time correlation coefficient as the reliability weight to perform a weighted average calculation on the point-by-point transformation ratio sequence to obtain the final transformation ratio result.

2. The integrated testing method for transformer DC resistance and turns ratio according to claim 1, characterized in that, The integrated testing method also includes a noise extraction step: Before injecting DC test current into the transformer windings, acquire the silence signals of the high-voltage acquisition channel and the low-voltage acquisition channel; The absolute values ​​of all sampling points of the silent signal are taken and the average value is calculated to obtain the high-voltage noise floor and the low-voltage noise floor, respectively.

3. The integrated testing method for transformer DC resistance and turns ratio according to claim 2, characterized in that, The calculation of the real-time signal-to-noise ratio values ​​corresponding to the high-voltage induced voltage and the low-voltage induced voltage includes: The short-time voltage amplitudes of the high-voltage induced voltage and the low-voltage induced voltage are calculated based on the first sliding window. Divide the short-time voltage amplitude of the high-voltage induced voltage by the high-voltage noise floor to obtain the high-voltage signal-to-noise ratio. Divide the short-time voltage amplitude of the low-voltage induced voltage by the low-voltage noise floor to obtain the low-voltage signal-to-noise ratio value; The high-voltage signal-to-noise ratio and the low-voltage signal-to-noise ratio together constitute the real-time signal-to-noise ratio.

4. The integrated testing method for transformer DC resistance and turns ratio according to claim 3, characterized in that, The calculation of the real-time correlation coefficient between the high-voltage induced voltage and the low-voltage induced voltage includes: calculating the Pearson correlation between the high-voltage induced voltage and the low-voltage induced voltage using the same first sliding window, and using the Pearson correlation as the real-time correlation coefficient.

5. The integrated testing method for transformer DC resistance and turns ratio according to claim 4, characterized in that, Determining the effective start time of the valid data set includes: in response to the real-time correlation coefficient being continuously greater than the first reference threshold, the starting point of the first sliding window is determined as the effective start time.

6. The integrated testing method for transformer DC resistance and turns ratio according to claim 5, characterized in that, Determining the effective termination time of a valid data set includes: after determining the effective start time, judging whether the real-time signal-to-noise ratio value drops to the next reference threshold, and judging whether the real-time correlation coefficient is continuously less than the first reference threshold; the earliest time that triggers the judgment condition is determined as the effective termination time.

7. The integrated testing method for transformer DC resistance and turns ratio according to claim 1, characterized in that, Before using the real-time correlation coefficient as a reliability weight to calculate the weighted average of the point-by-point variation sequence, the integrated testing method also includes a weight optimization step: the real-time correlation coefficient at each sampling time in the effective dataset is truncated to obtain the reliability weight corresponding to each sampling time.

8. The integrated testing method for transformer DC resistance and turns ratio according to claim 2, characterized in that, Before calculating the ratio of high-voltage induced voltage to low-voltage induced voltage point by point, the integrated testing method also includes a data verification step: Obtain the absolute value of the low-voltage induced voltage at each sampling time within the valid data set; In response to an absolute value less than the low-voltage noise floor, the corresponding sampling time is removed from the valid data set; Count the number of valid points in the valid dataset after the elimination operation; The ratio calculation process is terminated when the number of valid points is less than the preset lower limit.

9. The integrated testing method for transformer DC resistance and turns ratio according to claim 1, characterized in that, The resistance value of the fixed discharge resistor is adjusted in stages based on the estimated inductance of the winding, so that the discharge time constant of the controlled discharge process is within the preset duration range.

10. An integrated testing system for transformer DC resistance and turns ratio, characterized in that, It includes a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the integrated testing method for transformer DC resistance and turns ratio as described in any one of claims 1-9 is implemented.

Images