Transformer winding hot-spot temperature soft measurement method and system based on current harmonics

Abstract

The present application relates to power transformer state monitoring and electric variable spectrum measurement technical field, especially transformer winding hotspot temperature soft measurement method and system based on current harmonic. Method includes: collecting transformer winding three-phase instantaneous current and top layer oil temperature, carrying out sliding window spectrum measurement to three-phase instantaneous current, obtaining current effective value and frequency of fundamental wave and each effective harmonic; constructing thermal response branch according to each frequency component and determining heat generation power and equivalent heat capacity, correcting branch thermal resistance according to comprehensive harmonic distortion rate, determining each branch temperature rise according to heat generation power, equivalent heat capacity and dynamic equivalent thermal resistance iteration; the temperature candidate value is formed by combining each branch temperature rise with top layer oil temperature, and the winding hotspot temperature is measured by carrying out extreme value retrieval to temperature candidate value. The present application can retain local rapid thermal response of high frequency harmonic, and realize main heat generation frequency band tracing.

Description

Technical Field

[0001] This invention relates to the field of power transformer condition monitoring and electrical variable spectrum measurement technology, and in particular to a soft measurement method and system for transformer winding hot spot temperature based on current harmonics. Background Technology

[0002] In the operation monitoring of power transformers, the winding hot spot temperature is an important parameter for evaluating insulation thermal life, load capacity, and operational safety. Existing methods for obtaining hot spot temperature mainly include direct temperature measurement based on pre-embedded optical fibers and indirect measurement based on load current, top oil temperature, and equivalent thermal models. Pre-embedded optical fibers can obtain local winding temperatures, but existing transformers usually lack the conditions for pre-embedding, and it is difficult to install temperature sensing elements inside the windings. Indirect measurement methods have the characteristics of non-invasiveness and are more suitable for online operation monitoring.

[0003] Power electronic loads and nonlinear loads introduce multiple harmonic frequency components into the winding current. Due to the skin effect and proximity effect, the distribution of currents of different frequencies in the winding conductors, AC losses, and the effective volume involved in rapid thermal response vary. Existing indirect measurement methods typically convert harmonic currents into equivalent fundamental frequency losses, or apply the heat generation effects of each frequency component to the lumped heat model of the winding stage. This makes it difficult to separately characterize the overall heating of the fundamental frequency and the local rapid heating of high-frequency harmonics. The local temperature rise peaks caused by high-frequency harmonics are easily over-smoothed.

[0004] Meanwhile, the thermal parameters in existing equivalent thermal models are usually fixed values, which are difficult to adapt to changes in the winding thermal response under non-stationary operating conditions such as nonlinear load switching and harmonic distortion. Therefore, how to use three-phase instantaneous current and top oil temperature to characterize the heat generation effect and thermal response characteristics corresponding to different frequency components, and to measure the winding hot spot temperature while maintaining the stability of temperature estimation, is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] To address the challenges in existing technologies that make it difficult to characterize the heat generation and thermal response characteristics of different frequency current components, and the difficulty of adapting fixed thermal parameters to non-stationary harmonic operating conditions, this invention provides a soft measurement method and system for transformer winding hot spot temperature based on current harmonics.

[0006] In a first aspect, the present invention provides a soft measurement method for transformer winding hot spot temperature based on current harmonics, employing the following technical solution: A soft measurement method for transformer winding hot spot temperature based on current harmonics includes: acquiring the three-phase instantaneous current and top oil temperature of the transformer winding; performing spectrum measurement on the three-phase instantaneous current based on a sliding sampling window to obtain the fundamental current RMS value, fundamental frequency, and the RMS values ​​and frequencies of each effective harmonic current; constructing thermal response branches corresponding to each frequency component based on the fundamental frequency and the frequencies of each effective harmonic; and determining the heat generation power of each thermal response branch based on the fundamental current RMS value and the RMS values ​​of each effective harmonic current; and further determining the heat generation power of each thermal response branch based on the frequency components. The equivalent heat capacity of each thermal response branch is determined by the rate component. The comprehensive harmonic distortion rate is calculated, and the dynamic equivalent thermal resistance of each thermal response branch is determined by the bounded nonlinear relationship based on the comparison between the comprehensive harmonic distortion rate and the preset reference distortion rate threshold. Taking the top oil temperature as the heat sink boundary, the branch temperature rise of each thermal response branch is determined iteratively based on the heat generation power, equivalent heat capacity and dynamic equivalent thermal resistance. The branch temperature rise is combined with the top oil temperature to form temperature candidate values. Extreme value retrieval is performed on each temperature candidate value to measure the hot spot temperature of the transformer winding.

[0007] Furthermore, the spectrum measurement includes: windowing and fast Fourier transforming the three-phase instantaneous currents within each sliding sampling window, performing amplitude compensation according to the one-sided amplitude normalization method and combined with the coherent gain of the window function, converting the amplitude-compensated spectrum components into the effective current value, and obtaining the effective value spectrum components of the three-phase current.

[0008] Further, obtaining the fundamental current RMS value, fundamental frequency, and the RMS current values ​​and frequencies of each effective harmonic includes: screening the fundamental frequency spectrum components of each phase within the neighborhood of the rated fundamental frequency and determining the fundamental frequency; screening the harmonic spectrum components of each phase using amplitude conditions, frequency consistency conditions, and continuous window conditions within the neighborhood of integer multiples of the fundamental frequency and determining the frequencies of each effective harmonic; performing root mean square synthesis on the fundamental current RMS value of each phase and the phase current RMS value of the same harmonic, respectively, and using the corresponding root mean square synthesis results as the fundamental current RMS value and the corresponding effective harmonic current RMS value, respectively.

[0009] Furthermore, the thermal response branches respectively characterize candidate local hot zones related to the corresponding frequency and effective heating depth; the thermal response branch corresponding to the fundamental frequency adopts a reference equivalent heat capacity; the equivalent heat capacity of the thermal response branches corresponding to each effective harmonic is determined by the reference equivalent heat capacity and an adjustment factor. The adjustment factor increases as the ratio of the fundamental frequency to the frequency of the corresponding effective harmonic increases, and the equivalent heat capacity of the thermal response branches corresponding to each effective harmonic is limited by the lower limit of the equivalent heat capacity, so that the equivalent heat capacity decreases as the frequency of the corresponding effective harmonic increases before reaching the lower limit of the equivalent heat capacity, and remains unchanged after reaching the lower limit of the equivalent heat capacity.

[0010] Further, determining the heat generation power of each thermal response branch includes: obtaining the single-phase fundamental equivalent loss resistance and the single-phase frequency-dependent equivalent loss resistance corresponding to each effective harmonic, determined according to the same single-phase reference; taking the square of the effective value of the fundamental current, the single-phase fundamental equivalent loss resistance, and the product of the three-phase quantity as the heat generation power of the thermal response branch corresponding to the fundamental current; and taking the square of the effective value of the current of the corresponding effective harmonic, the corresponding single-phase frequency-dependent equivalent loss resistance, and the product of the three-phase quantity as the heat generation power of the thermal response branch corresponding to each effective harmonic.

[0011] Furthermore, the overall harmonic distortion rate is determined based on the effective values ​​of the currents of each effective harmonic and the effective value of the fundamental current. When the effective value of the fundamental current reaches the reliability condition, the deviation of the overall harmonic distortion rate from the preset reference distortion rate threshold is determined, and the deviation is input into a bounded, monotonically increasing nonlinear relationship to obtain a dynamic equivalent thermal resistance correction factor. When the effective value of the fundamental current does not reach the reliability condition, the overall harmonic distortion rate of the current sliding sampling window is not calculated.

[0012] Furthermore, the dynamic equivalent thermal resistance of each thermal response branch is determined based on the corresponding initial equivalent thermal resistance and the dynamic equivalent thermal resistance correction factor. When the overall harmonic distortion rate is higher than the preset reference distortion rate threshold, the effective value of the fundamental current reaches the reliability condition, and the three-phase instantaneous current and the top oil temperature are both in an effective state, dynamic equivalent thermal resistance correction is performed, and the corresponding boundary value is maintained when the dynamic equivalent thermal resistance correction factor reaches the preset boundary and the aforementioned conditions continue to be met. When the overall harmonic distortion rate falls below the preset reference distortion rate threshold, or when the three-phase instantaneous current and the top oil temperature are both in an effective state but the effective value of the fundamental current does not reach the reliability condition, the dynamic equivalent thermal resistance is gradually reduced to the corresponding initial equivalent thermal resistance according to the preset reduction ratio.

[0013] Further, the iterative determination of the branch temperature rise of each of the thermal response branches includes: determining the net heat generation power based on the difference between the heat generation power and the heat dissipation power determined by the branch temperature rise and the dynamic equivalent thermal resistance; updating the branch temperature rise based on the net heat generation power, the equivalent heat capacity, and the iteration time interval; identifying harmonics that simultaneously satisfy the amplitude condition, the frequency consistency condition relative to an integer multiple of the fundamental frequency, and the continuous window condition as the current effective harmonics; when a certain harmonic is no longer identified as the current effective harmonic and the three-phase instantaneous current is in an effective state, setting the heat generation power of the corresponding thermal response branch to zero, and continuing to iteratively update the branch temperature rise of the corresponding thermal response branch until the branch temperature rise of the corresponding thermal response branch meets the preset exit condition.

[0014] Furthermore, the measurement of the transformer winding hot spot temperature includes: combining the branch temperature rise of the thermal response branch corresponding to the fundamental wave, the branch temperature rise of the thermal response branch corresponding to the current effective harmonic, and the branch temperature rise of the thermal response branch that has not yet met the preset exit condition with the top oil temperature to form candidate values ​​for the fundamental wave temperature and candidate values ​​for each harmonic temperature; performing extreme value retrieval on the candidate values ​​for the fundamental wave temperature and candidate values ​​for each harmonic temperature, and taking the maximum temperature candidate value as the transformer winding hot spot temperature.

[0015] Secondly, the present invention provides a soft measurement system for transformer winding hot spot temperature based on current harmonics, employing the following technical solution: A transformer winding hotspot temperature soft measurement system based on current harmonics includes: a current measurement module for acquiring the three-phase instantaneous current of the transformer winding; a top oil temperature input module for obtaining the top oil temperature; a spectrum measurement module for performing spectrum measurement on the three-phase instantaneous current based on a sliding sampling window, and obtaining the fundamental current RMS value, fundamental frequency, and the RMS values ​​and frequencies of each effective harmonic according to the current RMS value caliber; a branch management module for constructing thermal response branches corresponding to each frequency component based on the fundamental frequency and the frequencies of each effective harmonic; and a thermal parameter determination module for... The heat generation power of each thermal response branch is determined based on the effective value of the fundamental current and the effective values ​​of the currents of each effective harmonic. The equivalent heat capacity of each thermal response branch is determined based on each frequency component, and the comprehensive harmonic distortion rate is calculated to determine the dynamic equivalent thermal resistance of each thermal response branch. The temperature rise iteration module is used to update the branch temperature rise of each thermal response branch based on the heat generation power, equivalent heat capacity, and dynamic equivalent thermal resistance, with the top oil temperature as the heat sink boundary. The temperature output module is used to combine the temperature rise of each branch with the top oil temperature to form temperature candidate values, perform extreme value retrieval on each temperature candidate value, and measure the hot spot temperature of the transformer winding.

[0016] Compared with the prior art, the present invention has at least the following technical effects: (1) In view of the problem that the traditional indirect measurement method converts the harmonic current into the equivalent fundamental loss and the high-frequency local rapid heating is over-smoothed by the winding-level lumped heat model, the present invention constructs independent thermal response branches for the fundamental wave and each effective harmonic, and matches the heat generation power and equivalent heat capacity according to the frequency component difference, so as to more accurately match the differences in heat generation distribution and thermal response characteristics of different frequency currents under the skin effect and proximity effect: the fundamental wave branch corresponds to the slow heating process of the entire winding, and the high-frequency harmonic branch corresponds to the rapid temperature rise of the local conductor. The two are calculated in parallel to avoid the thermal response of different frequency components being averaged with each other, thereby effectively avoiding the smoothing of the local temperature rise peak caused by the high-frequency harmonic, and improving the accuracy of winding hot spot temperature measurement.

[0017] (2) In view of the shortcomings of the traditional equivalent thermal model with fixed thermal parameters and difficulty in adapting to nonlinear load switching and harmonic distortion fluctuations, the present invention dynamically corrects the equivalent thermal resistance of each thermal response branch through bounded nonlinear relationship based on the comparison results of the comprehensive harmonic distortion rate and the reference threshold. When the harmonic distortion rate increases and the winding thermal response state changes, the thermal resistance parameter is updated synchronously and adaptively. This not only ensures the stability of temperature estimation under steady-state conditions, but also improves the dynamic response speed under non-stationary conditions, avoids the estimation deviation of the fixed parameter model when the load fluctuates, and makes the model always fit the real thermal conduction characteristics of the winding.

[0018] (3) This invention only needs to collect the existing three-phase instantaneous current and top oil temperature data of the transformer, and can complete the winding hot spot temperature calculation by relying on algorithm modeling. There is no need to pre-embed optical fiber inside the winding, nor is it necessary to disassemble the existing transformer to install temperature measuring elements. The solution has low deployment cost and low implementation difficulty. It can be directly connected to the existing transformer online monitoring system, and is suitable for the winding hot spot temperature monitoring needs of a large number of old transformers in operation. It effectively avoids power outage losses and equipment safety risks caused by hardware modification.

[0019] (4) This method takes the top oil temperature as the unified heat sink boundary, iteratively calculates the branch temperature rise of each frequency branch and forms temperature candidate values, and locks the winding hot spot temperature through extreme value retrieval. It conforms to the physical law that the winding hot spot is the local highest temperature and determines the insulation thermal life limit. Compared with the average winding temperature output by the traditional lumped model, the calculation results can better reflect the real heat load borne by the insulation material. It can provide more reliable core parameters for transformer insulation life assessment, load capacity verification and operation risk warning, and effectively support the safe operation and management of transformers. Attached Figure Description

[0020] Figure 1 This is a flowchart of the method for soft measurement of transformer winding hot spot temperature based on current harmonics in an embodiment of the present invention. Detailed Implementation

[0021] 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.

[0022] This embodiment provides a soft measurement method for transformer winding hot spot temperature based on current harmonics. The method uses the three-phase instantaneous current of the transformer winding and the top oil temperature to obtain the effective values ​​and frequencies of the fundamental current and each effective harmonic. Thermal response branches are constructed for different frequency components, and the branch temperature rise of each thermal response branch is determined iteratively based on the heat generation power, equivalent heat capacity, and dynamic equivalent thermal resistance. Finally, the transformer winding hot spot temperature is measured by performing extreme value retrieval on the candidate temperature values.

[0023] like Figure 1 As shown, the method in this embodiment includes data acquisition and verification, sliding sampling window spectrum measurement, effective frequency screening and thermal response branch construction, heat generation power determination, equivalent heat capacity determination, dynamic equivalent thermal resistance determination, branch temperature rise iteration, and temperature candidate value extremum retrieval.

[0024] The specific sampling frequency, sliding sampling window length, amplitude threshold, number of consecutive windows, and thermal parameter values ​​in this embodiment are used to illustrate the implementation of the present invention and do not constitute a limitation on the scope of protection of the present invention. For transformers of different models, capacities, voltage levels, and cooling methods, corresponding parameters can be set according to experimental data and actual operating conditions.

[0025] S001: Collect three-phase instantaneous current and top oil temperature.

[0026] Three-phase current transformers or current sensors corresponding to the monitored windings are installed on the low-voltage or high-voltage side of the transformer to collect the instantaneous currents of phases A, B, and C respectively. The three-phase current sampling channels are triggered by the same sampling clock to ensure that the three-phase instantaneous currents are synchronized in time.

[0027] The top oil temperature is obtained from the transformer's existing top oil temperature sensor or from a separately set temperature acquisition channel.

[0028] In this embodiment, the sampling frequency of the three-phase instantaneous current is 10kHz, the sampling period of the top oil temperature is 1s, and the thermal state update time interval is 0.1s. Between two adjacent top oil temperature sampling times, the most recently obtained effective top oil temperature is used as the thermal sink boundary at the current thermal state update time.

[0029] During the data acquisition process, the validity of the three-phase instantaneous current and top oil temperature is checked. The validity check of the three-phase instantaneous current includes at least range check, clipping check, missing value check and time synchronization check. The validity check of the top oil temperature includes at least range check, timestamp check and update status check.

[0030] When any phase current experiences continuous sampling loss, sampling clipping, or time asynchrony, or when the top oil temperature exceeds the sensor's range, calibration fails, or a valid timestamp is lost, the current thermal state update cycle will be marked as an invalid input.

[0031] During periods of invalid input, the branch temperature rise and dynamic equivalent thermal resistance of each thermal response branch are frozen, the most recent valid transformer winding hot spot temperature is maintained, and no dynamic equivalent thermal resistance correction is performed using invalid data, nor is the thermal response branch exited based on invalid data.

[0032] Ambient temperature can be used as data for operation records or offline parameter verification, but it is not a necessary input for the iteration of transformer winding hot spot temperature in this embodiment.

[0033] S002: Perform sliding sampling window spectrum measurement on the three-phase instantaneous current.

[0034] Sliding sampling windows are established for the instantaneous currents of phases A, B, and C respectively. In this embodiment, the time length of the sliding sampling window is 0.2s, and the movement time between adjacent sliding sampling windows is 0.1s.

[0035] The number of sampling points contained in each sliding sampling window is determined according to the following formula: ; In the formula, The number of sampling points contained within a sliding sampling window, dimensionless; The sampling frequency of the three-phase instantaneous current is expressed in Hertz (Hz). The time length of the sliding sampling window is expressed in seconds, denoted as s.

[0036] when ,and At that time, the number of sampling points included in the sliding sampling window is .

[0037] The DC component of each phase instantaneous current within each sliding sampling window is removed, then multiplied by a window function and a fast Fourier transform is performed. This embodiment uses the Hanning window, but other window functions can also be used depending on the requirements of spectral leakage and frequency resolution.

[0038] The frequency resolution of the Fast Fourier Transform is determined according to the following formula: ; In the formula, This represents the frequency interval between adjacent discrete frequency points, expressed in Hz.

[0039] No. The frequencies corresponding to each discrete frequency point are determined according to the following relationship: ; In the formula, For the first The frequency corresponding to each discrete frequency point is expressed in Hz. is the frequency point index of the Fast Fourier Transform, which is a non-negative integer and dimensionless.

[0040] In this embodiment, the frequency resolution is .

[0041] The Fast Fourier Transform results are normalized using a one-sided amplitude method, and amplitude compensation is performed based on the coherence gain of the window function. For spectral components other than the DC component and the Nyquist frequency component, the compensated one-sided peak amplitude is determined according to the following relationship: ; In the formula, For the first Phase in The compensated single-sided peak amplitude at each frequency point, in A; For the first Phase in Fast Fourier Transform results at each frequency point; The window function coherent gain is dimensionless. It indicates phase A, phase B, or phase C; the absolute value symbol indicates the amplitude of the corresponding fast Fourier transform result.

[0042] When the compensated single-sided amplitude is expressed using the peak value, it is converted into the effective value of the current. ; In the formula, For the first Phase in The effective value of the current corresponding to the frequency component at each frequency point, in amperes (A).

[0043] When the spectrum measurement program outputs the spectrum amplitude directly according to the effective value, the conversion from peak value to effective value is no longer performed repeatedly.

[0044] When the spectral peak is located between adjacent discrete frequency points, parabolic interpolation or other peak interpolation can be performed using the peak frequency point and its adjacent frequency points to obtain the compensated peak frequency and peak amplitude.

[0045] S003: Obtain the fundamental frequency and each effective harmonic and construct the thermal response branch.

[0046] Within each sliding sampling window, the fundamental frequency components of each phase are first searched in the neighborhood of the nominal fundamental frequency.

[0047] For a transformer with a rated frequency of 50Hz, the fundamental peak value can be searched within the frequency range of 45Hz to 55Hz; for a transformer with a rated frequency of 60Hz, the fundamental search range can be set accordingly.

[0048] The fundamental frequency of the current sliding sampling window is determined based on the peak frequencies of the fundamental frequency spectral components of each phase; the fundamental frequency can be determined by the average value, median value, or quality-weighted result of the three-phase fundamental peak frequencies.

[0049] After obtaining the fundamental frequency, search for the spectral components of each harmonic within the neighborhood of integer multiples of the fundamental frequency; The frequency uniformity condition of subharmonics is determined according to the following relationship: ; In the formula, For the first Within the first sliding sampling window The peak frequency of the subharmonic is expressed in Hz. For the first The fundamental frequency within each sliding sampling window, in Hz; The harmonic order is an integer greater than 1 and dimensionless; Allowable deviation for frequency consistency, in Hz; This is the sequence number of the sliding sampling window.

[0050] In this embodiment, the allowable deviation for frequency consistency can be taken as half of the frequency resolution. It can also be set according to the peak interpolation error and the power grid frequency fluctuation range.

[0051] For a harmonic spectral component to be determined as the current effective harmonic, the following conditions must be met simultaneously: Its peak amplitude reaches the preset amplitude threshold; Its peak frequency satisfies the frequency uniformity condition relative to an integer multiple of the fundamental frequency; It reaches a preset number of occurrences within a preset number of consecutive sliding sampling windows.

[0052] In this embodiment, the amplitude threshold is 0.5% of the rated phase current and is not less than three times the standard deviation of the no-load sampling noise; the continuous window condition is set to at least three of the most recent five sliding sampling windows satisfy the amplitude condition and the frequency consistency condition; the highest harmonic order detected is forty.

[0053] The amplitude threshold, number of sliding sampling windows, number of valid occurrences, and highest harmonic order can be adjusted according to sensor noise, sampling frequency, and frequency band to be monitored.

[0054] The effective values ​​of the fundamental currents in phases A, B, and C were obtained respectively, and then the root mean square (RMS) synthesis was performed. ; In the formula, For the first The effective value of the fundamental current in each sliding sampling window, in A; , and The first The effective values ​​of the fundamental current of phases A, B, and C within a sliding sampling window, all in amperes (A); the number 3 indicates the number of three phases.

[0055] For the The effective harmonics are synthesized by performing root-mean-square synthesis of the effective values ​​of the corresponding three-phase harmonic currents: ; In the formula, For the first Within the first sliding sampling window The effective value of the current of the first effective harmonic, in amperes (A); , and The first Within the sliding sampling window, the first phase of phases A, B, and C... The effective values ​​of the second harmonic currents are all expressed in amperes (A).

[0056] A thermal response branch corresponding to the fundamental frequency component is established, and a corresponding harmonic thermal response branch is established for each effective harmonic.

[0057] Each thermal response branch characterizes a candidate local hot zone related to the corresponding frequency and effective heating depth; each thermal response branch stores the branch identifier, corresponding frequency, effective current value, heat generation power, equivalent heat capacity, initial equivalent thermal resistance, dynamic equivalent thermal resistance, and branch temperature rise.

[0058] When a certain harmonic first meets the current effective harmonic condition, the corresponding harmonic thermal response branch is established.

[0059] When a certain harmonic is no longer identified as the current effective harmonic, but the three-phase instantaneous current is still in an effective state, the corresponding thermal response branch is not immediately deleted. Instead, the heat generation power of the corresponding thermal response branch is set to zero, allowing it to continue to decay residual heat.

[0060] In this embodiment, when the branch temperature rise of the corresponding harmonic thermal response branch is below 0.05K for five consecutive thermal state updates, it is determined that the preset exit condition is met, and the corresponding thermal response branch is deleted or deactivated.

[0061] S004: Determine the heat generation power of each thermal response branch.

[0062] The equivalent loss resistance of a single-phase fundamental wave can be determined based on transformer load loss test, short-circuit test, or winding DC resistance and additional loss data.

[0063] The single-phase frequency-dependent equivalent loss resistance corresponding to each effective harmonic can be determined based on winding conductor structure calculations, frequency characteristic tests, or electromagnetic simulations verified by measured data; the single-phase frequency-dependent equivalent loss resistance is used to reflect the characteristics of AC loss caused by skin effect and proximity effect as frequency changes.

[0064] The heat generation power of the thermal response branch corresponding to the fundamental wave is determined according to the following relationship: ; In the formula, For the first The total three-phase heat generation power of the thermal response branch corresponding to the fundamental wave within each sliding sampling window, in W; This is the effective value of the fundamental current, in amperes (A). The single-phase fundamental equivalent loss resistance is determined according to the single-phase standard, and the unit is Ω; the number 3 indicates the number of three phases.

[0065] No. The heat generation power of the thermal response branch corresponding to the first effective harmonic is determined according to the following formula: ; In the formula, For the first Within the first sliding sampling window The total three-phase heat generation power of the thermal response branch corresponding to the first effective harmonic, in W; For the first The effective value of the current of the first effective harmonic, in amperes (A); This is the single-phase frequency-dependent equivalent loss resistance determined according to a single-phase reference at the corresponding frequency, in Ω.

[0066] In this embodiment, the heat generation power adopts the three-phase total power calibrator. Therefore, the corresponding initial equivalent thermal resistance is calibrated using the three-phase total heat generation power as input.

[0067] When using single-phase local thermal resistance, the heat generation power and corresponding candidate temperature values ​​of each phase should be calculated separately, and the single-phase local thermal resistance should not be mixed with the total heat generation power of the three phases.

[0068] When the imbalance of the three-phase current exceeds the preset imbalance threshold, the heat generation power of each phase can be calculated separately.

[0069] The fundamental heat generation power of each phase is determined according to the following formula: ; In the formula, It indicates phase A, phase B, or phase C.

[0070] The total heat generation power of the three-phase fundamental wave is determined according to the following formula: .

[0071] No. The heat generation power of each phase of the subharmonic is determined according to the following formula: .

[0072] No. The total three-phase heat generation power of the subharmonic is determined according to the following formula: .

[0073] Frequency-dependent equivalent loss resistance is an electrical loss parameter used to determine the heat power generated by the corresponding frequency component; the dynamic equivalent thermal resistance, described later, is a thermal parameter used to characterize the heat dissipation capacity of the corresponding candidate local hot zone relative to the top oil temperature heat sink. Both are determined through electrical loss testing and thermal response calibration, respectively, to avoid redundant compensation for the same loss effect.

[0074] S005: Determine the equivalent heat capacity of each thermal response branch.

[0075] The thermal response branch corresponding to the fundamental frequency adopts the reference equivalent heat capacity: ; In the formula, This represents the equivalent heat capacity of the thermal response branch corresponding to the fundamental frequency, expressed in J / K. The base equivalent heat capacity is expressed in J / K.

[0076] No. The equivalent heat capacity of the thermal response branch corresponding to the first effective harmonic is determined according to the following relationship: ; In the formula, For the first Within the first sliding sampling window The equivalent heat capacity of the thermal response branch corresponding to the first effective harmonic is expressed in J / K. This is the lower limit of the equivalent heat capacity, expressed in J / K. This is the fundamental frequency, measured in Hz. For the first The frequency of the first effective harmonic is expressed in Hz. The regulation index is greater than zero and is dimensionless.

[0077] exist Furthermore, when the effective harmonic frequency is higher than the fundamental frequency, the ratio of the fundamental frequency to the effective harmonic frequency is less than 1. As the effective harmonic frequency increases, the equivalent heat capacity that has not reached the lower limit of the equivalent heat capacity gradually decreases; once the calculated equivalent heat capacity reaches the lower limit of the equivalent heat capacity, the equivalent heat capacity remains unchanged.

[0078] In this embodiment, the reference equivalent heat capacity is taken as Adjustment index The lower limit of the equivalent heat capacity is taken as 1% of the reference equivalent heat capacity, that is... .

[0079] When the measured frequency of the same harmonic thermal response branch changes between adjacent sliding sampling windows, thus causing a change in the equivalent heat capacity, the branch temperature rise can be corrected according to the principle of continuous thermal state energy.

[0080] The thermal states before and after the adjustment satisfy: ; Therefore, we can conclude that: ; In the formula, and The values ​​are the equivalent heat capacity before and after adjustment, in J / K. The branch temperature rise before adjustment is expressed in Kelvin (K). This is the adjusted branch temperature rise for the next temperature rise iteration, in Kelvin (K).

[0081] In another embodiment, the measured fundamental frequency and harmonic frequency can also be smoothed to reduce short-term fluctuations in equivalent heat capacity between adjacent sliding sampling windows.

[0082] Each thermal response branch independently preserves its thermal state in the computational structure; radial or transverse thermal conduction may exist between different candidate local hot zones in the actual winding; for thermal coupling between branches that is not explicitly established, its influence can be identified through offline parameters and reflected in the frequency-dependent equivalent loss resistance, the reference equivalent thermal capacity, the regulation index, and the initial equivalent thermal resistance of each thermal response branch.

[0083] When identifying offline parameters, the initial value of the single-phase fundamental equivalent loss resistance can be obtained first by load loss test or short circuit test, and then the initial value of the single-phase frequency-related equivalent loss resistance can be obtained by conductor structure calculation, frequency characteristic test or electromagnetic simulation.

[0084] Subsequently, under representative operating conditions including steady-state load changes, single major harmonic changes, and multiple harmonic changes, the three-phase instantaneous current, top oil temperature, and reference hot spot temperature were simultaneously collected.

[0085] The reference hot spot temperature can be obtained from the pre-embedded optical fiber; when optical fiber temperature measurement is not available, thermal simulation results verified by temperature rise test can also be used as auxiliary calibration data.

[0086] Using the mean square error or root mean square error between the hot spot temperature obtained by soft measurement and the reference hot spot temperature as the optimization target, the equivalent parameters are adjusted under the constraint that the equivalent heat capacity, equivalent thermal resistance and regulation index are all positive.

[0087] Calibration and verification data are divided into different time periods and do not overlap. After the parameters are determined, the maximum error, root mean square error and temperature change trend are checked on the verification data. When the verification conditions cover the predetermined load range, cooling state, top oil temperature range and harmonic range, and the error meets the preset engineering requirements, the corresponding parameters are used for online soft measurement.

[0088] S006: Calculate the overall harmonic distortion rate and determine the dynamic equivalent thermal resistance.

[0089] Let the reliability lower limit of the fundamental current RMS value be... In this embodiment, the reliability lower limit is 1% of the rated phase current, and is not lower than the amplitude threshold used for effective frequency screening.

[0090] When the effective value of the fundamental current reaches the reliability condition, the comprehensive harmonic distortion rate is calculated according to the following formula: ; In the formula, For the first The overall harmonic distortion rate of each sliding sampling window, dimensionless; This is the current set of effective harmonic orders; For the first The effective value of the current of the first effective harmonic, in amperes (A); This is the effective value of the fundamental current, expressed in amperes (A).

[0091] When the effective value of the fundamental current does not meet the reliability condition, the comprehensive harmonic distortion rate of the current sliding sampling window is not calculated to avoid the abnormal amplification of the comprehensive harmonic distortion rate caused by using a small effective value of the fundamental current as the denominator.

[0092] Let the preset baseline distortion rate threshold be... In this embodiment, we take .

[0093] When the overall harmonic distortion rate exceeds the preset reference distortion rate threshold, the effective value of the fundamental current meets the reliability conditions, and both the three-phase instantaneous current and the top oil temperature are in an effective state, the distortion deviation is determined as follows: ; By inputting the distortion deviation into a bounded, monotonically increasing nonlinear relationship, the dynamic equivalent thermal resistance correction factor is obtained: ; In the formula, The dynamic equivalent thermal resistance correction factor is dimensionless. The maximum compensation amplitude coefficient is a positive number and dimensionless. is a nonlinear sensitivity coefficient, which is a positive number and dimensionless; It is the hyperbolic tangent function.

[0094] when and When all are positive, the dynamic equivalent thermal resistance correction factor increases monotonically with the distortion deviation and is limited by the range of the hyperbolic tangent function.

[0095] In this embodiment ,and .

[0096] Boundary constraints are imposed on the dynamic equivalent thermal resistance correction factor: ; In the formula, The dynamic equivalent thermal resistance correction factor after boundary constraints is dimensionless. and These are the preset lower boundary and the preset upper boundary, respectively, both of which are dimensionless parameters.

[0097] In this embodiment ,and .

[0098] No. The dynamic equivalent thermal resistance of a thermal response branch is determined according to the following formula: ; In the formula, For the first Within the first sliding sampling window The dynamic equivalent thermal resistance of each thermal response branch, in K / W; For the first The initial equivalent thermal resistance of the thermal response branch, in K / W; This represents the thermal response branch corresponding to the fundamental frequency or any harmonic frequency.

[0099] In this embodiment, the initial equivalent thermal resistance of each thermal response branch can be taken as... And adjust them separately by identifying offline parameters.

[0100] When the dynamic equivalent thermal resistance correction factor reaches the preset boundary, and the comprehensive harmonic distortion rate is higher than the preset reference distortion rate threshold, the fundamental current effective value reaches the reliability condition, and the input data is valid, the corresponding boundary value is maintained.

[0101] Since the output value of the hyperbolic tangent function may only asymptotically approach the upper boundary, a correction factor boundary tolerance can be set. When satisfied When this occurs, the dynamic equivalent thermal resistance correction factor is considered to have reached its upper boundary.

[0102] When the overall harmonic distortion rate falls below the preset reference distortion rate threshold, or when the three-phase instantaneous current and top oil temperature are both in an effective state but the effective value of the fundamental current does not meet the reliability conditions, the dynamic equivalent thermal resistance will gradually return to the corresponding initial equivalent thermal resistance: ; In the formula, The backoff ratio is a dimensionless value that is greater than 0 and not greater than 1.

[0103] In this embodiment After each rollback, determine... In the formula, This is the thermal resistance rollback termination tolerance, expressed in K / W.

[0104] In this embodiment When the above conditions are met, directly order And terminate the rollback.

[0105] When the three-phase instantaneous current or top oil temperature is in an invalid state, the dynamic equivalent thermal resistance correction or rollback is not performed, but the current dynamic equivalent thermal resistance is frozen.

[0106] S007: Iteratively determine the branch temperature rise of each thermal response branch.

[0107] The top oil temperature is used as the common heat sink boundary for all thermal response branches; the thermal state update time interval is set to... In this embodiment .

[0108] For the For each thermal response branch, the net heat production power is determined based on the difference between its heat production power and the heat dissipation power formed by the branch's temperature rise through dynamic equivalent thermal resistance. ; In the formula, For the first The first hot state update time. Net heat output of each thermal response branch, in W; The heat generation power of the corresponding thermal response branch is expressed in W. The branch temperature rise corresponds to the thermal response branch, in Kelvin (K). The dynamic equivalent thermal resistance of the corresponding thermal response branch is expressed in K / W.

[0109] The branch temperature rise is updated using an explicit differential method based on net heat output, equivalent heat capacity, and thermal state update time interval: ; In the formula, The equivalent heat capacity of the corresponding thermal response branch is expressed in J / K. The updated branch temperature rise is expressed in Kelvin; the larger value between zero and the calculation result is used to avoid negative temperature rise during the heat dissipation stage.

[0110] For the thermal response branch corresponding to the fundamental frequency, the temperature rise is updated according to the following relationship: ; In the formula, The updated temperature rise of the fundamental branch is expressed in Kelvin (K). This represents the heat output power of the fundamental frequency branch, expressed in W. The reference equivalent heat capacity is expressed in J / K. This represents the dynamic equivalent thermal resistance of the thermal response branch corresponding to the fundamental frequency, expressed in K / W.

[0111] For the The temperature rise of the thermal response branch corresponding to the subharmonic is updated according to the following relationship: ; In the formula, For the updated number Temperature rise of the second harmonic branch, in K; For the first The heat generation power corresponding to the second harmonic is expressed in W. For the first The equivalent heat capacity corresponding to the subharmonic is expressed in J / K. For the first The dynamic equivalent thermal resistance corresponding to the subharmonic is expressed in K / W.

[0112] The dimensional relationship of the above temperature rise increment is as follows: ; The explicit differential thermal state update time interval should be less than the minimum thermal time constant of each thermal response branch.

[0113] No. The thermal time constant of each thermal response branch is determined according to the following relationship: ; In the formula, For the first The thermal time constant of each thermal response branch, in seconds; The dynamic equivalent thermal resistance is expressed in K / W. This is the equivalent heat capacity, expressed in J / K.

[0114] This embodiment uses offline stability checks to ensure that the thermal state update time interval is significantly smaller than the thermal time constant of each thermal response branch.

[0115] When starting from a cold state or confirming no-load start-up, the initial branch temperature rise of each thermal response branch can be set to 0.

[0116] During hot start-up under load, the effective three-phase instantaneous current and top oil temperature within a preset time period before startup can be used for preheating iteration, or the branch temperature rise of each thermal response branch can be initialized using a calibrated steady-state thermal model. Before completing initialization, an initialization status flag can be output to avoid mistaking zero temperature rise for the actual winding thermal state.

[0117] When a certain harmonic is no longer identified as the current effective harmonic and the three-phase instantaneous current is in an effective state, the heat generation power of the corresponding thermal response branch is set to 0.

[0118] The corresponding thermal response branch continues to be updated according to the temperature rise iteration formula; at this time, the branch temperature rise gradually decreases under the action of heat dissipation power, rather than instantaneously returning to 0 when the harmonic spectrum components disappear.

[0119] When the branch temperature rise of the corresponding thermal response branch meets the preset exit condition, the corresponding thermal response branch is exited; in this embodiment, the branch temperature rise of five consecutive thermal state updates is all below 0.05K as the preset exit condition.

[0120] When the overall input is invalid, freeze the branch temperature rise of each thermal response branch and do not perform heat generation or heat dissipation updates.

[0121] S008: Form candidate temperature values ​​and perform extreme value retrieval.

[0122] The thermal response branches currently participating in the temperature candidate value calculation include: the thermal response branch corresponding to the fundamental wave; the thermal response branch corresponding to the current effective harmonic; and the thermal response branch that no longer belongs to the current effective harmonic but has not yet met the preset exit conditions and is still in the residual heat decay stage.

[0123] The candidate values ​​for the fundamental temperature are determined according to the following formula: ; In the formula, These are candidate values ​​for the fundamental temperature, in °C. The top oil temperature used at the current thermal state update moment is expressed in °C. This represents the branch temperature rise of the thermal response branch corresponding to the fundamental frequency, expressed in Kelvin (K).

[0124] No. The candidate values ​​for the subharmonic temperature are determined according to the following formula: ; In the formula, For the first Candidate temperature values ​​corresponding to the subharmonics, in °C; The temperature rise of the corresponding thermal response branch is expressed in K.

[0125] Temperature rise is expressed in K, and top oil temperature is expressed in °C. Since the temperature difference of 1 K is equal to the temperature difference of 1 °C, the branch temperature rise can be directly combined with the top oil temperature.

[0126] Let the set of harmonic thermal response branches currently retained be... The set includes the thermal response branch corresponding to the current effective harmonic and the thermal response branch that has not yet met the preset exit conditions.

[0127] The hot spot temperature of the transformer winding is determined according to the following formula: ; In the formula, The transformer winding hot spot temperature measured at the current thermal state update time, in °C; Candidate values ​​for fundamental temperature; For the first Candidate temperature values ​​corresponding to the subharmonics.

[0128] Since all candidate temperature values ​​use the same top oil temperature as the heat sink boundary, we can first perform extreme value retrieval on the branch temperature rise of each thermal response branch, and then combine it with the top oil temperature: ; This formula is equivalent to the formula for retrieving extreme values ​​after generating candidate temperature values.

[0129] In this embodiment, each thermal response branch is defined as a candidate local hot zone related to the corresponding frequency and effective heating depth. The branch temperature rise of each thermal response branch is combined with the top oil temperature to form a candidate temperature value, and then the maximum value is obtained from each candidate temperature value.

[0130] This embodiment does not directly algebraically sum the branch temperature rise of different thermal response branches to avoid mechanical accumulation when the number of thermal response branches increases; the coupling effects of radial heat transfer, transverse heat transfer and different heat-generating areas that are not explicitly established are equivalently reflected within the target operating range through the aforementioned offline parameter identification and verification process.

[0131] This embodiment also provides a soft measurement system for transformer winding hot spot temperature based on current harmonics, including a current measurement module, a top oil temperature input module, a spectrum measurement module, a branch management module, a thermal parameter determination module, a temperature rise iteration module, and a temperature output module.

[0132] The current measurement module includes a three-phase current sensor, a synchronous sampling circuit, and a timestamp unit, used to collect the instantaneous three-phase current of the transformer winding; the current measurement module can also perform range checks, clipping checks, missing value checks, and three-phase synchronization checks.

[0133] The top oil temperature input module is used to obtain the top oil temperature from the transformer's original temperature sensor, temperature monitoring device, or independent temperature acquisition channel, and to perform range checks, timestamp checks, and update status checks on the top oil temperature.

[0134] The spectrum measurement module is used to perform spectrum measurement of three-phase instantaneous current based on a sliding sampling window. The spectrum measurement module sequentially performs DC removal, windowing, fast Fourier transform, single-sided amplitude normalization, window function coherent gain compensation, peak interpolation, and RMS value conversion to obtain the RMS value of the fundamental current, the fundamental frequency, and the RMS values ​​and frequencies of each effective harmonic current.

[0135] The branch management module is used to construct thermal response branches corresponding to each frequency component based on the fundamental frequency and the frequencies of each effective harmonic.

[0136] The branch management module establishes or maintains the corresponding harmonic thermal response branch based on the amplitude condition, the frequency consistency condition relative to the fundamental frequency integer multiple, and the continuous window condition. When the harmonic no longer meets the current effective harmonic condition, the heat generation power of the corresponding thermal response branch is set to zero, and the corresponding thermal response branch is exited after the branch temperature rise of the corresponding thermal response branch meets the preset exit condition.

[0137] The thermal parameter determination module is used to determine the heat generation power of each thermal response branch based on the effective value of the fundamental current, the effective value of the current of each effective harmonic, the equivalent loss resistance of the single-phase fundamental current, and the equivalent loss resistance of each single-phase frequency.

[0138] The thermal parameter determination module is also used to determine the equivalent thermal capacity of each thermal response branch based on the fundamental frequency, the frequency of each effective harmonic, the reference equivalent thermal capacity, the adjustment index, and the lower limit of the equivalent thermal capacity; calculate the comprehensive harmonic distortion rate based on the effective value of the fundamental current and the effective value of the current of each effective harmonic; and determine the dynamic equivalent thermal resistance of each thermal response branch based on the comparison result between the comprehensive harmonic distortion rate and the preset reference distortion rate threshold.

[0139] The temperature rise iteration module is used to update the branch temperature rise of each thermal response branch based on the top oil temperature as the heat sink boundary, the heat generation power, the equivalent heat capacity, and the dynamic equivalent thermal resistance.

[0140] The temperature rise iteration module is also used to handle cold initialization, hot initialization, residual heat decay after harmonic exit, and branch state freezing during the period of invalid input.

[0141] The temperature output module is used to combine the branch temperature rise of the thermal response branch corresponding to the fundamental wave, the branch temperature rise of the thermal response branch corresponding to the current effective harmonic, and the branch temperature rise of the thermal response branch that has not yet met the preset exit conditions with the top oil temperature to form temperature candidate values. It then performs extreme value retrieval on each temperature candidate value and outputs the maximum temperature candidate value as the transformer winding hot spot temperature.

[0142] The above modules can be implemented using an industrial computer, embedded processor, transformer online monitoring terminal, or server; the processor is connected to the memory, which stores program instructions, frequency-dependent equivalent loss resistance, reference equivalent thermal capacity, lower limit of equivalent thermal capacity, adjustment index, initial equivalent thermal resistance, preset reference distortion rate threshold, and dynamic equivalent thermal resistance correction parameters.

[0143] When the processor executes the program instructions in the memory, it completes the acquisition of three-phase instantaneous current and top oil temperature, sliding sampling window spectrum measurement, thermal response branch construction, heat generation power and thermal parameters determination, branch temperature rise iteration, and temperature candidate extreme value retrieval, thereby measuring the hot spot temperature of the transformer winding.

[0144] 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. A soft measurement method for transformer winding hot spot temperature based on current harmonics, characterized in that, include: The three-phase instantaneous current and top oil temperature of the transformer winding are collected. The spectrum of the three-phase instantaneous current is measured based on the sliding sampling window to obtain the effective value of the fundamental current, the fundamental frequency, and the effective values ​​and frequencies of each effective harmonic current. The thermal response branches corresponding to each frequency component are constructed based on the fundamental frequency and the frequencies of each effective harmonic, and the heat generation power of each thermal response branch is determined based on the effective value of the fundamental current and the effective values ​​of the currents of each effective harmonic. The equivalent heat capacity of each thermal response branch is determined based on each frequency component. The comprehensive harmonic distortion rate is calculated, and the dynamic equivalent thermal resistance of each thermal response branch is determined by using the bounded nonlinear relationship based on the comparison between the comprehensive harmonic distortion rate and the preset reference distortion rate threshold. Using the top oil temperature as the heat sink boundary, the branch temperature rise of each thermal response branch is determined iteratively based on the heat generation power, equivalent heat capacity, and dynamic equivalent thermal resistance. The temperature rise of each branch is combined with the top oil temperature to form candidate temperature values. Extreme value retrieval is performed on each candidate temperature value to measure the hot spot temperature of the transformer winding.

2. The soft measurement method for transformer winding hot spot temperature based on current harmonics according to claim 1, characterized in that, The spectrum measurement includes: Windowing and fast Fourier transform are applied to the three-phase instantaneous currents within each sliding sampling window. Amplitude compensation is performed using a one-sided amplitude normalization method combined with the coherent gain of the window function. The amplitude-compensated spectral components are then converted into the effective current value, thus obtaining the effective spectral components of the three-phase current.

3. The soft measurement method for transformer winding hot spot temperature based on current harmonics according to claim 2, characterized in that, The acquisition of the fundamental current RMS value, fundamental frequency, and the RMS values ​​and frequencies of each effective harmonic current includes: Within the neighborhood of the rated fundamental frequency, the fundamental frequency of each phase is determined by screening the fundamental frequency spectrum components of each phase. Within the neighborhood of integer multiples of the fundamental frequency, the harmonic spectrum components of each phase are screened by using amplitude conditions, frequency consistency conditions, and continuous window conditions, and the frequencies of each effective harmonic are determined. The root mean square (RMS) synthesis is performed on the RMS values ​​of the fundamental current and the RMS values ​​of the currents of the same harmonic in each phase, and the corresponding RMS synthesis results are used as the RMS values ​​of the fundamental current and the RMS values ​​of the currents of the corresponding effective harmonics, respectively.

4. The soft measurement method for transformer winding hot spot temperature based on current harmonics according to claim 1, characterized in that, The thermal response branches respectively characterize candidate local thermal regions related to the corresponding frequency and effective heating depth; The thermal response branch corresponding to the fundamental wave adopts the reference equivalent heat capacity; The equivalent heat capacity of the thermal response branch corresponding to each effective harmonic is determined by the reference equivalent heat capacity and the adjustment factor. The adjustment factor increases as the ratio of the fundamental frequency to the frequency of the corresponding effective harmonic increases. The equivalent heat capacity of the thermal response branch corresponding to each effective harmonic is limited by the lower limit of the equivalent heat capacity, so that the equivalent heat capacity decreases as the frequency of the corresponding effective harmonic increases before reaching the lower limit of the equivalent heat capacity, and remains unchanged after reaching the lower limit of the equivalent heat capacity.

5. The soft measurement method for transformer winding hot spot temperature based on current harmonics according to claim 3, characterized in that, Determining the heat generation power of each thermal response branch includes: Obtain the single-phase fundamental equivalent loss resistance and the single-phase frequency-dependent equivalent loss resistance corresponding to each effective harmonic, determined according to the same single-phase reference. The square of the effective value of the fundamental current, the equivalent loss resistance of the single-phase fundamental current, and the product of the number of three phases are taken as the heat generation power of the thermal response branch corresponding to the fundamental current. The square of the effective value of the current corresponding to the effective harmonic, the product of the corresponding single-phase frequency-dependent equivalent loss resistance and the number of three phases, are respectively used as the heat generation power of the thermal response branch corresponding to each effective harmonic.

6. The soft measurement method for transformer winding hot spot temperature based on current harmonics according to claim 1, characterized in that, The overall harmonic distortion rate is determined based on the effective values ​​of the currents of each effective harmonic and the effective value of the fundamental current. When the effective value of the fundamental current reaches the reliability condition, the deviation of the comprehensive harmonic distortion rate from the preset reference distortion rate threshold is determined, and the deviation is input into a bounded monotonically increasing nonlinear relationship to obtain a dynamic equivalent thermal resistance correction factor. When the effective value of the fundamental current does not meet the reliability condition, the comprehensive harmonic distortion rate of the current sliding sampling window is not calculated.

7. The soft measurement method for transformer winding hot spot temperature based on current harmonics according to claim 6, characterized in that, The dynamic equivalent thermal resistance of each thermal response branch is determined based on the corresponding initial equivalent thermal resistance and the dynamic equivalent thermal resistance correction factor. When the overall harmonic distortion rate is higher than the preset reference distortion rate threshold, the effective value of the fundamental current reaches the reliability condition, and the three-phase instantaneous current and the top oil temperature are both in an effective state, dynamic equivalent thermal resistance correction is performed, and the corresponding boundary value is maintained when the dynamic equivalent thermal resistance correction factor reaches the preset boundary and the aforementioned conditions continue to be met. When the overall harmonic distortion rate falls below the preset reference distortion rate threshold, or when the three-phase instantaneous current and the top oil temperature are both in an effective state but the effective value of the fundamental current does not reach the reliability condition, the dynamic equivalent thermal resistance is gradually reduced to the corresponding initial equivalent thermal resistance according to the preset reduction ratio.

8. The soft measurement method for transformer winding hot spot temperature based on current harmonics according to claim 7, characterized in that, The iteration to determine the branch temperature rise of each of the thermal response branches includes: The net heat generation power is determined based on the difference between the heat generation power and the heat dissipation power determined by the branch temperature rise and the dynamic equivalent thermal resistance, and the branch temperature rise is updated based on the net heat generation power, the equivalent heat capacity and the iteration time interval. Harmonics that simultaneously satisfy the amplitude condition, the frequency consistency condition relative to an integer multiple of the fundamental frequency, and the continuous window condition are determined as the current valid harmonics. When a certain harmonic is no longer identified as the current valid harmonic and the three-phase instantaneous current is in an effective state, the heat generation power of the corresponding thermal response branch is set to zero, and the branch temperature rise of the corresponding thermal response branch continues to be iteratively updated until the branch temperature rise of the corresponding thermal response branch meets the preset exit condition.

9. The soft measurement method for transformer winding hot spot temperature based on current harmonics according to claim 8, characterized in that, The measured transformer winding hot spot temperature includes: The branch temperature rise of the thermal response branch corresponding to the fundamental wave, the branch temperature rise of the thermal response branch corresponding to the current effective harmonic, and the branch temperature rise of the thermal response branch that has not yet met the preset exit conditions are combined with the top oil temperature to form candidate values ​​for the fundamental wave temperature and candidate values ​​for each harmonic temperature. Extreme value retrieval is performed on the candidate values ​​of fundamental temperature and each harmonic temperature, and the maximum temperature candidate value is taken as the hot spot temperature of the transformer winding.

10. A soft-sensing system for transformer winding hot spot temperature based on current harmonics, characterized in that, include: The current measurement module is used to collect the instantaneous three-phase current of the transformer windings; Top layer oil temperature input module, used to obtain the top layer oil temperature; The spectrum measurement module is used to perform spectrum measurement on the three-phase instantaneous current based on a sliding sampling window, and to obtain the effective value of the fundamental current, the fundamental frequency, and the effective values ​​and frequencies of each effective harmonic current according to the effective value of the current. The branch management module is used to construct thermal response branches corresponding to each frequency component based on the fundamental frequency and the frequencies of each effective harmonic. The thermal parameter determination module is used to determine the heat generation power of each thermal response branch based on the effective value of the fundamental current and the effective values ​​of the current of each effective harmonic, determine the equivalent heat capacity of each thermal response branch based on each frequency component, and calculate the comprehensive harmonic distortion rate to determine the dynamic equivalent thermal resistance of each thermal response branch. The temperature rise iteration module is used to update the branch temperature rise of each thermal response branch based on the top oil temperature as the heat sink boundary, the heat generation power, the equivalent heat capacity, and the dynamic equivalent thermal resistance. The temperature output module is used to combine the temperature rise of each branch with the top oil temperature to form a temperature candidate value, perform extreme value retrieval on each temperature candidate value, and measure the hot spot temperature of the transformer winding.

Images