Design method and system for reducing distributed capacitance of high-frequency driving transformer

Abstract

The invention relates to the technical field of high-frequency transformers, in particular to a design method and system for reducing distributed capacitance of a high-frequency drive transformer, and the method comprises the following steps: collecting instantaneous voltage and current of a winding, fitting an energy density curve, analyzing the coupling strength of a capacitance region, extracting voltage response, smoothly correcting deviation, and positioning capacitance space distribution; detecting a current phase difference and an energy path, judging a capacitance concentration ratio and adjusting a layout coordinate, calculating a phase deviation and energy ratio comparison, optimizing a capacity and spacing ratio, and correcting an error to generate an optimization result. According to the invention, accurate optimization is realized by acquiring voltage and current and fitting an energy density curve, analyzing coupling strength and capacitance distribution, iteratively correcting deviation, adjusting the proportion of capacitance to spacing, balancing an electromagnetic coupling state, improving energy transfer efficiency, reducing parasitic effect, and enhancing system stability and anti-interference capability; and the energy conversion quality and the structural reliability are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of high-frequency transformer, in particular to a design method and system for reducing the distributed capacitance of a high-frequency driving transformer. BACKGROUND

[0002] The technical field of high-frequency transformer includes the design, manufacture and application of electromagnetic energy coupling devices for high-frequency electric energy transmission and conversion. The core is to achieve efficient electromagnetic energy conversion under high-frequency conditions through reasonable design of magnetic conductive materials, winding structure and insulation method. The overall technical system includes magnetic core material property analysis, winding coupling method design, insulation medium selection, parasitic parameter control and electromagnetic compatibility optimization, and is mainly applied to switching power supply, power converter, wireless power transmission and high-frequency driving circuit. The research focus of this field is to improve the electromagnetic coupling efficiency, reduce the loss and optimize the electric field distribution to ensure the reliability and stability of the equipment under high-frequency and high-voltage conditions.

[0003] Among them, the design method and system for reducing the distributed capacitance of the high-frequency driving transformer is a design scheme for reducing the parasitic capacitance effect by adjusting the structure parameters and controlling the material properties to reduce the distributed capacitance caused by the coupling structure between windings during the high-frequency driving process of the high-frequency transformer. The technical matters cover winding arrangement optimization, interlayer insulation medium thickness design, winding interlayer shielding layer arrangement and lead-out structure adjustment, etc. Specifically, the winding interlayer electric field distribution is redesigned by changing the winding geometric arrangement ratio, selecting low dielectric constant insulation materials and adding static shielding layer at specific positions, so as to effectively suppress the distributed capacitance at the system structure level.

[0004] The existing technology mainly relies on geometric adjustment of winding structure and thickness control of insulation medium to reduce the distributed capacitance in the design of high-frequency driving transformer, but lacks dynamic analysis of energy distribution and coupling effect under actual operating conditions. Since it is only based on static structure parameter optimization, it cannot reflect the nonlinear change of electric field distribution in real time, which leads to the fact that the parasitic capacitance effect is still easy to cause local energy aggregation and voltage stress increase under high-frequency and high-voltage operating conditions. The existing method has a lag in spatial layout, and the capacitance adjustment depends on experience design, lacks feedback correction mechanism, and it is difficult to balance the electromagnetic balance among multiple windings. Since the capacitance position and capacity cannot be matched and optimized under dynamic working conditions, the energy transmission efficiency is reduced, the loss is concentrated in the coupling channel, the system resonance point is offset, and the linear stability and electromagnetic compatibility performance of high-frequency energy transmission are affected. This limitation leads to high temperature rise, increased loss, limited reliability and life of the transformer under high-frequency driving. SUMMARY

[0005] In order to solve the technical problems existing in the prior art, the embodiment of the present application provides a design method for reducing the distributed capacitance of a high-frequency driving transformer, comprising the following steps: S1: collect the instantaneous voltage and current of the main and auxiliary windings of the transformer, calculate the phase difference of the voltage and current and fit it into an energy density curve, analyze the coupling strength of the capacitance region, and generate capacitance coupling data; S2: based on the capacitance coupling data, extract the instantaneous voltage response of the capacitance region and smooth it, iteratively calculate the deviation from the energy density curve, analyze the spatial position of the capacitance region, and generate a capacitance distribution adjustment set; S3: based on the capacitance distribution adjustment set, detect the current phase difference and energy transfer path of the main and auxiliary windings, calculate the concentration of the distributed capacitance of multiple windings, judge the deviation trend of the capacitance layout and adjust the coordinates in the magnetic core window, and generate a capacitance layout set; S4: based on the capacitance layout set, obtain the current amplitude and energy transfer ratio of multiple capacitance regions under high-frequency resonance, calculate the phase deviation of the distributed capacitance and compare it with the energy transfer ratio, adjust the capacitance capacity and spacing ratio, and generate a structure configuration set; S5: based on the structure configuration set, collect the resonance current and voltage waveform in the test stage, calculate the energy feedback of the distributed capacitance, compare the error with the capacitance coupling data and correct the capacitance position and capacity, and generate a distributed capacitance optimization design result.

[0006] As a further scheme of the present application, the capacitance coupling data includes the instantaneous voltage response of the capacitance region, the phase difference, and the energy density curve, the capacitance distribution adjustment set includes the smoothed voltage data of the capacitance region, the energy density curve deviation, and the spatial position of the capacitance region, the capacitance layout set includes the winding current phase difference, the energy transfer path, and the winding distributed capacitance concentration, the structure configuration set includes the current amplitude of the capacitance region under high-frequency resonance, the energy transfer ratio, and the capacitance phase deviation, and the distributed capacitance optimization design result includes the capacitance position correction value, the capacitance capacity correction value, and the energy feedback error parameter.

[0007] As a further scheme of the present application, the specific steps of S1 are as follows: S101: collect the instantaneous voltage and current signals of the main and auxiliary windings of the transformer and match them, calculate the phase angle difference, arrange the phase difference results in the same time period in time sequence, and normalize and denoise the phase difference sequence to generate phase difference sequence data; S102: based on the phase difference sequence data, perform point-by-point product calculation on the phase difference change rate and the voltage and current amplitude change rate in all time windows to obtain instantaneous energy density distribution values, and curve fit all the energy density distribution values in the time windows to obtain an energy density fitting curve set; S103: Call the energy density fitting curve set, filter the energy density peak section in the curve, extract the corresponding phase difference and voltage current amplitude ratio, calculate the energy concentration parameter of the multi-peak section, analyze the coupling ratio of the energy concentration parameter and the phase difference distribution, analyze the coupling strength, and generate the capacitive coupling data.

[0008] As a further scheme of the present application, the specific steps of S2 are: S201: Based on the capacitive coupling data, extract the instantaneous voltage response signal in the capacitive region, perform interval average operation on the time series data of the voltage response signal, and smooth the fluctuation component of the continuous sampling section, and reorder the time to generate the smooth voltage response sequence; S202: Call the smooth voltage response sequence, perform deviation calculation on the voltage amplitude and the corresponding amplitude of the energy density curve in all time periods and perform iterative accumulation, analyze the deviation value distribution of the voltage response and the energy density, and perform interval normalization on the deviation distribution to obtain the energy deviation distribution set; S203: According to the energy deviation distribution set, perform clustering calculation on the spatial distribution coordinates of the deviation amplitude in multiple capacitive regions, identify the spatial coordinate range of the deviation concentration area according to the clustering result, and weight and integrate the multi-region coordinates and the corresponding deviation amplitude to generate the capacitive distribution adjustment set.

[0009] As a further scheme of the present application, the specific steps of S3 are: S301: Based on the capacitive distribution adjustment set, detect the current signal of the main and auxiliary windings in the same time period, perform difference calculation on the instantaneous amplitude of the current signal sequence to extract the phase angle difference, and then perform interval average operation on the phase difference change rate according to the time sequence to generate the winding phase difference sequence; S302: Call the winding phase difference sequence, calculate the energy transfer direction vector for the current direction and amplitude at the corresponding time of multiple phase differences, and aggregate the energy transfer vectors of continuous time periods into path trajectories in time sequence, identify the multi-direction mutation points of the path trajectories and coordinate fitting to obtain the energy transfer path set; S303: According to the energy transfer path set, calculate the concentration of the distributed capacitors in the path direction in multiple windings, extract the spatial coordinates of the regions whose concentration exceeds the concentration threshold, and perform proportional adjustment according to the coordinate offset direction and the concentration amplitude and remap to the magnetic core window spatial coordinates to generate the capacitive layout set.

[0010] As a further scheme of the present application, the concentration threshold is obtained by statistically analyzing the concentration parameters of the distributed capacitors in multiple windings to obtain the concentration value set of all capacitive regions, performing interval distribution analysis on the set to determine the mean and variance of the concentration, and then calculating the upper limit of the concentration distribution according to the mean and variance.

[0011] As a further scheme of the present application, the specific steps of S4 are: S401: Based on the set of capacitor layouts, obtain the current signal of the multi-capacitor region under high-frequency resonance conditions, calculate the current amplitude within the same resonance period, and simultaneously analyze the energy transfer ratio sequence according to the current input and output energy of adjacent capacitor regions to generate a set of resonance energy ratios; S402: Call the set of resonance energy ratios, calculate the time offset between the current peak point and the voltage peak point for the phase information of the current signal of the multi-distributed capacitor region, and convert it into an angle difference as a phase offset value, and perform a corresponding comparison with the energy transfer ratio to obtain a set of capacitor phase offset difference values; S403: According to the set of capacitor phase offset difference values, perform a two-variable proportional adjustment on the capacity parameters and adjacent spacing parameters of the multi-distributed capacitor, correct the capacity increase / decrease coefficient according to the positive / negative direction of the phase offset difference value, and re-calculate the energy transfer ratio change after adjustment and perform balance matching to generate a structure configuration set.

[0012] As a further scheme of the present application, the specific steps of S5 are: S501: Based on the structure configuration set, collect the resonance current and voltage waveforms in the test phase, perform time integration operation according to the instantaneous product of voltage and current within the same resonance period, calculate the difference between the energy input and the energy return of the multi-distributed capacitor node and sort it to obtain an energy feedback sequence; S502: Call the energy feedback sequence, perform error determination operation on the capacity parameters and adjacent coupling position parameters of the multi-distributed capacitor, compare the difference between the energy feedback and the energy transfer reference value, judge the offset direction of the capacity and position according to the difference symbol, and generate error offset data; S503: According to the error offset data, perform parameter correction on the capacity value and spatial position of the multi-distributed capacitor, take the offset direction as the correction symbol and perform correction scale calculation according to the difference size, update the corrected capacity parameter set and position arrangement data set, and generate the distributed capacitor optimization design result.

[0013] As a further scheme of the present application, the energy transfer reference value is obtained by performing amplitude normalization and phase synchronization processing on the resonance current and voltage waveforms collected in the test phase, extracting the peak points and phase difference of the voltage and current signals within the same resonance period, calculating the instantaneous power change rate according to the time coincidence degree of the two, and performing periodic average operation on the power change rate sequence.

[0014] The design system for reducing the distributed capacitance of a high-frequency drive transformer comprises: The capacitor coupling module collects the instantaneous voltage and current of the main and auxiliary windings of the transformer, calculates the phase difference between the voltage and the current and fits the energy density curve, analyzes the coupling strength of the capacitor region, generates the capacitor coupling data and transmits the capacitor coupling data to the capacitor extraction module; The capacitor extraction module extracts the instantaneous voltage response of the capacitor region based on the capacitor coupling data, smoothes the instantaneous voltage response, iteratively calculates the deviation from the energy density curve, analyzes the spatial position of the capacitor region, generates the capacitor distribution adjustment set and transmits the capacitor distribution adjustment set to the capacitor layout module; The capacitor layout module detects the current phase difference and energy transfer path of the main and auxiliary windings based on the capacitor distribution adjustment set, calculates the distribution capacitor concentration of multiple windings, judges the deviation trend of the capacitor layout and adjusts the coordinates of the capacitor layout in the magnetic core window, generates the capacitor layout set and transmits the capacitor layout set to the structure adjustment module; The structure adjustment module obtains the current amplitude and energy transfer ratio of multiple capacitor regions under high-frequency resonance based on the capacitor layout set, calculates the phase deviation of the distribution capacitor and compares the phase deviation with the energy transfer ratio, adjusts the capacitor capacity and spacing ratio, generates the structure configuration set and transmits the structure configuration set to the optimization design module; The optimization design module collects the resonance current and voltage waveform in the test stage based on the structure configuration set, calculates the energy feedback of the distribution capacitor, compares the error with the capacitor coupling data and corrects the capacitor position and capacity, and generates the distribution capacitor optimization design result.

[0015] Compared with the prior art, the application has the advantages and positive effects that: In the application, by collecting the instantaneous voltage and current of the main and auxiliary windings and fitting the energy density curve, the coupling strength of the capacitor region can be dynamically analyzed to realize quantitative identification of the energy coupling process. On this basis, the instantaneous voltage response of the capacitor region is extracted and smoothed, and the capacitor spatial position is located by iteratively correcting the deviation, so as to realize accurate identification and adjustment of the distribution capacitor. Further, by detecting the current phase difference and energy transfer path, the distribution capacitor concentration between windings is judged and optimized to make the capacitor layout balanced. By comparing and analyzing the current amplitude and energy transfer ratio under high-frequency resonance, the capacitor phase deviation and energy transmission matching relationship are obtained, the capacitor capacity and spacing ratio are optimized, and the dynamic balance of the electromagnetic field coupling state is realized. In the test stage, the energy feedback of the resonance current and voltage waveform is compared and corrected, so that the capacitor parameters adaptively tend to be optimal. The overall processing logic forms a closed-loop control mechanism from data acquisition to capacitor distribution optimization, so that the electric field distribution is more uniform, the energy transfer efficiency is improved, the parasitic effect is reduced, the stability and anti-interference ability of the system under high-frequency conditions are enhanced, and the energy conversion quality and reliability of the structure design of the high-frequency driving transformer are improved. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiments description. Obviously, the drawings in the following description only show some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained based on these drawings without any creative effort.

[0017] Figure 1 For the step flowchart of the present application; Figure 2 For the S1 refinement diagram of the present application; Figure 3 For the S2 refinement diagram of the present application; Figure 4 For the S3 refinement diagram of the present application; Figure 5 For the S4 refinement diagram of the present application; Figure 6 For the S5 refinement diagram of the present application; Figure 7 For the system module diagram of the present application. DETAILED DESCRIPTION

[0018] The technical solutions in the present application will be described below in combination with the drawings.

[0019] In the embodiments of the present application, the words such as "example", "for example" and the like are used to represent as an example, illustration or description. Any embodiment or design scheme described as "example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the word "example" is intended to present the concept in a specific manner. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0020] In the embodiments of the present application, "image" and "picture" can be used interchangeably at times. It should be pointed out that when the distinction is not emphasized, the meanings expressed are consistent. "Of", "corresponding" and "corresponding" can be used interchangeably at times. It should be pointed out that when the distinction is not emphasized, the meanings expressed are consistent.

[0021] In the embodiments of the present application, sometimes the subscript such as W1 can be written in the form of non-subscript such as W1. When the distinction is not emphasized, the meanings expressed are consistent.

[0022] In order to make the technical problems, technical solutions and advantages of the present application more clear, the following will be described in detail in combination with the drawings and specific embodiments.

[0023] Please refer toFigure 1 The embodiment of the present application provides a design method for reducing distributed capacitance of a high-frequency driving transformer, comprising the following steps: S1: collecting instantaneous voltage and current of main and auxiliary windings of the transformer, calculating phase difference of the voltage and the current and fitting into an energy density curve, analyzing coupling strength of a capacitance region, and generating capacitance coupling data; S2: based on the capacitance coupling data, extracting instantaneous voltage response of the capacitance region and smoothing, iteratively calculating deviation from the energy density curve, analyzing spatial position of the capacitance region, and generating a capacitance distribution adjustment set; S3: based on the capacitance distribution adjustment set, detecting current phase difference and energy transmission path of the main and auxiliary windings, calculating concentration of distributed capacitance of multiple windings, judging deviation trend of the capacitance layout and adjusting coordinates in a magnetic core window, and generating a capacitance layout set; S4: based on the capacitance layout set, obtaining current amplitude and energy transmission ratio of multiple capacitance regions under high-frequency resonance, calculating phase deviation of the distributed capacitance and comparing with the energy transmission ratio, adjusting capacitance capacity and spacing ratio, and generating a structure configuration set; S5: based on the structure configuration set, collecting resonance current and voltage waveforms in a test stage, calculating energy feedback of the distributed capacitance, comparing with error of the capacitance coupling data and correcting capacitance position and capacity, and generating a distributed capacitance optimization design result.

[0024] The capacitance coupling data comprises instantaneous voltage response, phase difference and energy density curve of the capacitance region, the capacitance distribution adjustment set comprises smoothed voltage data, energy density curve deviation and spatial position of the capacitance region, the capacitance layout set comprises winding current phase difference, energy transmission path and winding distributed capacitance concentration, the structure configuration set comprises current amplitude, energy transmission ratio and capacitance phase deviation of the capacitance region under high-frequency resonance, and the distributed capacitance optimization design result comprises capacitance position correction value, capacitance capacity correction value and energy feedback error parameter.

[0025] Please refer to Figure 2 The specific steps of S1 are as follows: S101: collecting instantaneous voltage and current signals of main and auxiliary windings of the transformer and matching, calculating phase angle difference, arranging phase difference results in a same time period in time sequence, and normalizing and denoising the phase difference sequence to generate phase difference sequence data; For a 110 kV three-phase oil-immersed power transformer, the instantaneous voltage and current signals of the main and auxiliary windings were synchronously collected by high-frequency sensors installed on the bushings of the transformer high-voltage side (main winding) and low-voltage side (auxiliary winding) at a sampling frequency of 10 kHz. At a certain time point t1, the collected instantaneous voltage of the main winding was 105.2 kV, and the instantaneous current was 98.7 A; the instantaneous voltage of the auxiliary winding was 10.1 kV, and the instantaneous current was 1012.5 A. These signals with precise time stamps were matched to ensure that each data point was strictly aligned in time. The process of calculating the phase angle difference first determined the phase of each signal. Taking the power frequency of 50 Hz as an example, the signal period is 20 ms. By detecting the center time points of the voltage and current signals passing through zero twice in a row, the zero-crossing points are determined. Within a sampling period, the main winding voltage signal crosses zero at time point t = 2.05 ms, while the main winding current signal crosses zero at t = 2.23 ms, with a time difference of 0.18 ms. This time difference corresponds to the main winding voltage and current phase difference, which is calculated as (0.18 ms / 20 ms) x 360° = 3.24°. Similarly, if the time difference of the zero-crossing points of the auxiliary winding voltage and current signals is 0.15 ms, then the phase difference is (0.15 ms / 20 ms) x 360° = 2.70°. The phase difference is the difference between the two phase differences, i.e., 3.24° - 2.70° = 0.54°. This calculation is repeated for consecutive sampling points to form a phase difference result time sequence containing 1000 data points, for example, the first five values of the sequence are [0.54°, 0.55°, 0.53°, 0.56°, 0.54°]. Subsequently, the sequence is normalized. First, find the maximum and minimum values in the sequence of 1000 data points. Set the maximum value in the measured sequence to 1.88°, and the minimum value to 0.21°. For the first data point 0.54° in the sequence, the normalized calculation is (0.54 - 0.21) / (1.88 - 0.21) = 0.1976. The same calculation is performed for each data point in the sequence to generate a normalized phase difference sequence. The denoising process uses a 5-point moving average method. For a certain point in the normalized sequence and its two adjacent points, a total of five points, the arithmetic mean value is calculated, and the average value is taken as the new value of the point. For example, for the third point in the normalized sequence and its adjacent point values [0.1976, 0.2036, 0.1916, 0.2096, 0.1976], the new value of the third point after denoising is (0.1976 + 0.2036 + 0.1916 + 0.2096 + 0.1976) / 5 = 0.2000. This process is applied to the entire normalized sequence to generate a smooth phase difference sequence data. As shown in Table 1, some of the original collected signals and processed phase difference sequence data are shown.

[0026] Table 1 Transformer signal collection and processing data sample table Time stamp Main winding voltage Main winding current Initial phase difference Normalized value Sequence value after denoising 1.0 105.2 98.7 0.54 0.1976 0.1995 1.1 106.1 99.2 0.55 0.2036 0.2000 1.2 104.9 98.5 0.53 0.1916 0.2008 1.3 106.5 99.6 0.56 0.2096 0.2015 1.4 105.3 98.8 0.54 0.1976 0.2021 Table 1 lists the signal acquisition values in a short period of time and the final phase difference sequence data generated after matching, calculation, normalization and denoising.

[0027] S102: Based on the phase difference sequence data, the instantaneous energy density distribution value is calculated by point-by-point multiplication of the phase difference change rate and the voltage and current amplitude change rate in all time windows, and the energy density distribution values of all time windows are curve fitted to obtain an energy density fitting curve set; Based on the generated phase difference sequence data, the voltage and current amplitude data collected synchronously are combined to perform the calculation of the instantaneous energy density distribution value. First, the time window required for calculation is determined. The width of the time window is set to 100 milliseconds, which is determined by experimental verification. The experimental process includes applying simulated internal winding inter-turn short circuit and slight insulation degradation and other transient disturbances to the target model 110kV transformer, and observing the time when the electrical characteristics fully develop and tend to be stable. Through analysis of the data of 50 different disturbance experiments, it is found that the complete development time of the disturbance characteristics is between 80 and 120. Therefore, 100 milliseconds is selected as the time window, which can completely capture the transient event and avoid introducing too much steady-state background data. Within a 100 millisecond time window, the phase difference rate of change is first calculated. Take the denoised phase difference sequence generated in S101, and calculate the difference between adjacent two data points. For example, at time points t1 and t2 (time interval is 0.1ms), the phase difference sequence values are 0.2000 and 0.2003 respectively, and the phase difference rate of change at this point is 0.2003-0.2000=0.0003. Similarly, the rate of change of the voltage amplitude and the current amplitude in the time window is calculated. The voltage amplitude takes the corresponding time point voltage effective value, and if the voltage effective values at t1 and t2 are 10.15kV and 10.17kV respectively, then the voltage amplitude rate of change is 10.17-10.15=0.02kV. If the current effective values at t1 and t2 are 1015.5A and 1015.2A respectively, then the current amplitude rate of change is 1015.2-1015.5=-0.3A. Next, a point-by-point product calculation is performed at each time point to obtain the instantaneous energy density distribution value. At time point t2, the value is the product of the phase difference rate of change, the voltage amplitude rate of change and the current amplitude rate of change, i.e. 0.0003×0.02×(-0.3)=-0.0000018. This calculation is applied to all data points in the entire 100 millisecond time window to generate a sequence of instantaneous energy density distribution values. For example, a window may generate a sequence of 1000 energy density values, such as [-0.0000018, 0.0000021, -0.0000015,...]. Finally, the energy density distribution values of all time windows are curve fitted. The 1000 energy density distribution value sequences generated in each time window are regarded as a group of discrete points, and a smooth energy density distribution curve is generated using the cubic spline interpolation method. This method constructs a piecewise cubic polynomial, so that the curve has second-order derivative continuity at the connection point. This process is independently performed on each 100 millisecond time window divided from the continuous collected data, thereby obtaining a series of energy density distribution curves evolving over time, which together constitute the energy density fitting curve set.

[0028] S103: Call the energy density fitting curve set, screen the energy density peak section in the curve, extract the corresponding phase difference and voltage current amplitude ratio, calculate the energy concentration parameter of the multi-peak section, analyze the coupling ratio of the energy concentration parameter and the phase difference distribution, analyze the coupling strength, and generate the capacitive coupling data; The generated energy density fitting curve set is called, and the energy density peak section in the curve is screened. The screening basis is a preset energy density peak threshold. The setting process of the threshold is as follows: collect transformer operation data under two working conditions, 50 groups for each working condition, and each group of data has a time length of 1 second. The first working condition is normal steady-state operation and conventional operation (such as tap switch switching), and the second working condition is a simulated internal capacitive coupling fault. All data are processed by S101 and S102 to calculate the energy density curve. In 100 groups of data, the maximum energy density under normal working condition is 3.2×10⁻ 5 , and the minimum energy density under fault working condition is 8.1×10⁻ 5 when the peak value appears. To ensure effective differentiation, the peak value screening threshold is set to 5.0×10⁻ 5 , which is between the normal maximum value and the fault minimum value. When screening, each curve in the energy density fitting curve set is traversed, and the continuous part of the curve with a value greater than 5.0×10⁻ 5 is identified as an energy density peak section. For example, a certain curve has a section with continuous values greater than 5.0×10⁻ 5 , and the highest point reaches 9.5×10⁻ 5The segment is then filtered out. Subsequently, the original phase difference value and the voltage-to-current amplitude ratio at the time point corresponding to the peak segment are extracted. If a peak segment corresponds to a time range of t = 50.2 ms to t = 55.8 ms, the phase difference sequence data generated by S101 in the time range, and the ratio of the voltage effective value to the current effective value at the same time are extracted. For example, at t = 53.5 ms, the phase difference is 1.25°, the voltage effective value is 10.2 kV, and the current effective value is 980 A, so the voltage-to-current amplitude ratio is 10200 V / 980 A = 10.41. Next, the energy concentration parameter of multiple peak segments is calculated. For a curve with peak segments, the energy density values in all peak segments are summed to obtain the total peak energy. At the same time, the energy density values of all points on the curve are summed to obtain the total window energy. The energy concentration parameter is the ratio of the total peak energy to the total window energy. If the total peak energy of a curve is set to 0.0085 and the total window energy is 0.0105, the energy concentration parameter is 0.0085 / 0.0105 = 0.8095. Then, the coupling ratio of the energy concentration parameter and the phase difference distribution is analyzed. The phase difference distribution is quantified by calculating the standard deviation of all phase difference values in the peak segment. If the phase difference data corresponding to a peak segment is [1.25°, 1.28°, 1.24°, 1.26°], the standard deviation is calculated as 0.017. The coupling ratio is defined as the energy concentration parameter divided by the phase difference standard deviation, i.e. 0.8095 / 0.017 = 47.62. Finally, the coupling strength is analyzed. The division standard of coupling strength is established by calculating the coupling ratio of the above 100 groups of experimental data. The results show that the coupling ratio of normal working conditions is less than 15, and the coupling ratio of fault working conditions is more than 40. Accordingly, the coupling strength interval is set: the coupling ratio less than 15 is "weak coupling", 15 to 40 is "moderate coupling", and more than 40 is "strong coupling". The coupling ratio obtained in the current example is 47.62, which belongs to "strong coupling". The generated capacitive coupling data is a structured record, containing: {energy concentration: 0.8095, phase difference standard deviation: 0.017, coupling ratio: 47.62, coupling strength: strong}.

[0029] Please refer to Figure 3 The specific steps of S2 are: S201: Based on the capacitive coupling data, extract the instantaneous voltage response signal in the capacitive region, perform interval average operation on the time series data of the voltage response signal, smooth the fluctuation components of the continuous sampling segments, and reorder them by time to generate a smoothed voltage response sequence; Based on the generated capacitive coupling data, the event showing "strong" coupling is extracted. The corresponding time range of this event is t = 50.2 ms to t = 55.8 ms. The instantaneous voltage response signal of the auxiliary winding is extracted from the raw sampling in this time range, forming a time series data containing 56 data points, for example, the first five values of the sequence are [10.21 kV, 10.23 kV, 10.19 kV, 10.25 kV, 10.24 kV]. The interval average operation is performed on the time series data of the voltage response signal. The interval width of the operation is set to 5 sampling points, i.e. 0.5 ms. The basis for setting this width is to analyze the main frequency of the high-frequency noise component in the voltage response signal of the 110 kV transformer under simulated capacitive fault, and it is found that most of the noise frequencies are higher than 2 kHz, and their periods are less than 0.5 ms. Therefore, the average interval of 0.5 ms is selected. For the first interval in the sequence (the first to the fifth data point), its value is [10.21, 10.23, 10.19, 10.25, 10.24] kV, and the arithmetic mean value is (10.21 + 10.23 + 10.19 + 10.25 + 10.24) / 5 = 10.224 kV. This result is taken as the first point of the new sequence, corresponding to the midpoint time of the original interval. Repeat this operation for all subsequent non-overlapping 5-point intervals until all data points are processed, generating a new sequence with a length of 11 data points. Next, the fluctuation component of the continuous sampling segment is smoothed and corrected. This process uses the weighted moving average method with a window size of 3 points. The weight coefficients are set based on experimental data, and the smoothing effect of 50 groups of known fault signals is evaluated to set the center point weight to 0.6 and the adjacent point weight to 0.2. This set of weights [0.2, 0.6, 0.2] retains the main trend while suppressing residual short-term fluctuations. For a point in the sequence after interval averaging and its two adjacent points, for example, the values of the second point in the sequence and its adjacent points are [10.224, 10.268, 10.255] kV, then the corrected value of the point is 10.224 * 0.2 + 10.268 * 0.6 + 10.255 * 0.2 = 10.2586 kV. Apply this calculation to the entire interval-averaged sequence. Finally, the smoothed voltage values are reordered according to their corresponding time stamps. Since all the above processing procedures are performed in chronological order, the generated sequence maintains the chronological order. This sequence is the smoothed voltage response sequence.

[0030] S202: Call the smoothed voltage response sequence, perform deviation calculation and iterative accumulation on the corresponding amplitude of the voltage amplitude and energy density curve for all time periods, analyze the deviation value distribution of the voltage response and the energy density, and perform interval normalization on the deviation distribution to obtain an energy deviation distribution set; The generated smooth voltage response sequence is called, for example, at time point t = 53.5 ms, the corresponding smooth voltage amplitude is 10.2601 kV. At the same time, the energy density fitting curve generated by S102 is called, at the same time point t = 53.5 ms, the corresponding energy density amplitude is 9.5 x 10⁻ 5 Before performing the deviation calculation, the two kinds of data are normalized. The normalization of the smooth voltage response sequence first needs to determine its maximum and minimum values in the entire time period, set to 10.35 kV and 10.18 kV, respectively, then the normalized value of 10.2601 kV is (10.2601-10.18) / (10.35-10.18)=0.4712. The normalization of the energy density curve, also find its maximum and minimum value, set to 9.8 x 10⁻ 5 and 0.5 x 10⁻ 5 , respectively, then the normalized value of 9.5 x 10⁻ 5 is (9.5-0.5) / (9.8-0.5)=0.9677. The deviation value of this point is the difference between the two: 0.4712-0.9677=-0.4965. Repeat this deviation calculation for all corresponding data points in the entire time period, and perform iterative accumulation. Set the initial accumulated deviation value to 0. At the first time point, the calculated deviation value is -0.4520, so the accumulated deviation is 0+(-0.4520)=-0.4520. At the second time point, the calculated deviation value is -0.4965, so the new accumulated deviation is -0.4520+(-0.4965)=-0.9485. This process continues until the last data point in the time period is processed, forming a cumulative deviation sequence. Then, analyze the deviation value distribution of the voltage response and the energy density. By checking the entire cumulative deviation sequence, determine its distribution range. Set the minimum value of the cumulative deviation sequence to -15.82 and the maximum value to 1.25 in the analyzed time period. This indicates that in most of the time, the relative increase in energy density exceeds that of the voltage response. Finally, interval normalize the deviation distribution. Use the maximum value 1.25 and the minimum value -15.82 found above to process each value in the cumulative deviation sequence. For the cumulative deviation value -0.9485, the normalized value is (-0.9485-(-15.82)) / (1.25-(-15.82))=14.8715 / 17.07=0.8712. Perform this calculation for all points in the cumulative deviation sequence to generate the energy deviation distribution set within the interval distribution.

[0031] S203: According to the energy deviation distribution set, perform clustering calculation on the spatial distribution coordinates of the deviation amplitude in the multi-capacitance region, identify the spatial coordinate range of the deviation concentrated area according to the clustering result, and integrate the multi-region coordinates and the corresponding deviation amplitude, to generate a capacitance distribution adjustment set; According to the generated energy deviation distribution set, spatial coordinate information is introduced. The transformer high-voltage winding is uniformly arranged with 5 monitoring points along its axial direction (Z axis), and the coordinates are Z=0.5m, 1.0m, 1.5m, 2.0m and 2.5m respectively. Each monitoring point is subjected to the complete processing flow of S101 to S202 to generate its own energy deviation distribution set. Now the deviation amplitude in the multi-capacitance region is processed, that is, the monitoring points judged as "strong coupling" in S103 are selected. The three monitoring points located at Z=1.0m, Z=1.5m and Z=2.5m appear "strong coupling" events, and the maximum deviation amplitudes (after normalization) in the corresponding energy deviation distribution set are 0.92, 0.95 and 0.88 respectively. Cluster calculation is performed on the spatial distribution coordinates of these three deviation central points. The clustering basis is the Euclidean distance between points. In this one-dimensional coordinate case, it is the absolute value of the coordinate difference. The distance threshold of clustering is set to 0.6m. The setting of this threshold is based on the analysis of the physical structure of the 110kV transformer winding. The typical height of a single winding cake is about 0.4-0.5m, so a threshold of 0.6m can cluster the abnormal points occurring on the same or adjacent winding cakes into a class. The distance between monitoring point 1 (Z=1.0m) and monitoring point 2 (Z=1.5m) is |1.5-1.0|=0.5m, which is less than the threshold 0.6m, so they are classified into a class (cluster A). The distance between monitoring point 1 (Z=1.0m) and monitoring point 3 (Z=2.5m) is |2.5-1.0|=1.5m, which is greater than the threshold 0.6m. The distance between monitoring point 2 (Z=1.5m) and monitoring point 3 (Z=2.5m) is |2.5-1.5|=1.0m, which is greater than the threshold 0.6m. Therefore, the clustering result is: cluster A contains {Z=1.0m, Z=1.5m}, and cluster B contains {Z=2.5m}. According to the clustering result, the spatial coordinate range of the deviation central region is identified. The spatial coordinate range of cluster A is Z from 1.0m to 1.5m. The spatial coordinate range of cluster B is only Z=2.5m. Finally, the multi-region coordinates and the corresponding deviation amplitudes are integrated by weighting. For each cluster, the weighted center coordinate is calculated. The weight is the maximum deviation amplitude corresponding to the point. For cluster A, it contains two points, and the weighted center coordinate is (1.0m x 0.92+1.5m x 0.95) / (0.92+0.95)=(0.92+1.425) / 1.87=2.345 / 1.87=1.254m. For cluster B, since it only contains one point, the center coordinate is the point coordinate Z=2.5m. The generated capacitance distribution adjustment set is shown in Table 2.

[0032] Table 2 Capacitance distribution adjustment set table Cluster number Monitoring point coordinate (m) included Maximum deviation amplitude Weighted center coordinate (m) A 1.0，1.5 0.92，0.95 1.254 B 2.5 0.88 2.500 As shown in Table 2, the capacitance distribution adjustment set lists the identified deviation cluster area (cluster), the center position of each area, and the original monitoring point information constituting the area.

[0033] Referring to Figure 4 The specific steps of S3 are as follows: S301: Based on the capacitance distribution adjustment set, the current signals of the main and auxiliary windings in the same time period are detected, the instantaneous amplitude of the current signal sequence is differentially calculated to extract the phase angle difference, and then the phase difference change rate is interval-averaged according to the time sequence to generate a winding phase difference sequence; Based on the generated capacitor distribution adjustment set, one of the deviation set areas is selected for analysis. Taking cluster A in Table 2 above as an example, the weighted center coordinates are Z = 1.254 m, and the corresponding time period is t = 50.2 ms to t = 55.8 ms. In this time period, the signals of the current sensors installed at the corresponding positions of the main winding and the auxiliary winding of the transformer (near Z = 1.254 m) are detected. At t = 51.0 ms, the instantaneous current of the main winding is 101.5 A, and the instantaneous current of the auxiliary winding is 1035.2 A. The two current signal sequences are processed to extract the phase angle difference. This is achieved by identifying the time difference between the zero-crossing points of the two signals. Within one power frequency cycle (20 ms), if the main winding current signal changes from negative to positive at t = 51.35 ms, and the auxiliary winding current signal changes from negative to positive at t = 51.41 ms, the time difference between them is 0.06 ms. The phase angle difference corresponding to this time difference is calculated as (0.06 ms / 20 ms) x 360° = 1.08°. Repeat this calculation for all sampling points in the entire time period (t = 50.2 ms to t = 55.8 ms) to obtain a phase difference time sequence. For example, the first five values of the sequence are [1.08°, 1.10°, 1.07°, 1.11°, 1.09°]. Then, the phase difference change rate is calculated based on the time sequence, and the change rate is subjected to interval averaging operation. First, the difference between adjacent two phase difference data points is calculated as the instantaneous change rate. For the second and first points in the sequence, the change rate is 1.10°-1.08° = 0.02°. This calculation is applied to the entire phase difference sequence to obtain a change rate sequence. Then, the interval average of the change rate sequence is calculated. The average interval width is set to 5 sampling points (0.5 ms). The basis for setting this width is that in the physical simulation of internal capacitor coupling faults of 110 kV transformers, the significant change period of energy transfer mode is not less than 0.5 ms. For the first interval of the change rate sequence, the values are [0.02°, -0.03°, 0.04°, -0.02°, 0.03°], and the arithmetic mean is calculated as (0.02-0.03+0.04-0.02+0.03) / 5 = 0.008°. This average value is taken as the first point of the new sequence. Move along the entire change rate sequence, generate a winding phase difference sequence.

[0034] S302: Call the winding phase difference sequence, calculate the energy transfer direction vector for the current direction and amplitude of the multi-phase difference corresponding time, and aggregate the energy transfer vectors of consecutive time periods in chronological order to obtain a path trajectory, identify the multi-direction mutation points of the path trajectory and coordinate fitting to obtain an energy transfer path set; The generated winding phase difference sequence is called, and combined with the current direction and amplitude collected at the same time. At a specific time, for example, t = 51.5ms, the main winding instantaneous current is -85.7A, and the auxiliary winding instantaneous current is -875.4A. At this time, the current direction is negative. The energy transfer direction vector is defined as a two-dimensional vector (main winding instantaneous current, auxiliary winding instantaneous current), that is, (-85.7, -875.4). The angle and length of the vector in the two-dimensional current space represent the direction and intensity of the energy transfer between the two windings. All energy transfer vectors calculated in a continuous time period (t = 50.2ms to t = 55.8ms) are aggregated in chronological order to form a path trajectory of energy transfer. The trajectory shows the dynamic process of energy exchange over time in the two-dimensional current space. For example, the path point sequence is (-85.7, -875.4), (-65.2, -665.9), (25.1, 258.3). Next, identify the multi-direction mutation points in the path trajectory. The identification criterion for the mutation point is that the angle between the energy transfer direction vectors of the two consecutive time points changes more than a preset angle threshold. The threshold is set by experimental data. Analyzing 50 groups of simulated internal capacitance fault transformer data, calculating the energy transfer path, it is found that under all fault conditions, the maximum change rate of the vector angle is greater than 30°, while under normal operating conditions, the value does not exceed 8°. Therefore, the angle threshold of the direction mutation is set to 20°. In the above path point sequence, the angle of vector (-65.2, -665.9) is about 264.4°, and the angle of vector (25.1, 258.3) is about 84.4°, the angle change between them is |84.4-264.4|=180°. The value is much larger than 20°, so point (25.1, 258.3) is identified as a mutation point. Traverse the entire path trajectory to find all mutation points that meet this condition. Fit the coordinates of all identified mutation points. Here, coordinate fitting refers to recording the coordinates (main winding current, auxiliary winding current) of all mutation points in the two-dimensional current space to form a set. This set represents the key state points where energy exchange changes dramatically during the fault. The obtained energy transfer path set is an ordered list of these mutation point coordinates, for example: (25.1, 258.3), (95.4, -960.1).

[0035] S303: According to the energy transfer path set, calculate the concentration of the distributed capacitance in the winding in the path direction, extract the spatial coordinates of the region whose concentration exceeds the concentration threshold, and perform proportional adjustment and remapping to the magnetic core window space coordinates according to the coordinate offset direction and the concentration amplitude, to generate a capacitance layout set; According to the generated set of energy transmission paths and the determined concentrated region spatial coordinates (e.g. the center Z = 1.254 m of cluster A), the concentration of the region distribution capacitance in the path direction is calculated. The calculation process of the concentration is as follows: first, the number of identified mutation points in the energy transmission paths belonging to the cluster A region within the analysis time period (t = 50.2 ms to t = 55.8 ms) is counted, which is set to 12. Second, the average modulus of the energy transmission vectors at these mutation points is calculated. The modulus of each vector is the square root of the sum of the squares of the primary winding current and the secondary winding current. If the sum of the moduli of the 12 mutation point vectors is 12600, then the average modulus is 12600 / 12 = 1050. The concentration is defined as the product of the number of mutation points and the average modulus, i.e. 12 x 1050 = 12600. Threshold judgment is performed on the calculated concentration. The setting of the concentration threshold is based on the analysis of transformer design specifications and historical failure data. 100 cases of historical data are selected, of which 50 are normal operation and 50 are capacitive fault. The above calculation process is performed on these data, and the maximum concentration under normal operation is 8000, while the minimum concentration under fault is 15000. Accordingly, the concentration threshold is set to 10000. The currently calculated concentration 12600 exceeds the threshold, so the region needs to be adjusted. The spatial coordinates of the region are extracted, i.e. Z = 1.254 m. Next, proportional adjustment is performed according to the coordinate offset direction and the concentration amplitude. The coordinate offset direction is predefined according to the transformer structure. For the problem of excessive inter-turn capacitance of the winding, the offset direction is set to increase the inter-turn insulation distance of the winding, i.e. along the radial direction of the winding outward. The adjustment amplitude (i.e. the amount of insulation increase) is proportional to the degree of concentration exceeding the threshold. The setting of the proportional adjustment coefficient is based on finite element simulation experiments: in the model, for every 0.1 mm increase in inter-turn insulation distance, the calculated concentration can be reduced by about 2000. Therefore, the proportional coefficient is set to 0.1 mm / 2000 = 0.00005 mm / unit concentration. The amount of concentration exceeding the threshold is 12600-10000 = 2600. The required adjustment amplitude is 2600 x 0.00005 mm = 0.13 mm. Finally, this adjustment amount is mapped back to the three-dimensional spatial coordinates of the magnetic core window. The original coordinates are Z = 1.254 m, and the radial coordinates are set to R = 0.8 m. The adjustment instruction is to increase the inter-turn insulation thickness of the winding located at (Z = 1.254 m, R = 0.8 m) by 0.13 mm. The same calculation is performed on all regions that exceed the threshold to generate the set of capacitance layout adjustment instructions.

[0036] Table 3 Set of Capacitance Layout Adjustment Instructions Region center coordinate (Z axis, m) Calculate concentration Adjustment direction Suggested adjustment amount 1.254 12600 Increase insulation radially outward 0.13 2.500 9500 No 0.00 See Figure 5 , the specific steps of S4 are: S401: Based on the set of capacitance layouts, obtain the current signal of the multi-capacitance region under high-frequency resonance conditions, calculate the current amplitude in the same resonance period, and analyze the energy transfer ratio sequence according to the current input and output energy of adjacent capacitance regions to generate a set of resonance energy ratios; Based on the generated set of capacitance layouts, select the region that needs to be adjusted, i.e. the region with cluster number A in Table 3, with the center coordinates at Z=1.254m. In order to obtain the current signal of this region under high-frequency resonance conditions, a sweep signal with a frequency of 5kHz and an amplitude of 50V is injected into the transformer winding. This frequency is selected according to the main resonance frequency range (4kHz to 6kHz) exhibited by 110kV transformers in turn-to-turn capacitance fault simulation. By installing current sensors at Z=1.254m (region A) and its physically adjacent Z=1.0m (region B) as the energy input reference point, the current signals are synchronously collected. In a resonance period (1 / 5kHz=0.2ms), the current amplitude is calculated. The maximum absolute value of the current signal in a certain period is determined. In a certain period, the maximum amplitude of the current signal of region B is 1.25A, and the maximum amplitude of the current signal of region A is 1.08A. At the same time, the energy transfer ratio is analyzed according to the current input and output energy of adjacent capacitance regions. In this scenario, region B is regarded as the energy input end, and region A is regarded as the energy output end. The energy transfer is quantified by the square ratio of current amplitude in a single resonance period. This method is based on the fact that in a pure resistance or resonance network, power is proportional to the square of current. The energy transfer ratio is calculated as the square of (output current amplitude / input current amplitude). Taking the above data as an example, the energy transfer ratio is (1.08A / 1.25A)²=0.864²=0.7465. Repeat this calculation for 10 consecutive resonance periods to obtain an energy transfer ratio sequence containing 10 data points, for example: [0.7465, 0.7481, 0.7459, 0.7502, 0.7477, 0.7495, 0.7468, 0.7488, 0.7510, 0.7480]. This sequence reflects the efficiency and stability of energy transfer from one region to the next under high-frequency excitation. Integrate this sequence to generate a set of resonance energy ratios.

[0037] S402: Call the set of resonance energy ratios, calculate the time offset between the current peak point and the voltage peak point for the current signal phase information of the multi-distributed capacitance region, and convert it to an angle difference as the phase offset value, and perform corresponding comparison with the energy transfer ratio to obtain a set of capacitance phase offset difference values; The generated resonance energy ratio set is called and combined with the voltage signal collected synchronously under the same high-frequency resonance condition. For the phase information of the current signal of the multi-distributed capacitance region, the time offset between the current peak point and the voltage peak point is calculated. Taking region A (Z = 1.254 m) as an example, in the first resonance period, through the data recorded by the high-frequency oscilloscope or the data acquisition card, it is identified that the voltage signal of the region reaches the peak at time t = 0.045 ms, while the current signal reaches the peak at t = 0.080 ms. The time offset between the two is 0.080 ms-0.045 ms = 0.035 ms. Convert this time offset to an angle difference as the phase offset value. The reference for conversion is the period of the resonance signal, which is 0.2 ms. The calculation of the angle difference is (time offset / resonance period) x 360°. Substituting the numerical value, the phase offset value is (0.035 ms / 0.2 ms) x 360° = 63°. This phase offset value reflects the degree of deviation of the capacitive characteristics of the distributed capacitance in the region under high frequency, and the phase offset of an ideal pure capacitor should be 90°. Compare the calculated phase offset value with the energy transfer ratio corresponding to the resonance period in S401. In the first period, the energy transfer ratio is 0.7465, and the phase offset value is 63°. In order to quantify this deviation, a reference phase offset value is introduced, which is set to 90° for an ideal capacitor. Calculate the difference between the actual value and the reference value, i.e. 90°-63° = 27°. This difference is called the phase offset difference value. For the energy transfer ratio sequence in S401, the phase offset difference value of each of the 10 resonance periods is calculated to form a phase offset difference value sequence. For example, the calculated sequence is [27°, 26°, 28°, 25°, 27°, 26°, 28°, 27°, 25°, 26°]. This sequence eventually forms the capacitance phase offset difference value set.

[0038] S403: According to the capacitance phase offset difference value set, the capacitance parameter and the adjacent spacing parameter of the multi-distributed capacitance are adjusted in two variables, the capacitance increase / decrease coefficient is corrected according to the positive / negative direction of the phase offset difference value, and the energy transfer ratio change is recalculated and balanced after adjustment to generate a structure configuration set; According to the generated set of phase shift difference values, the capacity parameters and adjacent spacing parameters of the multi-distributed capacitance are subjected to bivariate proportional adjustment. The average value in the set of phase shift difference values is selected as the adjustment basis, and the average value is (27+26+28+25+27+26+28+27+25+26) / 10=26.5°. The value is positive, indicating that the capacitive capacity of the region is insufficient, and the capacitance capacity needs to be increased. The positive and negative directions of the phase shift difference value are corrected according to the increase and decrease coefficient of the capacitance. The setting of the increase and decrease coefficient is determined through a series of finite element simulation experiments. The experiment is carried out on the 110kV transformer winding model, the inter-turn capacitance value is changed, and the phase shift change under high frequency is observed. Experimental data show that for every 1% increase in capacitance capacity, the phase shift difference value decreases by about 4°. Therefore, the increase and decrease coefficient of the capacitance is set to 1% / 4°=0.25% per degree. Based on the average phase shift difference value of 26.5°, the percentage of capacitance capacity that needs to be increased is 26.5°×0.25% / °=6.625%. At the same time, the capacitance capacity is inversely proportional to the spacing between the plates. To increase the capacitance by 6.625%, the adjacent inter-turn insulation spacing needs to be reduced. If the original design spacing is 2.0mm, then the new spacing is 2.0mm / (1+0.06625)=1.876mm. After adjustment, the amount of change in energy transfer ratio is calculated again through simulation and balance matching. The new spacing parameter 1.876mm is substituted into the simulation model, and the 5kHz resonance signal is applied again to calculate the adjusted energy transfer ratio. The simulation results show that the energy transfer ratio between region A and region B becomes 0.96. This process is used to verify the effectiveness of the adjustment. The benchmark for balance matching is a pre-set energy transfer ratio target interval, which is set by analyzing the resonance test data of 50 healthy transformers, and the energy transfer ratio of each transformer is between 0.95 and 1.05. The newly calculated energy transfer ratio 0.96 is within this interval, indicating that the adjustment meets the requirements of balance matching. The adjustment results are integrated to generate a structure configuration set.

[0039] Table 4 Final structure configuration set table Region center coordinate (Z axis, m) Original spacing Calculate adjustment amount Configure spacing Verify energy transfer ratio 1.254 2.000 -0.124 1.876 0.96 As shown in Table 4, the structure configuration set provides a complete adjustment path and verification results from the original design parameters to the final recommended configuration parameters, providing specific data support for the structural optimization of the transformer.

[0040] Please refer to Figure 6 , the specific steps of S5 are: S501: Based on the structure configuration set, the test stage resonance current and voltage waveforms are collected, time integration operation is performed according to the instantaneous product of voltage and current in the same resonance period, the difference between the energy input and the energy return of the multi-distributed capacitance node is calculated and sorted to obtain an energy feedback sequence; Based on the generated set of structural configurations, including the configuration spacing of 1.876 mm for Z = 1.254 m region in Table 4, the testing phase is entered. In this phase, the same 5 kHz, 50 V resonant excitation signal as in S401 is applied to the transformer prototype with the configuration adjustment, and the current and voltage waveform data at the Z = 1.254 m node are synchronously collected. Within one complete resonant cycle (0.2 ms), the instantaneous voltage and current values are collected with a time step of 0.01 ms. The time integration operation is performed according to the instantaneous product of voltage and current within the same resonant cycle. This operation is used to calculate the net energy exchange for the cycle. First, the cycle is divided into two parts: the energy input stage (instantaneous power is positive) and the energy return stage (instantaneous power is negative). For example, at one sampling point, the instantaneous voltage is 45.2 V and the instantaneous current is 1.15 A, so the instantaneous power is 45.2 V x 1.15 A = 51.98 W. At another sampling point, the instantaneous voltage is -25.8 V and the instantaneous current is 0.45 A, so the instantaneous power is -11.61 W. By summing the instantaneous powers of 20 sampling points within the entire cycle, the result of time integration can be approximated. The energy input is obtained by accumulating all positive power values, for example, 0.25 mJ. The energy return is obtained by accumulating the absolute values of all negative power values, for example, 0.20 mJ. The difference between the energy input and the energy return for the multiple distributed capacitance nodes is calculated. For the Z = 1.254 m node, the difference is 0.25 mJ - 0.20 mJ = 0.05 mJ. This positive value indicates that 0.05 mJ of net energy is absorbed or dissipated by the node within the resonant cycle. The same test and calculation are performed for other key distributed capacitance nodes, such as Z = 1.0 m and Z = 1.5 m. The energy difference calculated at the Z = 1.0 m node is 0.02 mJ, and the energy difference calculated at the Z = 1.5 m node is -0.01 mJ (indicating that the node outputs net energy). These difference values are sorted by their numerical values to obtain the energy feedback sequence: [-0.01 mJ, 0.02 mJ, 0.05 mJ].

[0041] S502: Call the energy feedback sequence to perform error determination operation on the capacity parameters of the multiple distributed capacitances and the adjacent coupling position parameters, compare the energy feedback with the energy transmission reference value, judge the deviation direction of the capacity and position according to the sign of the difference, and generate error deviation data; The generated energy feedback sequence [-0.01 mJ, 0.02 mJ, 0.05 mJ] is called, and an error judgment operation is performed on the capacity parameter of the distribution capacitor corresponding to each value in the sequence and the adjacent coupling position parameter. The core of this operation is to compare the energy feedback value of each node with a preset energy transfer reference value. The setting of the energy transfer reference value is as follows: 20 110kV transformers meeting the factory standard are subjected to the same 5kHz resonance test, and the average net energy loss of the key nodes in a resonance period is measured. The statistical data shows that the average net energy loss of these healthy transformers is 0.01 mJ, and the standard deviation is 0.005 mJ. Therefore, the energy transfer reference value is set to 0.01 mJ. If the energy feedback value of a node is in the interval [0.005 mJ, 0.015 mJ], it is judged to be "normal", and if it is out of this interval, it is judged to have an error. The energy feedback value is compared with the reference value. Take the Z=1.254m node as an example, its energy feedback value is 0.05mJ. Calculate the difference between it and the reference value: 0.05mJ-0.01mJ=0.04mJ. Determine the offset direction of the capacity and position according to the sign of the difference. The judgment logic is based on the following rules: if the difference is positive (such as 0.04mJ), it means that the actual energy loss is higher than the reference value. This is due to the additional resistive loss caused by the non-ideal capacitive, indicating that the current capacitor value is too low or the equivalent series resistance is too high. The correction direction is to further increase the capacitor capacity, that is, to reduce the turn-to-turn spacing. If the difference is negative (such as the energy feedback value of Z=1.5m node is -0.01mJ, the difference is -0.01mJ-0.01mJ=-0.02mJ), it means that the net energy is in the output state, which is much lower than the normal loss. This is due to the undesired resonance coupling with other parts, indicating that the current capacitor value is too high. The correction direction is to reduce the capacitor capacity, that is, to increase the turn-to-turn spacing. Based on this, the error offset data of Z=1.254m node is (offset direction: reduce spacing, difference size: 0.04mJ). The error offset data of Z=1.5m node is (offset direction: increase spacing, difference size: -0.02mJ). These data constitute the error offset data.

[0042] S503: According to the error offset data, the capacity value and the spatial position of the multi-distribution capacitor are corrected, the offset direction is taken as the correction symbol, and the correction scale is calculated according to the difference size, the corrected capacity parameter set and the position arrangement data set are updated, and the distribution capacitor optimization design result is generated; According to the generated error offset data, the capacity value and spatial position of the multi-distributed capacitor are finally corrected. The correction process takes the determined offset direction as the correction sign and calculates the correction scale according to the difference value. The calculation of the correction scale requires a correction coefficient. The coefficient is determined by parameter scanning experiment on the finite element electromagnetic field simulation model of the transformer winding. In the simulation, the inter-turn insulation spacing is adjusted by 0.001 mm step, and the corresponding single-cycle net energy loss change is recorded. Experimental data show that for every 0.001 mm change in spacing, the net energy loss value changes by about 0.008 mJ. Therefore, the correction coefficient is set to 0.001 mm / 0.008 mJ=0.125 mm / J or 0.000125 mm / mJ. The correction calculation is performed for the Z=1.254 m node. The difference value of the error offset data is 0.04 mJ, and the offset direction is "reduce spacing" (the correction sign is negative). Therefore, the correction scale of the spacing is -0.04 mJ×0.000125 mm / mJ=-0.005 mm. The spacing determined in S403 for the Z=1.254 m node is 1.876 mm. The final corrected spacing is 1.876 mm-0.005 mm=1.871 mm. The correction calculation is performed for the Z=1.5 m node. The difference value of the error offset data is -0.02 mJ, and the offset direction is "increase spacing" (the correction sign is positive). The correction scale of the spacing is |-0.02 mJ|×0.000125 mm / mJ=0.0025 mm. The original spacing is set to 2.0 mm, and the final corrected spacing is 2.0 mm+0.0025 mm=2.0025 mm. Update the corrected capacity parameter set and position arrangement data set. The capacity of the capacitor is inversely proportional to the spacing. For the Z=1.254 m node, set the capacity of the capacitor to 350.0 pF when the spacing is 1.876 mm, then update the capacity to 350.0 pF×(1.876 mm / 1.871 mm)=350.93 pF. The final correction parameters of all nodes are summarized to generate the distributed capacitor optimization design result.

[0043] Table 5 Distributed capacitor optimization design result table Region center coordinate (Z axis, m) Final revised spacing (mm) Final capacitance capacity (pF) 1.254 1.871 350.93 1.500 2.003 330.84 As shown in Table 5, the distributed capacitor optimization design result provides accurate structure parameters after final testing and fine tuning as the final basis for transformer manufacturing.

[0044] Please refer to Figure 7 , the design system for reducing the distributed capacitance of a high-frequency drive transformer, comprising: A capacitor coupling module acquires the instantaneous voltage and current of the main and auxiliary windings of the transformer, calculates the phase difference between the voltage and current and fits it into an energy density curve, analyzes the coupling strength of the capacitor region, generates capacitor coupling data and passes it to the capacitor extraction module; The capacitance extraction module extracts the instantaneous voltage response of the capacitance region and smoothes based on the capacitance coupling data, iteratively calculates the deviation from the energy density curve, analyzes the spatial position of the capacitance region, generates a capacitance distribution adjustment set and transmits to the capacitance layout module; The capacitance layout module detects the current phase difference of the main and auxiliary windings and the energy transmission path based on the capacitance distribution adjustment set, calculates the distribution capacitance concentration of multiple windings, judges the deviation trend of the capacitance layout and adjusts the coordinates in the magnetic core window, generates a capacitance layout set and transmits to the structure adjustment module; The structure adjustment module obtains the current amplitude and energy transmission ratio of multiple capacitance regions under high-frequency resonance based on the capacitance layout set, calculates the phase deviation of the distribution capacitance and compares it with the energy transmission ratio, adjusts the capacitance capacity and spacing ratio, generates a structure configuration set and transmits to the optimization design module; The optimization design module collects the resonance current and voltage waveform in the test stage based on the structure configuration set, calculates the energy feedback of the distribution capacitance, compares the error with the capacitance coupling data and corrects the capacitance position and capacity, and generates a distribution capacitance optimization design result.

[0045] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method of designing a high frequency drive transformer with reduced distributed capacitance, characterized by, The method comprises the following steps: S1: collecting the instantaneous voltage and current of the main and auxiliary windings of the transformer, calculating the phase difference of the voltage and current and fitting into an energy density curve, analyzing the coupling strength of the capacitor region, and generating capacitor coupling data; S2: based on the capacitor coupling data, extracting the instantaneous voltage response of the capacitor region and smoothing, iteratively calculating the deviation from the energy density curve, analyzing the spatial position of the capacitor region, and generating a capacitor distribution adjustment set; S3: based on the capacitor distribution adjustment set, detecting the current phase difference and energy transfer path of the main and auxiliary windings, calculating the concentration degree of the distributed capacitance of multiple windings, judging the deviation trend of the capacitor layout and adjusting the coordinates in the magnetic core window, and generating a capacitor layout set; S4: based on the capacitor layout set, obtaining the current amplitude and energy transfer ratio of multiple capacitor regions under high-frequency resonance, calculating the phase shift of the distributed capacitance and comparing it with the energy transfer ratio, adjusting the capacitance and spacing ratio, and generating a structure configuration set; S5: based on the structure configuration set, collecting the resonance current and voltage waveform in the test stage, calculating the energy feedback of the distributed capacitance, comparing the error with the capacitor coupling data and correcting the capacitor position and capacity, and generating a distributed capacitance optimization design result.

2. The design method of reducing distributed capacitance of a high-frequency drive transformer according to claim 1, characterized by, The capacitor coupling data includes the instantaneous voltage response, phase difference, and energy density curve of the capacitor region. The capacitor distribution adjustment set includes the smoothed voltage data, energy density curve deviation, and spatial position of the capacitor region. The capacitor layout set includes the winding current phase difference, energy transfer path, and winding distributed capacitance concentration degree. The structure configuration set includes the current amplitude, energy transfer ratio, and capacitor phase shift under high-frequency resonance of the capacitor region. The distributed capacitance optimization design result includes the capacitor position correction value, capacitor capacity correction value, and energy feedback error parameter.

3. The design method of reducing distributed capacitance of a high-frequency drive transformer according to claim 1, characterized by, The specific steps of S1 are as follows: S101: Collecting the instantaneous voltage and current signals of the main and auxiliary windings of the transformer and matching them, calculating the phase angle difference, arranging the phase difference results in the same time period in chronological order, and normalizing and denoising the phase difference sequence to generate phase difference sequence data; S102: Based on the phase difference sequence data, performing point-by-point product calculation on the phase difference change rate and voltage and current amplitude change rate in all time windows to obtain instantaneous energy density distribution values, and fitting the energy density distribution values of all time windows to obtain an energy density fitting curve set; S103: Calling the energy density fitting curve set, filtering the energy density peak section in the curve, extracting the corresponding phase difference and voltage and current amplitude ratio, calculating the energy concentration degree parameter of multiple peak sections, analyzing the coupling ratio of the energy concentration degree parameter and the phase difference distribution, analyzing the coupling strength, and generating capacitor coupling data.

4. The design method of reducing distributed capacitance of a high-frequency drive transformer according to claim 1, characterized by, The specific steps of S2 are as follows: S201: Based on the capacitor coupling data, extracting the instantaneous voltage response signal in the capacitor region, performing interval average operation on the time sequence data of the voltage response signal, smoothing the fluctuation components of the continuous sampling section, and reordering them by time to generate a smoothed voltage response sequence; S202: Call the smooth voltage response sequence, perform bias calculation and iterative accumulation for the corresponding amplitude of the voltage amplitude and energy density curve in all time periods, analyze the bias value distribution of voltage response and energy density, interval normalize the bias distribution, and obtain the energy bias distribution set; S203: According to the energy bias distribution set, perform clustering calculation on the spatial distribution coordinates of the bias amplitude in the multi-capacitance region, identify the spatial coordinate range of the bias concentration area according to the clustering result, and integrate the multi-region coordinates and the corresponding bias amplitude, generate the capacitance distribution adjustment set.

5. The design method of reducing distributed capacitance of a high-frequency drive transformer according to claim 1, characterized by, The specific steps of S3 are: S301: Based on the capacitance distribution adjustment set, detect the current signal of the main and auxiliary windings in the same time period, perform difference calculation on the instantaneous amplitude of the current signal sequence to extract the phase angle difference, and then perform interval average operation on the phase difference change rate according to the time sequence, to generate the winding phase difference sequence; S302: Call the winding phase difference sequence, calculate the energy transfer direction vector according to the current direction and amplitude of multiple phase differences at corresponding time, and aggregate the energy transfer vectors of consecutive time periods into path trajectories in time sequence, identify the multi-direction mutation points of the path trajectory and coordinate fitting, to obtain the energy transfer path set; S303: According to the energy transfer path set, calculate the concentration degree of the distributed capacitance in the multi-winding in the path direction, extract the spatial coordinates of the region whose concentration degree exceeds the concentration threshold, and perform proportional adjustment and remapping to the magnetic core window spatial coordinates according to the coordinate offset direction and concentration degree amplitude, to generate the capacitance layout set.

6. The design method of reducing distributed capacitance of a high-frequency drive transformer according to claim 5, wherein The concentration threshold is obtained by statistically analyzing the concentration degree parameters of the multi-winding distributed capacitance, obtaining the concentration degree value set of all capacitance regions, performing interval distribution analysis on the set to determine the mean and variance of the concentration degree, and then calculating the upper limit of the concentration degree distribution according to the mean and variance.

7. The design method of reducing distributed capacitance of a high-frequency drive transformer according to claim 1, characterized by, The specific steps of S4 are: S401: Based on the capacitance layout set, obtain the current signal of the multi-capacitance region under high-frequency resonance condition, calculate the current amplitude in the same resonance period, and analyze the energy transfer ratio sequence according to the current input and output energy of adjacent capacitance regions, to generate the resonance energy ratio set; S402: Call the resonance energy ratio set, calculate the time offset between the current peak point and the voltage peak point according to the current signal phase information of the multi-distributed capacitance region, convert it into an angle difference as a phase offset value, and perform corresponding comparison with the energy transfer ratio, to obtain the capacitance phase offset difference value set; S403: According to the capacitance phase offset difference value set, perform bivariate proportional adjustment on the capacitance parameters and adjacent spacing parameters of the multi-distributed capacitance, correct the capacitance increase and decrease coefficients according to the positive and negative directions of the phase offset difference value, and then re-calculate the energy transfer ratio change and perform balance matching, to generate the structure configuration set.

8. The design method of reducing distributed capacitance of a high-frequency drive transformer according to claim 1, characterized by, The specific steps of S5 are: S501: Based on the structure configuration set, collect the resonance current and voltage waveforms in the test phase, perform time integration operation according to the instantaneous product of voltage and current in the same resonance period, calculate the difference between the energy input and the energy return of the multi-distributed capacitance node and sort it, to obtain the energy feedback sequence; S502: calling the energy feedback sequence, performing error judgment operation on the capacity parameters and adjacent coupling position parameters of the multi-distributed capacitor, comparing the energy feedback with the energy transmission reference value, judging the deviation direction of the capacity and position according to the difference sign, and generating error deviation data; S503: according to the error deviation data, performing parameter correction on the capacity value and spatial position of the multi-distributed capacitor, taking the deviation direction as the correction sign and performing correction scale calculation according to the difference value, updating the corrected capacity parameter set and position arrangement data set, and generating the distributed capacitor optimization design result.

9. The design method of reducing distributed capacitance of a high-frequency drive transformer according to claim 8, characterized by, The energy transmission reference value is obtained by performing amplitude normalization and phase synchronization processing on the resonant current and voltage waveform collected in the test stage, extracting the peak point and phase difference of the voltage and current signal in the same resonant period, calculating the instantaneous power change rate according to the time coincidence degree of the two, and performing period average operation on the power change rate sequence.

10. A system for designing a reduced distributed capacitance high frequency drive transformer, characterized by, The system is used to realize the design method of reducing the distributed capacitance of the high-frequency drive transformer according to any one of claims 1-9, and the system comprises: The capacitor coupling module collects the instantaneous voltage and current of the transformer main and auxiliary windings, calculates the voltage and current phase difference and fits it into an energy density curve, analyzes the coupling strength of the capacitor region, generates capacitor coupling data and transmits it to the capacitor extraction module; The capacitor extraction module extracts the instantaneous voltage response of the capacitor region based on the capacitor coupling data and smoothes it, iteratively calculates the deviation from the energy density curve, analyzes the spatial position of the capacitor region, generates a capacitor distribution adjustment set and transmits it to the capacitor layout module; The capacitor layout module detects the current phase difference and energy transmission path of the main and auxiliary windings based on the capacitor distribution adjustment set, calculates the concentration of the distributed capacitance of multiple windings, judges the deviation trend of the capacitor layout and adjusts the coordinates in the magnetic core window, generates a capacitor layout set and transmits it to the structure adjustment module; The structure adjustment module obtains the current amplitude and energy transmission ratio of the multi-capacitor region under high-frequency resonance based on the capacitor layout set, calculates the phase deviation of the distributed capacitor and compares it with the energy transmission ratio, adjusts the capacitor capacity and spacing ratio, generates a structure configuration set and transmits it to the optimization design module; The optimization design module collects the resonant current and voltage waveform in the test stage based on the structure configuration set, calculates the energy feedback of the distributed capacitor, compares the error with the capacitor coupling data and corrects the capacitor position and capacity, and generates the distributed capacitor optimization design result.

Images