Method for assessing the aging status of epoxy resin insulation in dry-type transformers

Abstract

This invention provides a method for assessing the aging state of epoxy resin insulation in dry-type transformers. The method includes: acquiring frequency domain dielectric spectrum data of epoxy resin samples at different temperatures and aging times; fitting the frequency domain dielectric spectrum data using the Havriliak-Negami model to extract characteristic frequency parameters characterizing the insulation state; establishing a temperature correction model based on the characteristic frequency parameters of unaged samples and their corresponding measurement temperatures, and obtaining the initial activation energy of the epoxy resin; calculating the activation energy at each aging time based on the temperature correction model, the initial activation energy, the characteristic frequency parameters at different aging times, and their corresponding measurement temperatures, and obtaining the evolution curve of the activation energy changing with aging time; acquiring key feature points on the evolution curve, determining the threshold of the rate of change of activation energy relative to the initial activation energy of the material at each feature point, and constructing an aging assessment result including multiple aging levels. This achieves a quantitative assessment of the aging state.

Description

Technical Field

[0001] This invention relates to the technology of monitoring the insulation condition and diagnosing faults in power equipment, and more specifically, to a method for assessing the aging condition of epoxy resin insulation in dry-type transformers. Background Technology

[0002] Dry-type transformers, due to their compact structure, safety, reliability, and maintenance-free operation, have become core equipment in power distribution networks. Their long-term operational reliability primarily depends on the electrical strength, mechanical properties, and thermal stability of the key insulating material—epoxy resin. Under the long-term influence of multiple aging factors, the molecular structure of epoxy resin gradually deteriorates, leading to a decline in insulation performance and potentially even insulation breakdown, seriously threatening the safe and stable operation of the power system. Currently, assessment methods for epoxy resin insulation aging largely rely on traditional methods such as thermogravimetric analysis and dielectric spectroscopy testing. These methods are insufficient for quantitatively assessing its aging state. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to address the above-mentioned technical defects in the prior art by providing a method for evaluating the aging status of epoxy resin insulation in dry-type transformers. The method includes the following steps: Step S1: Obtain frequency domain dielectric spectrum data of epoxy resin samples at different temperatures and aging times; Step S2: Fit the frequency domain dielectric spectrum data using the Havriliak-Negami model to extract characteristic frequency parameters representing the insulation state; Step S3: Based on the characteristic frequency parameters of the unaged sample and its corresponding measurement temperature, establish a temperature correction model and obtain the initial activation energy of the epoxy resin. Step S4: Based on the temperature correction model, the initial activation energy, the characteristic frequency parameters at different aging times and their corresponding measurement temperatures, calculate the activation energy at each aging time and obtain the evolution curve of the activation energy with aging time. Step S5: Obtain key feature points on the evolution curve, determine the threshold of the rate of change of activation energy relative to the initial activation energy of the material at each feature point, and construct an aging assessment result including multiple aging levels based on the rate of change threshold.

[0004] Furthermore, the frequency domain dielectric spectrum data includes the measurement frequency and its corresponding dielectric loss value; Step S2 includes: The Havriliak-Negami model is fitted based on the measured frequency and its corresponding dielectric loss value to obtain the fitting curve of the dielectric loss spectrum, and the characteristic frequency parameter is extracted from the fitting curve; wherein, the characteristic frequency parameter is the peak frequency.

[0005] Furthermore, the Havriliak-Negami model includes: In the formula, For high frequency dielectric constant, For relaxation strength, For relaxation time, For the distribution width parameter, For asymmetric parameters, DC conductivity, DC conductance index parameter, ω is the angular frequency, and j is the imaginary unit.

[0006] Further, step S3 includes: Step S3.1: Obtain the peak frequency of the unaged sample at multiple different temperatures; perform linear fitting with each measured temperature value as the abscissa and the natural logarithm of the corresponding peak frequency as the ordinate to obtain the temperature correction curve; Step S3.2: Using the temperature correction curve, normalize the peak frequencies at each temperature to the same reference temperature; then perform linear fitting with the reciprocal of the measured temperature as the abscissa and the natural logarithm of the normalized peak frequency as the ordinate; according to the linear form of the Arenius formula, calculate the initial activation energy of the epoxy resin from the slope of the fitted curve.

[0007] Further, S3.2 includes: According to the linearized form of the Arenius formula: ln(f) = ln(A) - Ea / (R) T), where f is the peak frequency, T is the absolute temperature, A is the pre-exponential factor, Ea is the activation energy, and R is the gas constant; The slope k of the temperature correction curve satisfies the following relationship: k = - / R; Calculate the initial activation energy of the epoxy resin. for: =-k R.

[0008] Further, step S4 includes: Step S4.1: For each aging time, obtain the peak frequency of the sample at that aging time. and its measured temperature ; Step S4.2: Place the... , With the initial activation energy and the initial peak frequency of the unaged sample. and its measured temperature Substituting the values ​​into the activation energy calculation model, the activation energy at that aging time is calculated. ; By summarizing the activation energy data corresponding to all aging times, the evolution curve of activation energy as a function of aging time is obtained.

[0009] Further, step S4.1 includes: for each aging time, acquiring the frequency domain dielectric spectrum data of the epoxy resin sample at that aging time; Based on the frequency domain dielectric spectrum data, the Havriliak-Negami model was used for fitting, and the peak frequency of the sample at that aging time was extracted from the fitted curve. and its corresponding measurement temperature .

[0010] Furthermore, the activation energy calculation model includes the traditional Arenius formula and the improved Arenius formula, and step S4.2 includes: Determine the peak frequency after aging Relative to the initial peak frequency The degree of change; If the peak frequency after aging Relative to the initial peak frequency If the degree of change exceeds a preset threshold, the activation energy Ea corresponding to the current state is calculated using the traditional Arenius formula. If the peak frequency after aging Relative to the initial peak frequency If the degree of change does not exceed a preset threshold, then the improved Arenius formula is used to calculate the activation energy corresponding to the current state. ; In the formula, This is the initial activation energy.

[0011] Further, step S5 includes: Based on the activation energy evolution curve, key feature points marking the transition of the aging stage are obtained; Calculate the activation energy Ea of the material at each key feature point relative to the initial activation energy. The rate of change ΔEa, where ΔEa = (Ea - Ea1) / Ea1; Multiple threshold ranges are set corresponding to the rate of change ΔEa, and each threshold range defines an independent aging level, thereby constructing the aging assessment result.

[0012] Furthermore, the procedure before step S5 includes: The data points on the activation energy evolution curve are fitted using the least squares method to obtain the fitted evolution curve.

[0013] The beneficial effects of this invention are that it provides a method for evaluating the aging state of epoxy resin insulation in dry-type transformers. The method includes the following steps: Step S1: Obtaining frequency domain dielectric spectrum data of epoxy resin samples at different temperatures and aging times; Step S2: Fitting the frequency domain dielectric spectrum data using the Havriliak-Negami model to extract characteristic frequency parameters characterizing the insulation state; Step S3: Establishing a temperature correction model based on the characteristic frequency parameters of the unaged sample and its corresponding measurement temperature, and obtaining the initial activation energy of the epoxy resin; Step S4: Calculating the activation energy at each aging time based on the temperature correction model, the initial activation energy, the characteristic frequency parameters at different aging times and their corresponding measurement temperatures, and obtaining the evolution curve of the activation energy changing with aging time; Step S5: Obtaining key feature points on the evolution curve, determining the threshold of the rate of change of activation energy relative to the initial activation energy of the material at each feature point, and constructing an aging evaluation result including multiple aging levels based on the rate of change threshold. This invention achieves a quantitative and staged evaluation of the aging state of epoxy resin insulation. Attached Figure Description

[0014] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a logic flowchart of the method for assessing the aging status of epoxy resin insulation in dry-type transformers according to the present invention. Figure 2 These are curves showing the relationship between frequency and dielectric loss at different measurement temperatures; Figure 3 It is a temperature correction curve; Figure 4 This is the principle for calculating the activation energy of epoxy resin insulation materials; Figure 5 It is a graph showing the relationship between activation energy and aging time; Figure 6 These are fitted evolution curves of epoxy resin after aging at 150℃ for different times. Detailed Implementation

[0015] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the invention are now described in detail with reference to the accompanying drawings. In the following description, specific details such as particular structures and techniques are set forth for illustrative purposes and not for limitation, so as to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known devices, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0016] Figure 1 This is a logic flowchart of the method for assessing the aging status of epoxy resin insulation in dry-type transformers according to the present invention. like Figure 1 As shown, this invention provides a method for assessing the aging status of epoxy resin insulation in dry-type transformers, the method comprising the following steps: Step S1: Obtain frequency domain dielectric spectrum data of epoxy resin samples at different temperatures and aging times; Furthermore, the frequency domain dielectric spectrum data includes the measurement frequency and its corresponding dielectric loss value; In this step, it is important to note that obtaining comprehensive and accurate frequency-domain dielectric spectroscopy (FDS) data is the cornerstone of all subsequent analysis and evaluation. This process systematically places the material under controlled multi-stress conditions to obtain information on the evolution of its dielectric properties with aging, specifically including: In the sample preparation and standardization stage, an epoxy resin formulation consistent with the insulation material of the actual dry-type transformer was selected, and accelerated thermal aging experiments were conducted using multiple temperature gradients. Samples were grouped and placed in a high-precision programmable aging chamber, with a series of aging temperature gradients (e.g., 120℃, 135℃, 150℃, 165℃) set above the normal operating temperature. Isothermal aging for different durations (e.g., 0h, 48h, 96h, 240h, 312h, 360h, 408h) was performed at each temperature point to obtain a complete sample sequence covering the initial state to the deep aging state. At least three parallel samples were set for each aging condition to ensure the statistical reliability of the data. Finally, a broadband dielectric spectrometer was used to perform precise frequency domain dielectric spectrum measurements in an electromagnetically shielded environment, with the scanning frequency range set to [missing information]. Hz to The system measures the frequency at multiple constant temperatures (e.g., 30℃, 40℃, 50℃, 60℃) and automatically records the real and imaginary parts of the complex dielectric constant at each frequency point, and calculates the dielectric loss factor tanδ, thus establishing a frequency-dielectric parameter dataset for each sample. Figure 2 As shown, it is the dielectric spectrum curve of epoxy resin during the aging process at different temperatures, which intuitively shows the relationship between dielectric loss and frequency and temperature.

[0017] Step S2: The frequency domain dielectric spectrum data is fitted using the Havriliak-Negami model to extract characteristic frequency parameters representing the insulation state; In this step, it should be noted that by fitting the measured spectral data with a physical model, the intrinsic parameters characterizing the micro-relaxation dynamics of the material can be quantitatively extracted from the complex curves. Here, the Havriliak-Negami model refers to the Havriliak-Negami dielectric relaxation model.

[0018] Step S2 includes: fitting the Havriliak-Negami model based on the measured frequency and its corresponding dielectric loss value to obtain the fitting curve of the dielectric loss spectrum, and extracting the characteristic frequency parameter from the fitting curve; wherein, the characteristic frequency parameter is the peak frequency.

[0019] Furthermore, the Havriliak-Negami model includes: In the formula, For high frequency dielectric constant, For relaxation strength, For relaxation time, For the distribution width parameter, For asymmetric parameters, DC conductivity, DC conductance index parameter, ω is the angular frequency, and j is the imaginary unit.

[0020] The core of the model fitting objective and parameter extraction steps lies in using the HN model to perform nonlinear least-squares fitting on the frequency-dielectric loss curve obtained in step S1. Seven physical parameters in the model are adjusted through optimization algorithms (such as differential evolution algorithm): high-frequency limiting dielectric constant, relaxation intensity, relaxation time, symmetric broadening parameter describing the width of the relaxation time distribution, asymmetric broadening parameter describing the asymmetry of the relaxation time distribution, DC conductivity, and the frequency exponent describing the low-frequency electrode or interface polarization effect. This minimizes the residual between the model's calculated curve and the experimental data. During the fitting process, physically reasonable boundary constraints (such as 0<α≤1, 0<β≤1, τ>0) are set to ensure accuracy and stability. The fitting starts with a reasonable set of initial values ​​and is iteratively optimized until convergence. After successful fitting, the frequency of its maximum value—frequency f_p—is extracted. This frequency corresponds to the apex position of the main relaxation peak on the dielectric loss spectrum. For each dielectric spectrum of each sample (corresponding to a specific aging state and measurement temperature), the above fitting operation is performed independently, thereby systematically extracting the peak frequency data under all states.

[0021] Step S3: Based on the characteristic frequency parameters of the unaged sample and its corresponding measurement temperature, establish a temperature correction model and obtain the initial activation energy of the epoxy resin. In this step, it is important to note that, in order to eliminate the influence of different measurement temperatures on frequency parameters and establish a unified benchmark for state comparison, a modified model describing the relationship between temperature and relaxation frequency must be constructed. Simultaneously, obtaining the activation energy of the material in its unaged (initial) state serves as the absolute reference zero point for the entire aging assessment system.

[0022] Further, step S3 includes: Step S3.1: Obtain the peak frequencies of the unaged sample at multiple different temperatures; perform linear fitting with each measured temperature value as the abscissa and the natural logarithm of the corresponding peak frequency as the ordinate to obtain a temperature correction curve; such as Figure 3 As shown.

[0023] Step S3.2: Using the temperature correction curve, normalize the peak frequencies at each temperature to the same reference temperature; then perform linear fitting with the reciprocal of the measured temperature as the abscissa and the natural logarithm of the normalized peak frequency as the ordinate; according to the linear form of the Arenius formula, calculate the initial activation energy of the epoxy resin from the slope of the fitted curve.

[0024] Furthermore, S3.2 includes: the linearized form of the Arenius formula: ln(f) = ln(A) - Ea / (R) T) Where f is the peak frequency, T is the absolute temperature, A is the pre-exponential factor, Ea is the activation energy, and R is the gas constant; The slope k of the temperature correction curve satisfies the following relationship: k = - / R; Calculate the initial activation energy of the epoxy resin. for: =-k R.

[0025] Specifically, step S3.1 aims to establish a quantitative relationship between peak frequency and temperature. For unaged samples, dielectric spectra are measured at multiple different temperatures (T_i, i=1,2,3…), and the corresponding peak frequencies (f_i) are extracted. A scatter plot is drawn with the reciprocal of absolute temperature (1 / T_i) on the x-axis and the natural logarithm of the corresponding peak frequency (lnf_i) on the y-axis. According to the Arenius dynamics theory, for thermally activated relaxation processes, lnf should have a linear relationship with 1 / T. A linear regression is performed on the data points using the least squares method, and the resulting straight line is the temperature correction curve, with the equation lnf=lnA-Ea / (R·T). This curve enables the conversion between measured frequencies at different temperatures.

[0026] Step S3.2 then uses the above relationship to solve for the initial activation energy. The linearized form based on the Arenius formula is ln(f) = ln(A) - Ea / (R). The slope k of the temperature correction curve is physically equal to -Ea / R. Therefore, the initial activation energy Ea1 of the epoxy resin can be directly calculated from the slope. =-k R, where R is the gas constant (8.314 J / (mol·K)). By measuring multiple sets of parallel unaged samples and calculating the average value, a stable and reliable initial activation energy reference value for this material system can be obtained; for example, it was experimentally determined to be 90.03 kJ / mol.

[0027] Step S4: Based on the temperature correction model, initial activation energy, characteristic frequency parameters at different aging times and their corresponding measurement temperatures, calculate the activation energy at each aging time and obtain the evolution curve of activation energy with aging time. In this step, it should be noted that activation energy is a macroscopic physical quantity characterizing the energy barrier that molecular chain segments need to overcome for motion. Its value changes with microstructural changes caused by material aging (such as increased crosslinking density, chain breakage, oxidation, etc.). This step assigns a quantitative activation energy value to each aging state point through a calculation formula, thereby constructing an evolutionary trajectory describing the aging process.

[0028] Further, step S4 includes: Step S4.1: For each aging time, obtain the peak frequency of the sample at that aging time. and its measured temperature Step S4.2: ... , With initial activation energy and the initial peak frequency of the unaged sample. and its measured temperature Substituting the values ​​into the activation energy calculation model, the activation energy at that aging time is calculated. The activation energy data corresponding to all aging times were summarized to obtain the evolution curve of activation energy as a function of aging time.

[0029] By establishing a relationship equation between the peak frequency of the fitted HN model and the measured temperature, the frequency measured at room temperature can be equivalently converted to the corresponding frequency at the actual operating temperature of the transformer, thereby achieving temperature correction.

[0030] Then, the initial activation energy and the activation energy threshold at the end of the lifespan are determined based on the activation energy change law. Specifically, a fitting curve of activation energy change with aging time is established based on experimental data. The starting value of the curve (when the aging time is zero) is determined as the initial activation energy of the material, and its reliability is verified by multiple sets of parallel samples. The determination of the lifespan threshold is combined with the performance degradation criterion, referring to the national standard GB / T 11026.1-2016, and using the time-temperature equivalence principle to establish the activation energy equivalent aging time master curve. By extrapolating the evolution curve, the activation energy value corresponding to the insulation failure critical point is accurately predicted, and this value is used as the threshold.

[0031] Further, step S4.1 includes: for each aging time, acquiring the frequency domain dielectric spectrum data of the epoxy resin sample at that aging time; based on the frequency domain dielectric spectrum data, using the Havriliak-Negami model for fitting, and extracting the peak frequency of the sample at that aging time from the fitted curve. and its corresponding measurement temperature .

[0032] Furthermore, the activation energy calculation model includes the traditional Arenius formula and the improved Arenius formula, and step S4.2 includes: determining the peak frequency after aging. Relative to the initial peak frequency The degree of change; if the peak frequency after aging Relative to the initial peak frequency If the degree of change exceeds the preset threshold, the activation energy Ea corresponding to the current state is calculated using the traditional Arenius formula. If the peak frequency after aging Relative to the initial peak frequency If the degree of change does not exceed the preset threshold, then the modified Arenius formula is used to calculate the activation energy corresponding to the current state. ; In the formula, Ea1 is the initial activation energy.

[0033] The calculated activation energy of epoxy resin insulation materials reflects the combined effect of the modified Arenius formula and the traditional Arenius formula, and is applied in various scenarios. Specifically, in the aging state assessment process, for each aging time point, the peak frequency of the sample under that state must first be obtained. and its measured temperature Subsequently, , With the known initial activation energy and the initial peak frequency of the unaged sample and measuring temperature This is then substituted into the activation energy calculation model. The key step here is to dynamically select the calculation formula based on the degree of frequency change: determining the peak frequency after aging. Relative to the initial frequency The degree of change. If the degree of change exceeds a preset threshold (i.e., and The differences are significant. If the activation energy is to be calculated using the traditional Arenius formula, then the current activation energy can be calculated. Conversely, if the degree of change does not exceed the preset threshold (i.e. In this case, the modified Arenius formula is used for calculation. Therefore, when the peak frequency of dielectric loss varies only slightly, the modified formula effectively compensates for the limitations of the traditional formula. The limitation of time-lapse failure; when the peak frequency varies significantly, the traditional formula is applicable. For example... Figure 4 and Figure 5 As shown, Figure 4 The principle for calculating the activation energy of epoxy resin insulating materials. Figure 5 The graph shows the relationship between activation energy and aging time. Table 1 shows the key differences between the traditional and improved Arenius formulas, including the calculation differences between the two formulas and their corresponding applicable scenarios.

[0034] Table 1 Comparison of Traditional and Improved Arenius Formulas Furthermore, before step S5, the process includes: performing curve fitting on the data points of the activation energy evolution curve using the least squares method to obtain the fitted evolution curve. The fitted and smoothed curve more clearly reflects the changing trend, such as... Figure 6 As shown, it is the fitted evolution curve of epoxy resin after aging at 150℃ for different times.

[0035] Step S5: Obtain key feature points on the evolution curve, determine the threshold of the rate of change of activation energy relative to the initial activation energy of the material at each feature point, and construct an aging assessment result including multiple aging levels based on the rate of change threshold.

[0036] In this step, it's important to note that transforming the continuous activation energy evolution curve into discrete aging levels with clear engineering guidance is the final step in achieving intuitive condition assessment and operational decision support. This is accomplished by defining the relative rate of change of activation energy and setting scientifically defined grading thresholds.

[0037] Further, step S5 includes: obtaining key feature points that mark the turning point of the aging stage based on the activation energy evolution curve; calculating the rate of change ΔEa of the activation energy Ea of the material at each key feature point relative to the initial activation energy Ea1, where ΔEa=(Ea-Ea1) / Ea1; setting multiple threshold ranges corresponding to the rate of change ΔEa, with each threshold range defining an independent aging level, thereby constructing the aging assessment result.

[0038] Based on the evolution curve of activation energy over aging time, the ΔEa threshold can be mapped to specific feature locations: in a specific embodiment, such as Figure 6 As shown, when ΔEa reaches 20% (corresponding to an activation energy of approximately 108 kJ / mol), the curve begins to deviate from the "initial stable region," exhibiting its first inflection point. ΔEa at 30% (approximately 117 kJ / mol) marks the beginning of the acceleration phase of the curve's ascent. When ΔEa reaches 60% (approximately 144 kJ / mol), the curve enters a steep ascent phase, nearing the end of its lifespan (activation energy approximately 166.83 kJ / mol). All rates of change are referenced to the baseline activation energy of the unaged material, Ea = 90.03 kJ / mol. Extrapolating from the curve, the activation energy corresponding to approximately 800 hours of aging at 150°C is approximately 166.83 kJ / mol, representing a rate of change of approximately 85%. Therefore, setting ΔEa > 60% as the "severe aging" threshold provides a safety margin for the end of the lifespan (>60% to 85% represents a severe warning stage) and avoids premature misjudgment due to measurement errors or material fluctuations.

[0039] From a theoretical mechanism perspective: In the early stages of aging (ΔEa < 20%), the activation energy changes little, mainly reflecting material fluctuations or very early cross-linking, without causing significant performance degradation. The energy barrier for molecular chain segment motion does not change much, and the characteristic frequency of the dielectric spectrum does not shift significantly, falling within a safe operating range. In the 20% to 30% range, the corresponding curve begins to rise significantly, marking the entry of aging into a detectable accelerated phase. Microstructures such as cross-linking density and chain segment motion begin to undergo observable changes, and the dielectric loss tanδ begins to rise. In the 30% to 60% range, the slope of the curve further increases, reflecting the transition of the aging mechanism from a "competitive stage" of coexistence of cross-linking and degradation to a "degradation-dominated" stage. Material performance deteriorates rapidly, and insulation strength begins to decline. When ΔEa > 60%, the material enters a severe degradation stage, with main chain breakage, increased oxidation products, and a sharp decline in dielectric strength and mechanical properties, significantly increasing the risk of failure. By identifying the above key characteristic points, calculating the rate of change of activation energy relative to the initial activation energy ΔEa at each point, and setting multiple corresponding threshold ranges, an aging state assessment system covering multiple independent aging levels is constructed.

[0040] It is understood that the above embodiments only illustrate preferred embodiments of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can freely combine the above technical features without departing from the concept of the present invention, and can also make several modifications and improvements, all of which fall within the protection scope of the present invention. Therefore, all equivalent transformations and modifications made with respect to the scope of the claims of the present invention should fall within the scope of the claims of the present invention.

Claims

1. A method for evaluating an aging state of an epoxy resin insulation of a dry-type transformer, characterized by, The method includes the following steps: Step S1: Obtain frequency domain dielectric spectrum data of epoxy resin samples at different temperatures and aging times; Step S2: Fit the frequency domain dielectric spectrum data using the Havriliak-Negami model to extract characteristic frequency parameters representing the insulation state; Step S3: Based on the characteristic frequency parameters of the unaged sample and its corresponding measurement temperature, establish a temperature correction model and obtain the initial activation energy of the epoxy resin. Step S4: Based on the temperature correction model, the initial activation energy, the characteristic frequency parameters at different aging times and their corresponding measurement temperatures, calculate the activation energy at each aging time and obtain the evolution curve of the activation energy with aging time. Step S5: Obtain key feature points on the evolution curve, determine the threshold of the rate of change of activation energy relative to the initial activation energy of the material at each feature point, and construct an aging assessment result including multiple aging levels based on the rate of change threshold.

2. The method of claim 1, wherein the method is characterized by: The frequency domain dielectric spectrum data includes the measurement frequency and its corresponding dielectric loss value; Step S2 includes: The Havriliak-Negami model is fitted based on the measured frequency and its corresponding dielectric loss value to obtain the fitting curve of the dielectric loss spectrum, and the characteristic frequency parameter is extracted from the fitting curve; wherein, the characteristic frequency parameter is the peak frequency.

3. The method of claim 1, wherein the method is characterized by: The Havriliak-Negami model includes: wherein is the high-frequency dielectric constant, is the relaxation strength, is the relaxation time, is the distribution width parameter, is the asymmetry parameter, is the direct current conductivity, is the direct current conductivity exponent parameter, is the angular frequency, and j is the imaginary unit.

4. The dry-type transformer epoxy resin insulation aging state evaluation method according to claim 2, characterized by, Step S3 includes: Step S3.1: Obtain the peak frequency of the unaged sample at multiple different temperatures; perform linear fitting with each measured temperature value as the abscissa and the natural logarithm of the corresponding peak frequency as the ordinate to obtain the temperature correction curve; Step S3.2: Using the temperature correction curve, normalize the peak frequencies at each temperature to the same reference temperature; then perform linear fitting with the reciprocal of the measured temperature as the abscissa and the natural logarithm of the normalized peak frequency as the ordinate; according to the linear form of the Arenius formula, calculate the initial activation energy of the epoxy resin from the slope of the fitted curve.

5. The dry-type transformer epoxy resin insulation aging state evaluation method according to claim 4, characterized by, Step S3.2 includes: According to the linearized form of the Arrhenius equation: ln(f) = ln(A) - Ea / (R T), where f is the peak frequency, T is the absolute temperature, A is the pre-exponential factor, Ea is the activation energy, and R is the gas constant; The slope k of the temperature correction curve satisfies the relationship: k=- / R; calculating an initial activation energy of the epoxy resin is: = -k R.

6. The dry-type transformer epoxy resin insulation aging state evaluation method according to claim 5, characterized by, Step S4 includes: Step S4.1: For each aging time, obtain the peak frequency of the sample at that aging time. and its measured temperature ; Step S4.2: The peak frequency and its measured temperature With the initial activation energy and the initial peak frequency of the unaged sample. and its measured temperature Substituting the values ​​into the activation energy calculation model, the activation energy at that aging time is calculated. ; By summarizing the activation energy data corresponding to all aging times, the evolution curve of activation energy as a function of aging time is obtained.

7. The method for assessing the aging status of epoxy resin insulation in dry-type transformers according to claim 6, characterized in that, Step S4.1 includes: for each aging time, acquiring the frequency domain dielectric spectrum data of the epoxy resin sample at that aging time; Based on the frequency domain dielectric spectrum data, the Havriliak-Negami model was used for fitting, and the peak frequency of the sample at that aging time was extracted from the fitted curve. and its corresponding measured temperature .

8. The method for assessing the aging status of epoxy resin insulation in dry-type transformers according to claim 6, characterized in that, The activation energy calculation model includes the traditional Arenius formula and the improved Arenius formula, and step S4.2 includes: Determine the peak frequency after aging Relative to the initial peak frequency The degree of change; If the peak frequency after aging Relative to the initial peak frequency If the degree of change exceeds a preset threshold, the activation energy corresponding to the current state is calculated using the traditional Arenius formula. ; If the peak frequency after aging Relative to the initial peak frequency If the degree of change does not exceed a preset threshold, then the improved Arenius formula is used to calculate the activation energy corresponding to the current state. ; In the formula, This is the initial activation energy.

9. The method for assessing the aging status of epoxy resin insulation in dry-type transformers according to claim 1, characterized in that, Step S5 includes: Based on the activation energy evolution curve, key feature points marking the transition of the aging stage are obtained; Calculate the rate of change ΔEa of the activation energy Ea of the material at each key feature point relative to the initial activation energy Ea1, where ΔEa = (Ea - Ea1) / Ea1; Multiple threshold ranges are set corresponding to the rate of change ΔEa, and each threshold range defines an independent aging level, thereby constructing the aging assessment result.

10. The method for assessing the aging status of epoxy resin insulation in dry-type transformers according to claim 1, characterized in that, The procedure preceding step S5 also includes: The data points on the activation energy evolution curve are fitted using the least squares method to obtain the fitted evolution curve.

Images