A method for testing the explosion resistance of a power capacitor

Abstract

The present application relates to the technical field of capacitor, especially to a power capacitor explosion-proof test method. A multi-frequency harmonic current sequence is injected in a complex stress field, and a three-dimensional strain propagation characteristic of a capacitor shell is captured in real time. A strain-discharge space-time correlation model is established according to the three-dimensional strain propagation characteristic. A capacitor shell structure integrity attenuation curve is obtained according to the strain-discharge space-time correlation model. When the slope of the attenuation curve breaks through a preset critical topology form, a multi-stage linkage protection device is triggered to implement energy discharge. A plasma spectrum fingerprint formed by the self-healing characteristics of the metallized film in the energy discharge process is extracted, and a capacitor explosion-proof performance grade evaluation system is inversely reconstructed through the characteristic spectral line offset. The present application realizes a whole-process closed-loop test of the capacitor from stress loading, failure warning to performance evaluation, and improves the environmental simulation reality, failure warning accuracy and performance evaluation scientificity of the explosion-proof test.

Description

Technical Field

[0001] This invention relates to the field of capacitor technology, and in particular to a method for testing the explosion resistance of power capacitors. Background Technology

[0002] As power systems develop towards higher voltage and larger capacity, power capacitors, as key equipment for reactive power compensation and filtering, face severe challenges to their operational reliability. Traditional capacitor explosion-resistance testing methods have significant limitations: on the one hand, existing tests mostly employ a single stress loading method, which cannot realistically simulate the coupling effects of multiple physical fields such as temperature, electric, and mechanical vibration fields during actual operation; on the other hand, conventional tests only focus on the final failure threshold, lacking dynamic monitoring and early warning mechanisms for the capacitor failure process. Particularly in the assessment of the self-healing characteristics of metallized thin films, existing technologies struggle to capture the microscopic physicochemical changes during the self-healing process in real time, leading to significant errors in explosion-resistance performance evaluation. Furthermore, traditional strain detection methods are limited by spatial resolution and response speed, failing to accurately establish the spatiotemporal correlation between shell strain and internal discharge, thus restricting the accuracy of capacitor failure early warning. There is an urgent need in this field for a power capacitor explosion-resistance testing method that can simulate real operating conditions, achieve full-process dynamic monitoring, and establish a multi-parameter correlation evaluation system. This would address key technical issues in existing technologies, such as inaccurate simulation of composite stress fields, incomplete monitoring of the failure process, and incomplete evaluation of explosion-resistance performance, providing a more scientific testing method for the safe operation and reliability assessment of power capacitors. Summary of the Invention

[0003] This invention overcomes the shortcomings of the prior art and provides a method for testing the explosion resistance of power capacitors.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] This invention discloses a method for testing the explosion resistance of power capacitors, comprising the following steps:

[0006] Obtain the actual operating environment of the target capacitor, construct the composite stress field of the target capacitor when it is working in the actual operating environment, and inject a multi-band harmonic current sequence into the composite stress field;

[0007] The three-dimensional strain propagation characteristics of the capacitor shell are captured in real time by a distributed optical fiber sensor array, and a strain-discharge spatiotemporal correlation model is established based on the three-dimensional strain propagation characteristics.

[0008] The capacitor shell structure integrity decay curve is obtained based on the strain-discharge spatiotemporal correlation model. When the slope of the decay curve exceeds the preset critical topology, a multi-level linkage protection device is triggered to release energy.

[0009] The plasma spectral fingerprint formed by the self-healing properties of the metallized thin film during energy release is extracted. The explosion resistance performance evaluation system of the capacitor is reconstructed in reverse by the characteristic spectral line shift, and the evaluation results are generated.

[0010] The composite stress field includes synchronously superimposed alternating temperature gradient fields, asymmetric voltage shock waves, and mechanical vibration modes.

[0011] Preferably, the actual operating environment of the target capacitor is obtained, and a composite stress field is constructed when the target capacitor operates in the actual operating environment, specifically as follows:

[0012] By deploying a sensor network at preset locations on the target capacitor, the temperature fluctuation spectrum, the time-varying envelope characteristics of the voltage impulse waveform, and the frequency domain energy distribution of the mechanical vibration source in its actual operating environment are collected in real time.

[0013] The temperature fluctuation spectrum is decomposed into a fundamental temperature component and multiple subharmonic temperature components. At the same time, the leading transient parameters and subsequent oscillation modes are separated from the voltage impulse waveform, and the position of the main resonance peak in the frequency domain energy distribution of mechanical vibration is extracted.

[0014] The fundamental temperature component and the subharmonic temperature component are reconstructed in phase synchronization to form a multilayer alternating temperature gradient field with dynamic phase difference. At the same time, an asymmetric voltage shock wave time domain sequence is generated with the leading transient parameters as initial conditions and the oscillation mode as boundary constraints. The position of the main resonance peak is mapped to the excitation frequency of the mechanical vibration mode.

[0015] Based on the spatial distribution parameters of the multilayer alternating temperature gradient field, thermal deformation compensation correction is performed on the propagation path of the time-domain sequence of the asymmetric voltage shock wave, so that the leading edge of the voltage shock wave and the antinode node of the mechanical vibration mode form a spatiotemporal coupling effect.

[0016] By comparing the covariance matrix of the dynamic stress response spectrum under the spatiotemporal coupling effect with that of the actual environmental parameter set, the layering weight of the temperature gradient field, the damping coefficient of the voltage shock wave oscillation, and the amplitude-frequency ratio of the mechanical vibration excitation are iteratively adjusted until the multi-physics field synergy factor of the composite stress field reaches the preset spatiotemporal distribution consistency threshold.

[0017] Preferably, a multi-band harmonic current sequence is injected into the composite stress field, specifically as follows:

[0018] Real-time acquisition of dielectric polarization-relaxation response spectrum inside capacitor, and extraction of its dielectric relaxation time constant as dynamic change rate of dielectric loss under composite stress field.

[0019] The dielectric relaxation time constant is input into the multi-band harmonic generator, which divides the preset wideband harmonics into subsynchronous oscillation band, fundamental resonant band and supersynchronous resonant band, and generates an initial coding sequence of phase modulation parameters based on the cross-modulation effect between each band.

[0020] The phase modulation parameters are nonlinearly coupled with the dynamic change rate of dielectric loss to obtain the dielectric loss fluctuation rate.

[0021] The phase angle offset of the initial coding sequence is dynamically weighted by the dielectric loss volatility to form an adaptive coding sequence with dielectric-thermal coupling characteristics.

[0022] The carrier phase of each harmonic frequency band is controlled in real time according to the adaptive coding sequence to generate a multi-band harmonic current sequence with dynamic phase deflection characteristics.

[0023] Preferably, the three-dimensional strain propagation characteristics of the capacitor casing are captured in real time using a distributed fiber optic sensing array, and a strain-discharge spatiotemporal correlation model is established based on the three-dimensional strain propagation characteristics, specifically as follows:

[0024] A grating sensor array is deployed on the surface of the capacitor shell. The omnidirectional strain gradient vector on the shell surface is synchronously acquired by orthogonal polarization light interferometry demodulation technology, and the elastic waveguide phase change characteristics on its propagation path are extracted.

[0025] The omnidirectional strain gradient vector is decomposed into an elastic waveguide component and a viscoplastic creep component. The strain energy flux density propagation rate of each preset grid node of the capacitor shell is determined by utilizing the phase change characteristics of the elastic waveguide.

[0026] A wideband partial discharge signal acquisition system is used to capture the time-frequency ridge characteristics of discharge pulses inside the medium. The strain energy flux density propagation rate and the time-frequency ridge characteristics of the discharge pulses are spatiotemporally registered to establish a covariant tensor field between the strain propagation wavefront and the discharge electromagnetic pulse wavefront.

[0027] Singular value decomposition is performed on the covariant tensor field to extract the principal component coupling coefficient matrix of strain propagation rate and discharge pulse amplitude-frequency characteristics, and a strain-discharge spatiotemporal correlation feature surface is constructed based on the principal component coupling coefficient matrix.

[0028] A strain-discharge spatiotemporal correlation model is established based on the strain-discharge spatiotemporal correlation feature surface.

[0029] Preferably, the capacitor casing structural integrity attenuation curve is obtained based on the strain-discharge spatiotemporal correlation model. When the slope of the attenuation curve exceeds the preset critical topological shape, a multi-level linkage protection device is triggered to release energy. Specifically:

[0030] The principal component coupling coefficient matrix of strain propagation rate and discharge pulse amplitude-frequency characteristics is extracted from the strain-discharge spatiotemporal correlation model, and the dynamic stress accumulation factor of each preset grid node of the shell is determined based on the principal component coupling coefficient matrix.

[0031] The dynamic stress accumulation factor is compared point by point with the preset material fatigue damage threshold to screen out stress-exceeding nodes and generate a local failure probability distribution cloud map.

[0032] Based on the spatial gradient change trend of the local failure probability distribution cloud map, a nonlinear decay curve of the shell structure integrity is fitted, and the change of the second derivative of the decay curve is monitored in real time.

[0033] When the extreme point of the second derivative exceeds the curvature change threshold defined by the preset critical topological shape, the pre-trigger signal of the multi-level linkage protection device is activated.

[0034] Based on the intensity level matching of the pre-triggered signal, the conduction timing of the energy release channel is matched to prioritize the release of electromagnetic-mechanical coupling energy in the region where the stress over-limit node is located, while simultaneously adjusting the loading parameters of the composite stress field to suppress cascading failures.

[0035] Preferably, the plasma spectral fingerprint formed by the self-healing properties of the metallized thin film during energy release is extracted, and the explosion-proof performance rating system of the capacitor is reconstructed inversely by the characteristic spectral line shift, and the evaluation results are generated, specifically as follows:

[0036] A high-speed spectral acquisition device is used to capture transient plasma radiation generated by the energy discharge channel, and the time-varying intensity envelope of the characteristic spectral lines of metal vapor is extracted.

[0037] The time-varying intensity envelope was matched with a standard metallic element emission database to identify characteristic spectral line clusters formed by the recrystallization of the vapor-deposited material during the self-healing process;

[0038] By calculating the dynamic shift of the main peak wavelength in the characteristic spectral line cluster relative to the reference position, a mapping relationship between plasma expansion rate and thin film self-healing efficiency is established.

[0039] Based on the mapping relationship, the self-healing intensity gradient of the thin film in different stress regions is quantified to generate a three-dimensional self-healing efficiency distribution map.

[0040] By combining the statistical characteristics of anisotropic factors in the three-dimensional self-healing efficiency distribution map, a multi-level evaluation index matrix reflecting the overall explosion resistance performance of capacitors is constructed.

[0041] Based on the weight distribution of anisotropic factors in the multi-level evaluation index matrix, each evaluation index is compared and analyzed with the preset explosion-proof performance level threshold to generate the explosion-proof performance level evaluation result of the target capacitor.

[0042] The evaluation indicators include plasma expansion rate, thin film self-healing efficiency gradient, three-dimensional self-healing efficiency distribution uniformity, and anisotropic factor intensity distribution characteristics.

[0043] Preferably, the dynamic stress accumulation factor of each preset mesh node in the shell is determined based on the principal component coupling coefficient matrix, specifically as follows:

[0044] The eigenvectors of strain propagation rate and discharge pulse amplitude-frequency characteristics are extracted from the principal component coupling coefficient matrix, and the stress wave propagation phase difference of each preset grid node is calculated.

[0045] The stress response amplitude is obtained by convolving the stress wave propagation phase difference with the material damping characteristic parameters of the corresponding node.

[0046] Based on the integral of the stress response amplitude over time, and combined with the geometric deformation sensitivity coefficient of the preset mesh nodes, the initial stress accumulation is generated.

[0047] By introducing a stress transfer efficiency correction factor between adjacent nodes, the initial stress accumulation is spatially correlated and corrected to obtain the corrected nodal stress field distribution.

[0048] The dynamic stress accumulation factor of each grid node is output by calculating the ratio of the energy dissipation rate of each node in the stress field distribution to the material fatigue limit.

[0049] This invention addresses the technical deficiencies in the prior art and offers the following advantages: It obtains the actual operating environment of the target capacitor, constructs a composite stress field of the target capacitor under actual operating conditions, and injects a multi-band harmonic current sequence into the composite stress field; it captures the three-dimensional strain propagation characteristics of the capacitor shell in real time using a distributed fiber optic sensing array, and establishes a strain-discharge spatiotemporal correlation model based on these characteristics; it obtains the structural integrity attenuation curve of the capacitor shell based on the strain-discharge spatiotemporal correlation model, and triggers a multi-level linkage protection device to release energy when the slope of the attenuation curve exceeds a preset critical topological shape; it extracts the plasma spectral fingerprint formed by the self-healing properties of the metallized thin film during energy release, and reconstructs the explosion-proof performance rating system of the capacitor through the characteristic spectral line offset, generating evaluation results. This invention achieves a closed-loop test of the entire process of capacitor testing, from stress loading and failure warning to performance evaluation, improving the realism of the environment simulation, the accuracy of failure warning, and the scientific nature of performance evaluation in explosion-proof testing, providing a comprehensive and effective testing method for capacitor reliability design. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained from these drawings without creative effort.

[0051] Figure 1 A flowchart of the first method for an explosion resistance test of a power capacitor;

[0052] Figure 2 This is a flowchart of the second method of a test method for the explosion resistance of power capacitors;

[0053] Figure 3 This is a flowchart of the third method for an explosion resistance test of power capacitors. Detailed Implementation

[0054] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0055] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0056] like Figure 1 As shown, this invention discloses a method for testing the explosion resistance of power capacitors, comprising the following steps:

[0057] S102. Obtain the actual operating environment of the target capacitor, construct the composite stress field of the target capacitor when it is working in the actual operating environment, and inject a multi-band harmonic current sequence into the composite stress field.

[0058] S104. The three-dimensional strain propagation characteristics of the capacitor shell are captured in real time by a distributed optical fiber sensing array, and a strain-discharge spatiotemporal correlation model is established based on the three-dimensional strain propagation characteristics.

[0059] S106. Obtain the capacitor shell structure integrity attenuation curve according to the strain-discharge spatiotemporal correlation model. When the slope of the attenuation curve exceeds the preset critical topology, trigger the multi-level linkage protection device to release energy.

[0060] S108. Extract the plasma spectral fingerprint formed by the self-healing properties of the metallized thin film during energy release, and reconstruct the explosion resistance performance level evaluation system of the capacitor by the characteristic spectral line offset, and generate the evaluation results.

[0061] The composite stress field includes synchronously superimposed alternating temperature gradient fields, asymmetric voltage shock waves, and mechanical vibration modes.

[0062] It should be noted that this invention addresses the technical problems of unrealistic multi-physics coupling simulation, incomplete failure process monitoring, and inaccurate performance evaluation in capacitor explosion-proof testing. By constructing a composite stress field to simulate real working conditions, combined with strain-discharge dynamic correlation analysis and self-healing characteristic spectral detection, a closed-loop testing system for capacitors is achieved, encompassing the entire process from stress loading and failure early warning to performance evaluation. This improves the realism of environmental simulation, the accuracy of failure early warning, and the scientific rigor of performance evaluation in explosion-proof testing, providing a comprehensive and effective testing method for capacitor reliability design.

[0063] Preferably, the actual operating environment of the target capacitor is obtained, and a composite stress field of the target capacitor under actual operating conditions is constructed, such as... Figure 2 As shown, specifically:

[0064] S202. Through a sensor network deployed at a preset location on the target capacitor, the temperature fluctuation spectrum, the time-varying envelope characteristics of the voltage impulse waveform, and the frequency domain energy distribution of the mechanical vibration source in its actual operating environment are collected in real time.

[0065] S204. Decompose the temperature fluctuation spectrum into a fundamental temperature component and multiple subharmonic temperature components. At the same time, separate the leading transient parameters and subsequent oscillation modes from the voltage impulse waveform, and extract the position of the main resonance peak in the frequency domain energy distribution of mechanical vibration.

[0066] S206. The fundamental temperature component and the subharmonic temperature component are reconstructed in phase synchronization to form a multi-layer alternating temperature gradient field with dynamic phase difference. At the same time, an asymmetric voltage shock wave time domain sequence is generated with the leading transient parameters as initial conditions and the oscillation mode as boundary constraints. The position of the main resonance peak is mapped to the excitation frequency of the mechanical vibration mode.

[0067] S208. Based on the spatial distribution parameters of the multilayer alternating temperature gradient field, thermal deformation compensation correction is performed on the propagation path of the time-domain sequence of the asymmetric voltage shock wave, so that the leading edge of the voltage shock wave and the antinode node of the mechanical vibration mode form a spatiotemporal coupling effect.

[0068] S210. By comparing the dynamic stress response spectrum under the spatiotemporal coupling effect with the covariance matrix of the actual environmental parameter set, the layering weight of the temperature gradient field, the damping coefficient of the voltage shock wave oscillation and the amplitude-frequency ratio of the mechanical vibration excitation are iteratively adjusted until the multi-physics field synergy factor of the composite stress field reaches the preset spatiotemporal distribution consistency threshold.

[0069] In one specific embodiment of the invention, during the explosion-proof test of the filter capacitor, a sensor network consisting of 12 PT100 temperature sensors, 6 high-frequency voltage probes, and 8 MEMS accelerometers is deployed to collect in real time the temperature fluctuation spectrum (fundamental temperature component 55±3℃, subharmonic components including 5Hz / 10Hz harmonics), voltage impulse waveform (leading transient parameters including 0.2ms rise time and 8kV peak voltage, oscillation mode including 1200Hz frequency and 0.15 damping ratio), and mechanical vibration frequency domain energy distribution (main resonance peaks located at 85Hz and 250Hz) generated during capacitor operation. A three-layer alternating temperature gradient field (central region 60℃ / outer region 45℃, phase difference 60°) is generated through phase synchronous reconstruction, and an asymmetric voltage impulse wave time domain sequence (pulse width 2ms / repetition frequency 5 times / second) is constructed. Simultaneously, the 85Hz / 250Hz main resonance peak is mapped to the mechanical vibration excitation frequency. Based on the spatial distribution parameters of the temperature gradient field (axial gradient 5℃ / cm, radial gradient 3℃ / cm), thermal deformation compensation (compensation amount 0.3mm) is applied to the propagation path of the voltage shock wave to achieve spatiotemporal coupling between the leading edge of the 8kV shock wave and the antinode of the vibration wave within 3ms. By comparing the covariance matrix (dimension 12×12) of the dynamic stress response spectrum (main frequency band 800-1500Hz) and the actual environmental parameter set, the layering weight coefficient of the temperature gradient field is iteratively adjusted from 0.6 to 0.8, the voltage shock wave oscillation damping coefficient is optimized from 0.2 to 0.1, and the amplitude-frequency ratio of mechanical vibration excitation is corrected from 1:3 to 1:2. Ultimately, the multi-physics synergy factor of the composite stress field reaches 0.92, exceeding the preset spatiotemporal consistency threshold of 0.9, thus forming a composite stress field.

[0070] In summary, this method achieves spatiotemporal coordinated loading of temperature field, electric field and mechanical vibration field by dynamically decomposing environmental parameters and establishing a multi-physics coupling mechanism. This enables the composite stress field to realistically reproduce the actual operating conditions in terms of spatial distribution and dynamic response characteristics, thereby improving the environmental simulation accuracy and failure mechanism reproduction capability of explosion-proof tests and providing accurate multi-field coupling test conditions for capacitor reliability assessment.

[0071] Preferably, a multi-band harmonic current sequence is injected into the composite stress field, such as... Figure 3 As shown, specifically:

[0072] S302. Real-time acquisition of the dielectric polarization-relaxation response spectrum inside the capacitor, and extraction of the dynamic change rate of its dielectric relaxation time constant with dielectric loss under the action of composite stress field.

[0073] Among them, the polarization-relaxation response spectrum refers to the characteristic curve of the polarization intensity of a dielectric material changing with frequency under the action of an applied alternating electric field, reflecting the relaxation process of the polarization mechanism inside the dielectric and its corresponding dielectric response characteristics.

[0074] S304. Input the dielectric relaxation time constant into the multi-band harmonic generator, divide the preset wideband harmonics into subsynchronous oscillation band, fundamental resonant band and supersynchronous resonant band, and generate an initial coding sequence of phase modulation parameters based on the cross-modulation effect between each band.

[0075] S306. Nonlinearly couple the phase modulation parameters with the dynamic change rate of dielectric loss to obtain the dielectric loss fluctuation rate.

[0076] S308. The phase angle offset of the initial coding sequence is dynamically weighted by the dielectric loss fluctuation rate to form an adaptive coding sequence with dielectric-thermal coupling characteristics.

[0077] S310. Perform real-time sliding mode control on the carrier phase of each harmonic frequency band according to the adaptive coding sequence to generate a multi-band harmonic current sequence with dynamic phase deflection characteristics.

[0078] In one specific embodiment of the present invention, during the explosion-proof test of the filter capacitor, a broadband dielectric spectrometer (test frequency range 0.1Hz-10kHz) was used to collect the polarization-relaxation response spectrum of the polypropylene film dielectric at an operating temperature of 85℃ in real time. The dielectric relaxation time constant was measured to be 25ms±3ms, and the dielectric loss tangent was monitored to dynamically fluctuate within the range of 0.0012-0.0025. The acquired dielectric parameters were input into a multi-band harmonic generator, and the output harmonics were divided into: a subsynchronous oscillation band (15-45Hz), a fundamental resonant band (50Hz±0.5%), and a supersynchronous resonant band (55-2500Hz). Through cross-modulation of the second and third harmonics, an initial phase coding sequence (a step change of 0°-120° with a step size of 30°) was generated. The phase encoding parameters are coupled exponentially with the real-time monitored dynamic change rate of dielectric loss (fluctuation amplitude ±15%) to output a dielectric loss fluctuation parameter (range of 0.8-1.2 per unit). Then, based on this fluctuation parameter, the phase angle offset of the initial phase encoding sequence is dynamically adjusted (adjustment amplitude ±5°-±20°), and finally a multi-band harmonic current sequence containing a 50Hz fundamental frequency (phase accuracy ±0.5°) and various harmonics (such as 100Hz, 150Hz, etc.) is output, with the total harmonic distortion (THD) controlled within 8%.

[0079] Overall, this method achieves adaptive matching between harmonic current sequences and dynamic loss characteristics of capacitor dielectrics by acquiring dielectric response characteristics in real time and establishing a dielectric-thermal coupling control mechanism. This enables the injected harmonic current to accurately simulate the dynamic harmonic conditions in actual power grid operation, effectively improving the accuracy of capacitor performance testing and the effectiveness of reliability assessment in complex harmonic environments.

[0080] Preferably, the three-dimensional strain propagation characteristics of the capacitor casing are captured in real time using a distributed fiber optic sensing array, and a strain-discharge spatiotemporal correlation model is established based on the three-dimensional strain propagation characteristics, specifically as follows:

[0081] A grating sensor array is deployed on the surface of the capacitor shell. The omnidirectional strain gradient vector on the shell surface is synchronously acquired by orthogonal polarization light interferometry demodulation technology, and the elastic waveguide phase change characteristics on its propagation path are extracted.

[0082] Among them, the orthogonal polarization interferometry demodulation technology refers to the measurement method that uses the interference signal generated by two orthogonally polarized lasers in a grating sensor to demodulate strain information by detecting the change in optical path difference. It achieves high-precision measurement of strain vector by separating the phase difference between the TE and TM polarization modes.

[0083] The omnidirectional strain gradient vector is decomposed into an elastic waveguide component and a viscoplastic creep component. The strain energy flux density propagation rate of each preset grid node of the capacitor shell is determined by utilizing the phase change characteristics of the elastic waveguide.

[0084] A wideband partial discharge signal acquisition system is used to capture the time-frequency ridge characteristics of discharge pulses inside the medium. The strain energy flux density propagation rate and the time-frequency ridge characteristics of the discharge pulses are spatiotemporally registered to establish a covariant tensor field between the strain propagation wavefront and the discharge electromagnetic pulse wavefront.

[0085] Singular value decomposition is performed on the covariant tensor field to extract the principal component coupling coefficient matrix of strain propagation rate and discharge pulse amplitude-frequency characteristics, and a strain-discharge spatiotemporal correlation feature surface is constructed based on the principal component coupling coefficient matrix.

[0086] A strain-discharge spatiotemporal correlation model is established based on the strain-discharge spatiotemporal correlation feature surface.

[0087] It should be noted that, by performing Gaussian curvature analysis on the feature surface, the distribution characteristics of curvature extrema points characterizing the strain-discharge coupling strength are extracted. Based on the curvature extrema points, a spatiotemporal mapping relationship between the strain propagation wavefront and the discharge pulse wavefront is constructed, forming a set of associated feature vectors containing three-dimensional parameters of amplitude, frequency, and phase. Then, a manifold learning method is used to reduce the dimensionality of the feature vector set, extracting the principal component features that dominate the strain-discharge coupling behavior. Finally, a deep correlation model between the strain propagation path and the discharge development trajectory is constructed through a spatiotemporal convolutional neural network. The network input layer receives the feature vectors after principal component analysis, the hidden layer includes a spatiotemporal attention mechanism to capture key coupling nodes, and the output layer generates a probability distribution map reflecting the dynamic correlation strength between strain and discharge. This model optimizes the network weights through a backpropagation algorithm, ultimately outputting a strain-discharge correlation strength matrix with spatiotemporal resolution.

[0088] In one specific embodiment of the invention, during the explosion-proof test of the filter capacitor, 24 FBG grating sensors (15cm spacing) are evenly spaced along the axial direction of the shell. A 1550nm wavelength orthogonal polarized light interferometry system is used to synchronously acquire the omnidirectional strain gradient vector on the shell surface at a sampling rate of 100kHz, and the maximum principal strain gradient is measured to be 0.12με / mm. The elastic waveguide component (85%) and viscoplastic creep component (15%) are extracted by mode decomposition. Based on the elastic waveguide phase abrupt change characteristics (phase jump ≥5°), the strain energy flux density propagation rate is determined to be 120m / s±10m / s. Simultaneously, a 40MHz broadband partial discharge detection system is used to capture the time-frequency ridge characteristics of the discharge pulse (main frequency band 3-18MHz, pulse width 50-200ns). The strain energy flux density propagation rate and the discharge pulse characteristics are spatiotemporally registered (time alignment accuracy ≤1μs) to establish a 12×12-dimensional strain-discharge covariance tensor field. Principal component decomposition was performed on the tensor field, and the coupling coefficients of the first three principal components (weighting coefficients of 0.72, 0.18, and 0.07) were extracted to construct a three-dimensional associated feature surface (X-axis strain rate 0-200 m / s, Y-axis discharge amplitude 0-50 mV, Z-axis coupling strength 0-1 per unit). Finally, the curvature distribution of the feature surface (maximum curvature 0.35 mm) was determined. -1 Establish a spatiotemporal correlation mapping model between strain propagation path and discharge active region.

[0089] In summary, this method establishes a spatiotemporal coupling model of strain waves and discharge pulses, enabling dynamic correlation analysis between mechanical deformation of the capacitor casing and internal insulation degradation. It can accurately identify the spatiotemporal correspondence between strain concentration areas and discharge active areas, providing multi-dimensional condition assessment basis for early fault warning and failure location of capacitors.

[0090] Preferably, the capacitor casing structural integrity attenuation curve is obtained based on the strain-discharge spatiotemporal correlation model. When the slope of the attenuation curve exceeds the preset critical topological shape, a multi-level linkage protection device is triggered to release energy. Specifically:

[0091] The principal component coupling coefficient matrix of strain propagation rate and discharge pulse amplitude-frequency characteristics is extracted from the strain-discharge spatiotemporal correlation model, and the dynamic stress accumulation factor of each preset grid node of the shell is determined based on the principal component coupling coefficient matrix.

[0092] The dynamic stress accumulation factor is compared point by point with the preset material fatigue damage threshold to screen out stress-exceeding nodes and generate a local failure probability distribution cloud map.

[0093] Based on the spatial gradient change trend of the local failure probability distribution cloud map, a nonlinear decay curve of the shell structure integrity is fitted, and the change of the second derivative of the decay curve is monitored in real time.

[0094] When the extreme point of the second derivative exceeds the curvature change threshold defined by the preset critical topological shape, the pre-trigger signal of the multi-level linkage protection device is activated.

[0095] Based on the intensity level matching of the pre-triggered signal, the conduction timing of the energy release channel is matched to prioritize the release of electromagnetic-mechanical coupling energy in the region where the stress over-limit node is located, while simultaneously adjusting the loading parameters of the composite stress field to suppress cascading failures.

[0096] In one specific embodiment of the invention, during the testing of the filter capacitor, the coupling coefficients of the first three principal components (weighting coefficients 0.72, 0.18, and 0.07) are extracted from the strain-discharge spatiotemporal correlation model, and the dynamic stress accumulation factor (range 0.3-1.2 per unit value) of 36 grid nodes on the shell surface is calculated. The accumulation factor of each node is compared point by point with the preset fatigue damage threshold of aluminum alloy material (0.85 per unit value), and nodes 5 / 12 / 19 are selected as stress over-limit points (1.05-1.2 per unit value), generating a local failure probability distribution cloud map (probability value 0.65-0.92). Based on the spatial gradient change of the cloud map (maximum gradient 0.18 / mm), a shell structural integrity decay curve (nonlinear exponent α = 1.5) is fitted, and the change of the second derivative of the curve is monitored in real time. When an extreme point curvature reaches 0.45 / mm, the curve is considered complete. 2 (Breaking through the preset critical topological morphology threshold of 0.4 / mm) 2When the signal is triggered, the pre-trigger signal (level 3) of the three-level linkage protection device is activated. According to the signal level, the No. 3 energy discharge channel (current capacity 8kA / 10μs) is connected to release the coupled energy (peak discharge current 6.5kA) in the area where node 5 is located first. At the same time, the peak voltage shock wave of the composite stress field is dynamically adjusted from 8kV to 6kV and the mechanical vibration excitation frequency is reduced from 85Hz to 70Hz to suppress the failure propagation in adjacent areas.

[0097] In summary, by establishing a dynamic mapping relationship between the strain-discharge correlation model and the structural integrity decay curve, accurate prediction and graded early warning of capacitor shell failure process are achieved. When material damage accumulates to a critical state, a multi-level protection mechanism can be intelligently triggered to effectively avoid sudden bursting accidents. At the same time, through directional energy discharge and stress field feedback adjustment, the response accuracy and reliability of the protection system are improved, minimizing the risk of fault expansion.

[0098] Preferably, the plasma spectral fingerprint formed by the self-healing properties of the metallized thin film during energy release is extracted, and the explosion-proof performance rating system of the capacitor is reconstructed inversely by the characteristic spectral line shift, and the evaluation results are generated, specifically as follows:

[0099] A high-speed spectral acquisition device is used to capture transient plasma radiation generated by the energy discharge channel, and the time-varying intensity envelope of the characteristic spectral lines of metal vapor is extracted.

[0100] The time-varying intensity envelope was matched with a standard metallic element emission database to identify characteristic spectral line clusters formed by the recrystallization of the vapor-deposited material during the self-healing process;

[0101] The standard metal element emission database refers to a reference database that systematically records the characteristic emission spectral wavelengths and relative intensity distributions of various metal elements in plasma state through experimental determination, and is used for element identification and quantitative analysis in spectral analysis.

[0102] By calculating the dynamic shift of the main peak wavelength in the characteristic spectral line cluster relative to the reference position, a mapping relationship between plasma expansion rate and thin film self-healing efficiency is established.

[0103] Based on the mapping relationship, the self-healing intensity gradient of the thin film in different stress regions is quantified to generate a three-dimensional self-healing efficiency distribution map.

[0104] By combining the statistical characteristics of anisotropic factors in the three-dimensional self-healing efficiency distribution map, a multi-level evaluation index matrix reflecting the overall explosion resistance performance of capacitors is constructed.

[0105] Based on the weight distribution of anisotropic factors in the multi-level evaluation index matrix, each evaluation index is compared and analyzed with the preset explosion-proof performance level threshold to generate the explosion-proof performance level evaluation result of the target capacitor.

[0106] The evaluation indicators include plasma expansion rate, thin film self-healing efficiency gradient, three-dimensional self-healing efficiency distribution uniformity, and anisotropic factor intensity distribution characteristics.

[0107] In one specific embodiment of the invention, during the testing of the filter capacitor, a high-speed spectrometer (wavelength range 200-800 nm) at 50,000 frames / second was used to capture the plasma radiation generated by the discharge channel, detecting the time-varying intensity envelope (pulse width 50-200 μs, peak intensity 1500-3000 a.u.) of the characteristic spectral lines of Zn (481.05 nm, 636.23 nm). The spectral data was compared with the NIST standard database to identify the characteristic spectral line cluster formed by Zn recrystallization (main peak 481.05 nm ± 0.2 nm, accompanying peaks 472.22 nm / 492.40 nm). The dynamic redshift of the main peak wavelength relative to the reference position (481.05 nm) (maximum shift +0.15 nm) was measured, establishing a linear mapping relationship between the plasma expansion rate (120-180 m / s) and the film self-healing efficiency (75-92%). A three-dimensional self-healing efficiency distribution map (axial gradient 15% / cm, radial gradient 8% / cm) is generated based on the mapping relationship. An anisotropy factor (0.25-0.45) is extracted to construct an evaluation matrix. Finally, the plasma expansion rate (155m / s), self-healing efficiency gradient (12% / cm), distribution uniformity (85%), and anisotropy factor (0.32) are compared with preset thresholds to evaluate the explosion resistance performance level of the capacitor.

[0108] In summary, to address the limitations of traditional capacitor explosion-resistance performance evaluation methods, such as the inability to monitor the self-healing process of metallized thin films in real time and the lack of quantitative evaluation standards, this method analyzes the plasma spectral characteristics during energy release. This enables dynamic capture and quantitative evaluation of the film's self-healing behavior, establishing an explosion-resistance performance evaluation system based on multi-dimensional characteristic parameters. This system accurately reflects the self-healing capability and overall explosion-resistance performance of capacitors under different stress conditions, providing a scientific basis for capacitor reliability classification and service life prediction.

[0109] Preferably, the dynamic stress accumulation factor of each preset mesh node in the shell is determined based on the principal component coupling coefficient matrix, specifically as follows:

[0110] The eigenvectors of strain propagation rate and discharge pulse amplitude-frequency characteristics are extracted from the principal component coupling coefficient matrix, and the stress wave propagation phase difference of each preset grid node is calculated.

[0111] The stress response amplitude is obtained by convolving the stress wave propagation phase difference with the material damping characteristic parameters of the corresponding node.

[0112] Based on the integral of the stress response amplitude over time, and combined with the geometric deformation sensitivity coefficient of the preset mesh nodes, the initial stress accumulation is generated.

[0113] By introducing a stress transfer efficiency correction factor between adjacent nodes, the initial stress accumulation is spatially correlated and corrected to obtain the corrected nodal stress field distribution.

[0114] The dynamic stress accumulation factor of each grid node is output by calculating the ratio of the energy dissipation rate of each node in the stress field distribution to the material fatigue limit.

[0115] In one specific embodiment of the invention, during the testing of the filter capacitor, the first three eigenvectors (e.g., contribution rates of 72%, 18%, and 7%) are extracted from the strain-discharge principal component coupling coefficient matrix (3×36 dimensions), and the stress wave propagation phase difference (range 5°-25°) of 36 grid nodes on the shell surface is calculated. The phase difference of each node is convolved with the damping characteristic parameters of the aluminum alloy material (loss factor 0.02-0.05) to obtain the stress response amplitude (0.8-1.5 MPa). The response amplitude is integrated over 60 seconds (sampling rate 1 kHz), and combined with the geometric deformation sensitivity coefficient of each node (0.7-1.3 per unit), an initial stress accumulation (45-120 MPa·s) is generated. After spatial correction by introducing a stress transfer efficiency correction factor (0.6-0.9) between adjacent nodes, the node stress field distribution is obtained (peak stress nodes 12 / 19 / 25, value range 38-105 MPa). Finally, the dynamic stress accumulation factor is output by calculating the ratio of the energy dissipation rate of each node to the material fatigue limit.

[0116] In summary, by establishing a quantitative correlation model between strain-discharge characteristics and stress propagation, we have achieved accurate calculation and dynamic updating of the stress state at each node of the shell, and can accurately identify stress concentration areas and potential failure points.

[0117] In addition, the capacitor explosion resistance test method may also include the following steps:

[0118] A conductive polymer film with a cell membrane ion channel topology was prepared on the inner surface of the capacitor shell. By controlling the frequency of the alternating electric field, a dynamic conductivity distribution matching the spatial gradient of the composite stress field was formed on the surface of the film.

[0119] Based on the real-time conductivity distribution parameters of conductive polymer films, the voltage-dependent gating dynamics of sodium and potassium ion channels were simulated, and the spatiotemporal mapping relationship between the local conductivity change of the film and the propagation path of mechanical stress waves was established.

[0120] The spatiotemporal mapping relationship is nonlinearly superimposed with the thermo-electric coupling parameters of the composite stress field to trigger the ion flux oscillation behavior of the Hodgkin-Huxley model and extract the biological rhythm fundamental frequency component in its main frequency band.

[0121] The fundamental frequency component of biological rhythm is input into the biomimetic vibration load modulator, and the amplitude-frequency characteristics of the mechanical vibration mode are nonlinearly modulated by the phase sensitivity of ion flux oscillation to generate a chaotic vibration waveform with bifurcation characteristics.

[0122] The chaotic vibration waveform is coupled with the original mechanical vibration mode in the time domain to form a synthetic vibration load sequence. The impedance relaxation characteristics of the conductive polymer film are monitored simultaneously and the threshold of the gated dynamic parameters is adjusted to maintain the steady-state output of the chaotic vibration.

[0123] Specific Implementation: In the testing of the filter capacitor, a conductive polymer film with a biomimetic ion channel structure (channel diameter 50-80 nm, film thickness 20 μm) was formed on the inner surface of the shell using a polypyrrole / carbon nanotube composite material via electrochemical deposition. By adjusting the alternating electric field frequency (dynamically switching from 5-15 kHz), an axial gradient conductivity distribution (1.2-3.8 S / m gradient difference) was created on the film surface, spatially matching the temperature gradient of the composite stress field (axial 5 °C / cm). Real-time acquisition of film conductivity distribution data (sampling rate 1 MHz) was performed. Based on a sodium ion channel gating dynamics model (activation voltage threshold -55 mV), a spatiotemporal mapping matrix (12 × 36 dimensions) of conductivity change and mechanical stress wave propagation path (120 m / s) was established. This mapping matrix was nonlinearly superimposed with thermo-electric coupling parameters (Joule heating rate 2 °C / s, dielectric constant drift 0.3%) to trigger ion flux oscillations (dominant frequency 0.5-2 Hz, amplitude 15-30 μA / cm). 2 Extract the fundamental frequency component of biological rhythms (0.8 Hz / 25 μA / cm). 2 This component is input into a vibration modulator, which uses the oscillation phase lag characteristic (45°-90°) to nonlinearly modulate the 50Hz fundamental vibration (amplitude 0.5mm) to generate a chaotic vibration waveform (bifurcation period 3.2s). Finally, the chaotic waveform is coupled in the time domain with the original vibration (85Hz / 0.3mm) (coupling coefficient 0.7) to form a synthetic vibration load (peak acceleration 8g). The gate voltage threshold (dynamic adjustment range -60mV to -50mV) is simultaneously adjusted by feedback through the thin-film impedance relaxation spectrum (relaxation time 0.8-1.2ms), thereby maintaining the steady-state output of the chaotic vibration bifurcation characteristic.

[0124] The Hodgkin-Huxley model is a biophysical model describing the relationship between the dynamic changes of ion channels on the neuronal cell membrane and the generation of action potentials. It simulates the dynamic response process of cell membrane potential by quantifying the changes in conductivity of sodium and potassium ion channels. The thermo-electric coupling parameters include the rate of change of material conductivity caused by the temperature gradient field, the amount of dielectric constant drift caused by thermal expansion, and the rate of local temperature rise caused by the Joule heating effect.

[0125] In summary, this method constructs a biomimetic ion channel structure and a gated dynamics model to dynamically couple mechanical stress propagation with changes in electro-thermal parameters, generating chaotic vibration loads with biological rhythm characteristics. This achieves accurate simulation of nonlinear vibration environments under the synergistic effect of multiple physics fields, improving the realism and reliability of mechanical fatigue characteristic testing of capacitors under complex working conditions. At the same time, an adaptive control mechanism is formed through impedance relaxation characteristic feedback adjustment to ensure the stability of the dynamic response of the test system.

[0126] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0127] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0128] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for testing the explosion resistance of power capacitors, characterized in that, Includes the following steps: Obtain the actual operating environment of the target capacitor, construct the composite stress field of the target capacitor when it is working in the actual operating environment, and inject a multi-band harmonic current sequence into the composite stress field; The three-dimensional strain propagation characteristics of the capacitor shell are captured in real time by a distributed optical fiber sensor array, and a strain-discharge spatiotemporal correlation model is established based on the three-dimensional strain propagation characteristics. The capacitor shell structure integrity decay curve is obtained based on the strain-discharge spatiotemporal correlation model. When the slope of the decay curve exceeds the preset critical topology, a multi-level linkage protection device is triggered to release energy. The plasma spectral fingerprint formed by the self-healing properties of the metallized thin film during energy release is extracted. The explosion resistance performance evaluation system of the capacitor is reconstructed in reverse by the characteristic spectral line shift, and the evaluation results are generated.

2. A method of performing an explosion-proof test of a power capacitor according to claim 1, characterized in that, To obtain the actual operating environment of the target capacitor, a composite stress field is constructed when the target capacitor operates in the actual operating environment, specifically as follows: By deploying a sensor network at preset locations on the target capacitor, the temperature fluctuation spectrum, the time-varying envelope characteristics of the voltage impulse waveform, and the frequency domain energy distribution of the mechanical vibration source in its actual operating environment are collected in real time. The temperature fluctuation spectrum is decomposed into a fundamental temperature component and multiple subharmonic temperature components. At the same time, the leading transient parameters and subsequent oscillation modes are separated from the voltage impulse waveform, and the position of the main resonance peak in the frequency domain energy distribution of mechanical vibration is extracted. The fundamental temperature component and the subharmonic temperature component are reconstructed in phase synchronization to form a multilayer alternating temperature gradient field with dynamic phase difference. At the same time, an asymmetric voltage shock wave time domain sequence is generated with the leading transient parameters as initial conditions and the oscillation mode as boundary constraints. The position of the main resonance peak is mapped to the excitation frequency of the mechanical vibration mode. Based on the spatial distribution parameters of the multilayer alternating temperature gradient field, thermal deformation compensation correction is performed on the propagation path of the time-domain sequence of the asymmetric voltage shock wave, so that the leading edge of the voltage shock wave and the antinode node of the mechanical vibration mode form a spatiotemporal coupling effect. By comparing the covariance matrix of the dynamic stress response spectrum under the spatiotemporal coupling effect with that of the actual environmental parameter set, the layering weight of the temperature gradient field, the damping coefficient of the voltage shock wave oscillation, and the amplitude-frequency ratio of the mechanical vibration excitation are iteratively adjusted until the multi-physics field synergy factor of the composite stress field reaches the preset spatiotemporal distribution consistency threshold.

3. A method of performing an explosion-proof test of a power capacitor according to claim 1, characterized in that, Injecting a multi-band harmonic current sequence into the composite stress field specifically involves: Real-time acquisition of dielectric polarization-relaxation response spectrum inside capacitor, and extraction of its dielectric relaxation time constant as dynamic change rate of dielectric loss under composite stress field. The dielectric relaxation time constant is input into the multi-band harmonic generator, which divides the preset wideband harmonics into subsynchronous oscillation band, fundamental resonant band and supersynchronous resonant band, and generates an initial coding sequence of phase modulation parameters based on the cross-modulation effect between each band. The phase modulation parameters are nonlinearly coupled with the dynamic change rate of dielectric loss to obtain the dielectric loss fluctuation rate. The phase angle offset of the initial coding sequence is dynamically weighted by the dielectric loss volatility to form an adaptive coding sequence with dielectric-thermal coupling characteristics. The carrier phase of each harmonic frequency band is controlled in real time according to the adaptive coding sequence to generate a multi-band harmonic current sequence with dynamic phase deflection characteristics.

4. The method for explosion resistance testing of power capacitors according to claim 1, characterized in that, The three-dimensional strain propagation characteristics of the capacitor casing are captured in real time using a distributed fiber optic sensing array, and a strain-discharge spatiotemporal correlation model is established based on these characteristics. Specifically: A grating sensor array is deployed on the surface of the capacitor shell. The omnidirectional strain gradient vector on the shell surface is synchronously acquired by orthogonal polarization light interferometry demodulation technology, and the elastic waveguide phase change characteristics on its propagation path are extracted. The omnidirectional strain gradient vector is decomposed into an elastic waveguide component and a viscoplastic creep component. The strain energy flux density propagation rate of each preset grid node of the capacitor shell is determined by utilizing the phase change characteristics of the elastic waveguide. A wideband partial discharge signal acquisition system is used to capture the time-frequency ridge characteristics of discharge pulses inside the medium. The strain energy flux density propagation rate and the time-frequency ridge characteristics of the discharge pulses are spatiotemporally registered to establish a covariant tensor field between the strain propagation wavefront and the discharge electromagnetic pulse wavefront. Singular value decomposition is performed on the covariant tensor field to extract the principal component coupling coefficient matrix of strain propagation rate and discharge pulse amplitude-frequency characteristics, and a strain-discharge spatiotemporal correlation feature surface is constructed based on the principal component coupling coefficient matrix. A strain-discharge spatiotemporal correlation model is established based on the strain-discharge spatiotemporal correlation feature surface.

5. A method of performing an explosion-proof test of a power capacitor according to claim 1, characterized by, The capacitor casing structural integrity attenuation curve is obtained based on the strain-discharge spatiotemporal correlation model. When the slope of the attenuation curve exceeds the preset critical topological shape, a multi-level linkage protection device is triggered to release energy. Specifically: The principal component coupling coefficient matrix of strain propagation rate and discharge pulse amplitude-frequency characteristics is extracted from the strain-discharge spatiotemporal correlation model, and the dynamic stress accumulation factor of each preset grid node of the shell is determined based on the principal component coupling coefficient matrix. The dynamic stress accumulation factor is compared point by point with the preset material fatigue damage threshold to screen out stress-exceeding nodes and generate a local failure probability distribution cloud map. Based on the spatial gradient change trend of the local failure probability distribution cloud map, a nonlinear decay curve of the shell structure integrity is fitted, and the change of the second derivative of the decay curve is monitored in real time. When the extreme point of the second derivative exceeds the curvature change threshold defined by the preset critical topological shape, the pre-trigger signal of the multi-level linkage protection device is activated. Based on the intensity level matching of the pre-triggered signal, the conduction timing of the energy release channel is matched to prioritize the release of electromagnetic-mechanical coupling energy in the region where the stress over-limit node is located, while simultaneously adjusting the loading parameters of the composite stress field to suppress cascading failures.

6. A method of performing an explosion-proof test of a power capacitor according to claim 1, characterized by The plasma spectral fingerprint formed by the self-healing properties of the metallized thin film during energy release is extracted. The explosion-proof performance rating system of the capacitor is then reconstructed inversely using the characteristic spectral line shifts, and the evaluation results are generated. Specifically: A high-speed spectral acquisition device is used to capture transient plasma radiation generated by the energy discharge channel, and the time-varying intensity envelope of the characteristic spectral lines of metal vapor is extracted. The time-varying intensity envelope was matched with a standard metallic element emission database to identify characteristic spectral line clusters formed by the recrystallization of the vapor-deposited material during the self-healing process; By calculating the dynamic shift of the main peak wavelength in the characteristic spectral line cluster relative to the reference position, a mapping relationship between plasma expansion rate and thin film self-healing efficiency is established. Based on the mapping relationship, the self-healing intensity gradient of the thin film in different stress regions is quantified to generate a three-dimensional self-healing efficiency distribution map. By combining the statistical characteristics of anisotropic factors in the three-dimensional self-healing efficiency distribution map, a multi-level evaluation index matrix reflecting the overall explosion resistance performance of capacitors is constructed. Based on the weight distribution of anisotropic factors in the multi-level evaluation index matrix, each evaluation index is compared and analyzed with the preset explosion-proof performance level threshold to generate the explosion-proof performance level evaluation result of the target capacitor.

7. A method of performing an explosion-proof test of a power capacitor according to claim 5, wherein The dynamic stress accumulation factor of each preset mesh node in the shell is determined based on the principal component coupling coefficient matrix, specifically as follows: The eigenvectors of strain propagation rate and discharge pulse amplitude-frequency characteristics are extracted from the principal component coupling coefficient matrix, and the stress wave propagation phase difference of each preset grid node is calculated. The stress response amplitude is obtained by convolving the stress wave propagation phase difference with the material damping characteristic parameters of the corresponding node. Based on the integral of the stress response amplitude over time, and combined with the geometric deformation sensitivity coefficient of the preset mesh nodes, the initial stress accumulation is generated. By introducing a stress transfer efficiency correction factor between adjacent nodes, the initial stress accumulation is spatially correlated and corrected to obtain the corrected nodal stress field distribution. The dynamic stress accumulation factor of each grid node is output by calculating the ratio of the energy dissipation rate of each node in the stress field distribution to the material fatigue limit.

8. A method of performing an explosion-proof test of a power capacitor according to claim 1, characterized in that: The composite stress field includes synchronously superimposed alternating temperature gradient fields, asymmetric voltage shock waves, and mechanical vibration modes.

9. The explosion resistance test method for power capacitors according to claim 6, characterized in that: The evaluation indicators include plasma expansion rate, thin film self-healing efficiency gradient, uniformity of three-dimensional self-healing efficiency distribution, and anisotropic factor intensity distribution characteristics.

Images