Packaging material space charge sampling method and system based on burst circuit

Abstract

The application relates to the technical field of power equipment insulation detection, in particular to a packaging material space charge sampling method and system based on a Burst circuit, which comprises the following steps: constructing a Burst pulse power supply, and triggering the Burst pulse power supply to emit a high-frequency test pulse sequence to a packaging material sample based on polarity reversal of a polarization voltage; synchronously collecting original space charge waveforms generated by the packaging material sample under the test pulse excitation and corresponding instantaneous polarization voltages; performing waveform recovery processing on the collected original space charge waveforms, processing the original space charge waveforms by using a time-varying recovery matrix, and obtaining high-fidelity space charge distribution waveforms; and performing average noise reduction processing on the high-fidelity space charge distribution waveforms subjected to the waveform recovery processing, and outputting packaging material space charge distribution data. The dynamic evolution law of charge rapid injection, migration and dissipation in the polarity reversal process can be completely captured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of insulation detection of power equipment, in particular to a packaging material space charge sampling method and system based on a Burst circuit. BACKGROUND

[0002] As the core equipment of a DC power transmission system, a converter transformer is subjected to frequent polarity reversal electric fields during operation, and the dynamic evolution of space charges directly affects the aging process and service life of the insulation material. Therefore, accurately capturing the distribution and migration law of space charges in the packaging material during polarity reversal is of great significance for studying the insulation failure mechanism and ensuring the safe operation of the equipment.

[0003] The existing space charge test of packaging materials mostly uses a traditional pulse power combined with a PEA (Pulsed Electro-Acoustic), but there are two key problems: first, the low frequency of short-time pulses of the traditional pulse power makes it difficult to capture the dynamic evolution law of the rapid injection, migration and dissipation of charges during polarity reversal; second, the original electrical signals obtained by sampling are distorted, including medium absorption attenuation and dispersion in the sound wave propagation process, geometric diffusion attenuation caused by the coaxial structure, non-uniform changes in sound velocity caused by temperature gradient, and signal overshoot ringing caused by the low-pass filter system composed of sensors and amplifiers. The traditional waveform recovery algorithm mostly uses fixed parameters for static compensation, which cannot adapt to the dynamic changes of temperature and electric field during polarity reversal, resulting in poor accuracy of test results and inability to provide reliable data support for the study of insulation failure mechanism. SUMMARY

[0004] To solve the above problems, the present application provides a packaging material space charge sampling method based on a Burst circuit, which comprises:

[0005] A Burst pulse power is constructed, and the Burst pulse power is triggered to emit a high-frequency test pulse sequence to the packaging material sample based on polarity reversal of the polarization voltage;

[0006] The rising edge of each test pulse is used as a trigger signal to synchronously collect the original space charge waveform generated by the packaging material sample under the excitation of the test pulse and the corresponding instantaneous polarization voltage;

[0007] The collected original space charge waveform is subjected to waveform restoration processing, the waveform restoration processing including attenuation dispersion compensation, geometric diffusion compensation, temperature-sound speed dynamic mapping compensation and deconvolution overshoot compensation; attenuation dispersion compensation matrix, geometric diffusion compensation matrix and deconvolution compensation matrix are determined based on the waveform restoration processing, and the sound speed field is determined according to the temperature-sound speed dynamic mapping compensation, and the time-varying restoration matrix is obtained by synthesizing the attenuation dispersion compensation matrix, the geometric diffusion compensation matrix and the deconvolution compensation matrix, and the original space charge waveform is processed by using the time-varying restoration matrix to obtain a high-fidelity space charge distribution waveform.

[0008] The high-fidelity space charge distribution waveform subjected to waveform restoration processing is subjected to average noise reduction processing, and the space charge distribution data of the packaging material is output.

[0009] The packaging material space charge sampling method further includes pre-calibration of the packaging material sample before collecting the original space charge waveform generated by the packaging material sample under the test pulse excitation, and obtaining the sound wave propagation characteristic parameters, the sound wave propagation characteristic parameters including the sound wave attenuation coefficient α and the dispersion coefficient β.

[0010] The attenuation dispersion compensation is based on the pre-calibrated sound wave attenuation coefficient α and dispersion coefficient β, and an attenuation dispersion compensation matrix is constructed, and the operation is specifically as follows:

[0011] The signal frequency range [f1, f2] is subjected to N-point discretization to obtain a frequency vector f = [f1, f2,..., f n ] T ;

[0012] For each discrete frequency point fᵢ, based on the sound wave attenuation coefficient α(fᵢ) and the dispersion coefficient β(fᵢ), and the sound wave propagation distance vector r in the material, the frequency domain attenuation kernel function is constructed as:

[0013] ,

[0014] Where j is the imaginary unit;

[0015] According to the thickness of the packaging material sample, the sample thickness d is equally divided into N sampling units, and the discrete radius r n =j·Δr, Δr = d / N;

[0016] The center frequency f c is selected, and based on the frequency domain attenuation kernel function, an N×N diagonal form attenuation dispersion compensation matrix is constructed as:

[0017] ,

[0018] Where the matrix element is represented as ;

[0019] By introducing time-varying characteristics, the sound wave attenuation coefficient is updated based on a real-time temperature field, and a time-varying attenuation dispersion compensation matrix is obtained M atten (t i ) .

[0020] The geometric diffusion compensation constructs a corresponding diffusion attenuation compensation matrix according to the actual structure of the encapsulating material sample, and the compensation matrix comprises the following steps:

[0021] If the encapsulating material sample is a coaxial structure, the compensation matrix is as follows:

[0022] ,

[0023] wherein, a is the inner radius of the encapsulating material sample, r i is the radius of the discrete sampling point, and N is the total sampling point number of the original electric signal;

[0024] If the encapsulating material sample is a flat plate structure, the compensation matrix is as follows:

[0025] ,

[0026] wherein, d is the sampling distance in the thickness direction of the flat plate sample.

[0027] The specific operation of the temperature-sound velocity dynamic mapping compensation is as follows:

[0028] An ANSYS thermal simulation model matched with the structure of the encapsulating material sample is constructed, and the thermodynamic parameters of the ANSYS thermal simulation model, including thermal conductivity, specific heat capacity and electrical conductivity, are set;

[0029] Based on the real-time collected polarization voltage and polarization current, the ANSYS thermal simulation model is inversed through a Kalman filtering algorithm to obtain an instantaneous temperature field of the encapsulating material;

[0030] The inversed instantaneous temperature field is substituted into a sound velocity-temperature linear equation of the encapsulating material to map the temperature field into a sound velocity field, and a sound velocity distribution vector corresponding to the test pulse is obtained.

[0031] The time-varying recovery matrix formula is as follows:

[0032] ,

[0033] wherein, M decov represents a deconvolution compensation matrix; M atten is an attenuation dispersion compensation matrix; and M diff is a geometric diffusion compensation matrix; and vi vi represents the acoustic wave propagation speed of the encapsulating material at the i-th test pulse moment; v0 represents the reference acoustic speed of the encapsulating material.

[0034] The high-frequency test pulse sequence transmits no less than 100 pulses in 0.5 seconds, and the single pulse width is in nanoseconds.

[0035] The application further provides a packaging material space charge sampling system based on a Burst circuit, which comprises a Burst pulse power supply, a polarization voltage power supply, a charge test unit, an oscilloscope and a waveform processing module.

[0036] The Burst pulse power supply is used for sending a trigger signal according to a preset timing logic, forming an initial pulse, and performing power amplification and waveform shaping on the initial pulse, and outputs a high-voltage burst pulse of 0-100kV.

[0037] The polarization voltage power supply is used for outputting a polarization voltage to provide an electric field environment simulating an actual operation condition for the packaging material sample.

[0038] The charge measurement unit is used for placing the packaging material sample, and applying the high-voltage burst pulse and the polarization voltage; and collecting an acoustic signal generated by the space charge of the packaging material sample under the excitation of the high-voltage burst pulse, and converting the acoustic signal into an electric signal.

[0039] The oscilloscope is used for receiving the electric signal output by the charge test unit to form an original space charge waveform.

[0040] The waveform processing module is used for performing waveform recovery processing on the original space charge waveform formed by the oscilloscope to obtain space charge distribution data of the packaging material sample.

[0041] The Burst pulse power supply comprises a direct current power supply, a plurality of high-power discharge modules, a high-power pulse transformer, a Marx power source and an FPGA timing control unit; the direct current power supply output is electrically connected with the input ends of the plurality of high-power discharge modules; the output ends of the high-power discharge modules are electrically connected with the input end of the high-power pulse transformer; the output end of the high-power pulse transformer is electrically connected with the Marx power source; and the FPGA timing control unit is electrically connected with the direct current power supply, the Marx power source and the high-power discharge modules.

[0042] The high-power discharge module comprises an isolation silicon stack, a high-voltage energy storage pulse capacitor, a high-power thyristor, a trigger control unit and an isolation power supply unit; the isolation silicon stack is electrically connected with the high-voltage energy storage pulse capacitor, and the high-power thyristor is electrically connected with the high-voltage energy storage pulse capacitor, the trigger control unit and the isolation power supply unit respectively.

[0043] Beneficial effects: the application is a packaging material space charge sampling method and system based on a Burst circuit, and through a Burst pulse power supply, hundreds of high-voltage test pulse sequences are output in a key sub-second window of polarity reversal, so that the dynamic evolution law of charge rapid injection, migration and dissipation in the polarity reversal process can be completely captured. At the same time, four-stage dynamic compensation is adopted, combined with ANSYS real-time thermal simulation, to realize dynamic and accurate compensation of temperature, geometry, attenuation, overshoot and other factors, and the accuracy of measurement is improved; the test pulse sequence of the Burst pulse voltage is synchronized and aligned with the polarization voltage, so as to ensure the consistency of the sampling data and the physical process, and the output space charge distribution data can intuitively obtain the charge evolution in the whole process of polarity reversal of the packaging material. BRIEF DESCRIPTION OF DRAWINGS

[0044] The schemes and advantages of the present application will become apparent to those of ordinary skill in the art by reading the following detailed description of the preferred embodiments with reference to the accompanying drawings. The drawings are for purposes of illustrating the preferred embodiments and are not to be construed as limiting the present application.

[0045] In the drawings:

[0046] Figure 1 The flowchart of the packaging material space charge sampling method of the application;

[0047] Figure 2 The circuit structure diagram of the Burst pulse power supply;

[0048] Figure 3 The structure diagram of the packaging material test system;

[0049] Figure 4 The space charge distribution waveform comparison diagram before and after waveform recovery processing of the application;

[0050] Figure 5 The waveform diagram of single test pulse collection;

[0051] Figure 6 The waveform diagram after Burst mode average noise reduction processing. DETAILED DESCRIPTION

[0052] Exemplary embodiments of the present disclosure will be described in greater detail below with reference to the accompanying drawings.

[0053] EMBODIMENT

[0054] Referring toFigure 3 The space charge test of the encapsulating material is generally obtained through experiments by an encapsulating material test system, which comprises a high-voltage pulse power supply, a polarization voltage power supply, a charge test unit and an oscilloscope, the output ends of the high-voltage pulse power supply and the polarization voltage power supply are respectively coupled to the high-voltage electrode of the charge test unit by a capacitance and a resistance with matching values, the charge test unit is placed with an encapsulating material sample, and the output end of the charge test unit is electrically connected to the oscilloscope.

[0055] The core role of the polarization voltage power supply is to provide an electric field environment simulating the actual operation condition for the encapsulating material sample, and the working process is as follows: a basic voltage waveform required by the experiment is generated by a waveform generator, the basic waveform is then input to a power amplifier, and after being amplified to the polarization voltage amplitude set by the experiment by the power amplifier, the polarization voltage is output and applied to the charge measurement unit, so as to provide an electric field excitation for the generation and migration of the space charge in the sample.

[0056] The charge measurement unit is directly in contact with the encapsulating material sample, and the internal electrode structure and sample placement mode are as follows: the upper electrode is a semiconductor electrode, the lower electrode is an aluminum electrode, and the encapsulating material sample to be tested is closely placed between the upper and lower electrodes to ensure good contact between the electrodes and the sample and avoid signal distortion caused by poor contact. Meanwhile, the charge test unit also integrates a PVDF sensor and a signal amplifier inside, the PVDF sensor is used to convert the acoustic signal generated by the charge into an electric signal, and the signal amplifier is used to amplify the weak electric signal for providing a clear signal source for subsequent oscilloscope acquisition.

[0057] As the core inventive point of the technical scheme of the embodiment, the scheme adopts a Burst pulse power supply to replace the high-voltage pulse power supply in the traditional encapsulating material test system. The embodiment provides an encapsulating material space charge sampling system based on a Burst circuit, which comprises a Burst pulse power supply, a polarization voltage power supply, a charge test unit, an oscilloscope and a waveform processing module, the output ends of the Burst pulse power supply and the polarization voltage power supply are respectively coupled to the high-voltage electrode of the charge test unit by a capacitance and a resistance with matching values, the charge test unit is placed with an encapsulating material sample, the output end of the charge test unit is electrically connected to the oscilloscope, and the oscilloscope is electrically connected to the waveform processing module.

[0058] The Burst pulse power supply is used to send a trigger signal according to a preset timing logic, form an initial pulse, and perform power amplification and waveform shaping on the initial pulse to output a high-voltage burst pulse of 0-100 kV.

[0059] The polarization voltage power supply is used to output a polarization voltage to provide an electric field environment simulating the actual operation condition for the encapsulating material sample.

[0060] A charge measuring unit is used to place a packaging material sample and apply a high-voltage burst pulse and a polarization voltage; an acoustic signal generated by space charge of the packaging material sample under the high-voltage burst pulse excitation is collected, and the acoustic signal is converted into an electric signal;

[0061] An oscilloscope is used to receive the electric signal output by the charge collecting unit to form an original space charge waveform;

[0062] A waveform processing module is used to perform waveform recovery processing on the original space charge waveform formed by the oscilloscope to obtain space charge distribution data of the packaging material sample.

[0063] As shown in Figure 2 The burst pulse power supply includes a direct current power supply, a plurality of high-power discharge modules, a high-power pulse transformer, a Marx power source and an FPGA timing control unit; the output end of the direct current power supply is electrically connected with the input ends of the plurality of high-power discharge modules, the output ends of the high-power discharge modules are electrically connected with the input end of the high-power pulse transformer, the output end of the high-power pulse transformer is electrically connected with the Marx power source, and the FPGA timing control unit is electrically connected with the direct current power supply, the Marx power source and each high-power discharge module.

[0064] The direct current power supply is externally connected with a 220V alternating current power supply and performs AC-DC conversion to provide a direct current output voltage.

[0065] The high-power discharge module is used to receive a trigger signal of the FPGA timing control unit to generate an initial pulse.

[0066] The Marx power source is used to receive the initial pulse of the high-power discharge module, amplify and shape the initial pulse to output a high-voltage pulse.

[0067] The FPGA timing control unit is used to accurately control the working timing of the high-power discharge module and transmit a timing trigger signal to the high-power discharge module.

[0068] Each high-power discharge module includes an isolation silicon stack, a high-voltage energy storage pulse capacitor, a high-power thyristor, a trigger control unit and an isolation power supply unit; the isolation silicon stack and the high-voltage energy storage pulse capacitor are electrically connected, the isolation silicon stack is used to realize electrical isolation, and the high-voltage energy storage pulse capacitor is used to store electrical energy; the high-power thyristor is electrically connected with the high-voltage energy storage pulse capacitor, the trigger control unit and the isolation power supply unit, respectively, the high-power thyristor is used as a switching element to control the release of electrical energy, the trigger control unit receives a trigger signal of the FPGA timing control unit and controls the conduction and turn-off of the thyristor, and the isolation power supply unit provides isolation power supply for the internal circuit of the entire module.

[0069] The working principle of the Burst pulse power supply is that 220V voltage input is converted into 0-5kV direct current power supply to power each high-power discharge module. The FPGA timing control unit sends a trigger signal to each high-power discharge module according to the preset timing logic. After the trigger control unit in the high-power discharge module receives the signal, the high-power thyristor is turned on, the electric energy stored in the high-voltage energy storage pulse capacitor is released through the thyristor, and the initial pulse is formed. The initial pulse is transmitted to the Marx power source after being transformed by the high-power pulse transformer, the Marx power source further amplifies the power and shapes the waveform of the pulse, and finally outputs 0-100kV high-voltage burst pulse for packaging material space charge testing.

[0070] Based on the above packaging material space charge sampling system, the system is used to realize a packaging material space charge sampling method based on a Burst circuit, as shown in Figure 1 The method comprises the following steps:

[0071] S1, a Burst pulse power supply is constructed, and based on the start of polarity reversal of the polarization voltage, a high-frequency test pulse sequence is triggered to be emitted from the Burst pulse power supply to the packaging material sample;

[0072] The Burst pulse power supply is configured to emit a high-frequency test pulse sequence with a number of pulses not less than 100 within 0.5 seconds at the key moment of polarity reversal, so as to densely sample the charge distribution inside the packaging material at the moment and in the short time thereafter.

[0073] S2, the rising edge of each test pulse is taken as a trigger signal, and the original space charge waveform generated by the packaging material sample under the test pulse and the corresponding instantaneous polarization voltage are synchronously collected;

[0074] At the same time, since the waveform recovery processing depends on the intrinsic parameters of the material, before S2 tests the packaging material sample, the packaging material sample is also pre-calibrated to obtain the sound wave propagation characteristic parameters.

[0075] The sound wave attenuation coefficient a and the dispersion coefficient b are measured: a flat plate type standard sample with the same oil immersion process as the sample to be tested is prepared, and the ultrasonic transmission method is used to measure at a reference temperature of 25℃. A wideband ultrasonic transducer (center frequency 5MHz) is used to emit a Chirp signal, and a time domain signal is recorded by an oscilloscope. The attenuation coefficient is calculated based on the frequency spectrum ratio method as follows:

[0076] ,

[0077] Wherein, d is the thickness of the sample, P ref (f) is the reference signal spectrum without the sample, transmission signal spectrum when containing the sample.

[0078] The typical parameters of the encapsulating material are: α = 0.8~1.5 Np / (m·MHz 2 ), β = 0.5~1.2 Np / (m·MHz). The fitting of α and β as a quadratic function of frequency is stored as a compensation parameter library.

[0079] S3, performing waveform restoration processing on the collected original spatial charge waveform, the waveform restoration processing including attenuation dispersion compensation, geometric diffusion compensation, temperature-acoustic velocity dynamic mapping compensation, and deconvolution anti-overshoot compensation; determining an attenuation dispersion compensation matrix, a geometric diffusion compensation matrix, and a deconvolution compensation matrix based on the waveform restoration processing, and determining an acoustic velocity field based on the temperature-acoustic velocity dynamic mapping compensation, and synthesizing the attenuation dispersion compensation matrix, the geometric diffusion compensation matrix, and the deconvolution compensation matrix to obtain a time-varying restoration matrix, and applying the time-varying restoration matrix to process the original spatial charge waveform to obtain a high-fidelity spatial charge distribution waveform;

[0080] Attenuation dispersion compensation: based on the pre-calibrated acoustic wave attenuation coefficient α and dispersion coefficient β, an attenuation dispersion compensation matrix is constructed, and the operation is specifically as follows:

[0081] Discretize the signal frequency range [f1, f2] by N points to obtain a frequency vector f = [f1, f2,..., f n ] T ; the frequency range covers the bandwidth of the PVDF sensor, usually 0.5MHz~10MHz.

[0082] For each discrete frequency point fᵢ, based on the acoustic wave attenuation coefficient α(fᵢ) and the dispersion coefficient β(fᵢ), and the propagation distance vector r of the acoustic wave in the material, the frequency domain attenuation kernel function is constructed as:

[0083] ,

[0084] where j is the imaginary unit;

[0085] According to the thickness of the sample of the encapsulating material, the sample thickness d is equally divided into N sampling units, corresponding to the discrete radius r n =j·Δr, Δr = d / N;

[0086] Select the center frequency f c , based on the frequency domain attenuation kernel function, the attenuation dispersion compensation matrix in the form of N×N diagonal is constructed as:

[0087] ,

[0088] where the matrix element is represented as ;

[0089] By introducing time-varying characteristics, updating the sound wave attenuation coefficient based on the real-time temperature field, and obtaining the time-varying attenuation dispersion compensation matrix M atten (t i ) .

[0090] By multiplying the signal spectrum of the original space charge waveform with the time-varying attenuation dispersion compensation matrix:

[0091] ,

[0092] And then through inverse Fourier transform to obtain the time domain compensation signal. The left multiplication operation of the matrix is equivalent to applying an inverse exponential gain to each depth sampling point in the frequency domain, accurately compensating for the amplitude attenuation and phase distortion of the sound wave in the packaging material due to medium loss and frequency dispersion, and restoring the true space charge distribution profile.

[0093] Geometric diffusion compensation: according to the actual structure of the packaging material sample, a corresponding diffusion attenuation compensation matrix is constructed, which specifically includes:

[0094] When the packaging material sample is a coaxial structure, its compensation matrix is:

[0095] ,

[0096] Where, a is the inner radius of the packaging material sample, r i is the radius of the discrete sampling point, and N is the total number of original electrical signal sampling points. The compensation matrix is multiplied on the left of the original signal vector to eliminate the cylindrical diffusion attenuation.

[0097] When the packaging material sample is a flat plate structure, the compensation matrix is:

[0098] ,

[0099] Where, d is the sampling distance in the thickness direction of the flat plate sample.

[0100] The temperature-sound velocity dynamic mapping compensation is optimized for the thermodynamic properties of the packaging material, specifically including:

[0101] An ANSYS thermal simulation model matching the structure of the packaging material sample is constructed, and the thermodynamic parameters of the ANSYS thermal simulation model are set, including thermal conductivity, specific heat capacity, and electrical conductivity; wherein the thermal conductivity of mineral oil immersed paper λ = 0.18 W / (m·K), the specific heat capacity c = 1400 J / (kg·K), and the electrical conductivity σ(25℃) = 1×10-14 S / m; the density of the mineral oil-impregnated paper and the high-temperature modified oil paper is uniformly taken as 900 kg / m -15 S / m; the density of the mineral oil-impregnated paper and the high-temperature modified oil paper is uniformly taken as 900 kg / m 3 .

[0102] Based on the real-time collected polarization voltage and polarization current, the ANSYS thermal simulation model is inverted by Kalman filtering algorithm to obtain the instantaneous temperature field of the packaging material, and the inversion formula is:

[0103] ,

[0104] wherein, is the Kalman gain, (Δθ(r,t i-1 )) represents the current prediction value based on the previous temperature field Δθ(r,t i-1 ), and I(t i ) is the real-time collected polarization current.

[0105] The instantaneous temperature field inverted is substituted into the sound velocity-temperature linear equation of the packaging material to map the temperature field into a sound velocity field, so as to obtain the sound velocity distribution vector corresponding to the test pulse.

[0106] The sound velocity-temperature linear equation of the packaging material is:

[0107] ,

[0108] In the formula, v 0,oil-paper is the reference sound velocity of the packaging material, and is taken as 1700-1900 m / s; k v is the sound velocity temperature coefficient of the packaging material; v oil-paper (θ) represents the sound wave propagation speed of the packaging material at temperature θ; and θ represents temperature.

[0109] The instantaneous temperature field inverted is substituted into the above sound velocity-temperature linear equation to obtain the sound velocity distribution vector of the i-th test pulse at different sampling time points as:

[0110] ,

[0111] wherein, θ(r j ,t i ) represents the temperature of the j-th sampling point at the i-th test pulse time point.

[0112] Deconvolution de-overshoot compensation: the signal characteristic calibration of the packaging material sample is performed to determine the system function, and a deconvolution compensation matrix is obtained by introducing a filtering constraint, and the process is specifically as follows:

[0113] The off-line calibration is performed by using a standard sample of the encapsulating material: by preparing an oil-paper sample with a known step charge distribution, the oil-paper sample is implanted with a metal foil with a thickness of 10 μm at a thickness direction d / 2 to form a standard charge interface; a direct current polarization voltage is applied for a fixed duration, a single pulse excitation is transmitted by using a Burst pulse power supply, and a standard response waveform h of a PVDF sensor and a signal amplifier is recorded std (t);

[0114] The system function is obtained by performing Fourier transform on the standard response waveform:

[0115] ,

[0116] wherein δ(t) is a Fourier transform function of an ideal unit impulse charge distribution.

[0117] An N×N diagonal matrix is constructed at a frequency domain discrete point fk as:

[0118] ,

[0119] Since the characteristics of the PVDF sensor and the signal amplifier are stable, M decov can be regarded as a time-invariant matrix, therefore, in order to avoid amplification of high-frequency noise, a Wiener filter constraint is introduced, and a deconvolution compensation matrix is obtained as:

[0120] ,

[0121] wherein γ is an inverse of a signal-to-noise ratio; * represents a complex conjugate.

[0122] The attenuation dispersion compensation matrix, the geometric diffusion compensation matrix and the deconvolution compensation matrix are determined according to waveform recovery processing, a sound velocity distribution vector is obtained by temperature-sound velocity dynamic mapping compensation, and a time-varying recovery matrix is synthesized, and the formula is:

[0123] ,

[0124] wherein M decov represents the deconvolution compensation matrix; M atten is the attenuation dispersion compensation matrix; M diff is the geometric diffusion compensation matrix; v i represents a sound wave propagation speed of the encapsulating material at an i-th test pulse moment; v0 is a reference sound speed of the encapsulating material.

[0125] The original space charge waveform is processed by using the time-varying recovery matrix, and a high-fidelity space charge distribution waveform is obtained.

[0126] In order to verify the significant advantages of the method described in this embodiment in waveform recovery accuracy and noise reduction effect, based on the above-mentioned packaged material space charge sampling system, a packaged material sample with a thickness of 300 μm is selected for testing and verification.

[0127] Referring to Figure 4 , the figure shows the comparison between the original space charge waveform without processing (shown by the dashed line) and the high-fidelity space charge waveform after processing by the time-varying recovery matrix described in the present application (shown by the solid line) at the moment of polarity reversal.

[0128] As can be seen from the dashed line in Figure 4 , the original signal has serious distortion:

[0129] Amplitude attenuation: with the increase of propagation distance, the signal amplitude decreases exponentially, and the geometric diffusion caused by the coaxial structure makes the far-end signal extremely weak;

[0130] Waveform broadening: affected by medium dispersion, the peak value of the original waveform is broadened, the resolution is reduced, and the position of the interface charge cannot be clearly distinguished;

[0131] Overshoot ringing: there is obvious reverse overshoot at the electrode interface, which masks the accumulated charge near the interface.

[0132] In contrast, Figure 4 the solid line in shows that after four-stage dynamic compensation in step S3 of the present application, the signal amplitude is accurately corrected, the charge peak value at the far-end cathode interface is restored to the true level; the waveform edge is significantly sharpened, the full width at half maximum is reduced by about 45%, which greatly improves the spatial resolution; the overshoot ringing of the sensor system is effectively eliminated by the deconvolution compensation matrix, and the interface charge distribution is clearly visible; and thanks to the temperature-sound speed dynamic mapping, the position of the waveform peak on the time axis is corrected, eliminating the positioning error caused by the change of sound speed due to the Joule heat during the polarity reversal process.

[0133] S4, the high-fidelity space charge distribution waveform after waveform recovery processing is subjected to average noise reduction processing, and the space charge distribution data of the packaged material is output.

[0134] The average noise reduction processing is to sum and average the recovered waveforms corresponding to 100 burst pulses to generate a space charge distribution snapshot every 0.5 seconds, and the average process is synchronized and aligned with the polarization voltage.

[0135] Referring to Figures 5-6 , the waveform graph of a single test pulse and the waveform graph after 100-pulse average noise reduction processing in the burst mode of the present application are shown. During the polarity reversal process, the space charge signal is extremely weak and is easily affected by high-voltage power supply switching noise and environmental electromagnetic interference. As Figure 5As shown in the curve, a large amount of random white noise is mixed in the single collected signal, and the signal-to-noise ratio is only about 8 dB, resulting in some small body charge signals being submerged in the noise floor and difficult to identify.

[0136] The present application uses the high-frequency pulse sequence emitted by the Burst pulse power source within 0.5 seconds to perform the synchronous average processing of the S4 step. Since the acoustic signal generated by the space charge is coherent with the trigger signal in time, while the random noise is incoherent, according to the signal superposition principle, the signal-to-noise ratio is proportional to the square root of the average number N. For example, if the average number N is 100, the signal-to-noise ratio is improved by 10 times. Figure 6 As shown in the curve, after 100 times of average processing, the amplitude of the random noise is greatly reduced, and the signal-to-noise ratio is improved to about 28 dB. The waveform after noise reduction has a flat baseline, and the weak body charge packet (marked by an arrow in the figure) originally submerged in the noise is clearly visible.

Claims

1. A method for space charge sampling of encapsulating materials based on a Burst circuit, characterized in that, The method comprises the following steps: Constructing a burst pulse power supply, and triggering the burst pulse power supply to emit a high-frequency test pulse sequence to a packaging material sample based on a polarity reversal of a polarization voltage; Taking the rising edge of each test pulse as a trigger signal to synchronously collect an original space charge waveform generated by the packaging material sample under the excitation of the test pulse and a corresponding instantaneous polarization voltage; Performing waveform recovery processing on the collected original space charge waveform, wherein the waveform recovery processing comprises attenuation dispersion compensation, geometric diffusion compensation, temperature-sound velocity dynamic mapping compensation, and deconvolution overshoot compensation; Based on the waveform recovery processing, an attenuation dispersion compensation matrix, a geometric diffusion compensation matrix, and a deconvolution compensation matrix are determined respectively, and a sound velocity field is determined according to the temperature-sound velocity dynamic mapping compensation; a time-varying recovery matrix is synthesized from the attenuation dispersion compensation matrix, the geometric diffusion compensation matrix, and the deconvolution compensation matrix, and the original space charge waveform is processed by using the time-varying recovery matrix to obtain a high-fidelity space charge distribution waveform; The attenuation dispersion compensation is based on a pre-calibrated sound wave attenuation coefficient α and a dispersion coefficient β to construct an attenuation dispersion compensation matrix, and the operation is specifically as follows: N-point discrete is performed on the signal frequency range [f1, f2] to obtain a frequency vector f = [f1, f2,..., fN]. n ] T ; For each discrete frequency point fᵢ, a frequency domain attenuation kernel function is constructed based on the sound wave attenuation coefficient α(fᵢ) and the dispersion coefficient β(fᵢ), and a propagation distance vector r of the sound wave in the material, wherein the frequency domain attenuation kernel function is as follows: Wherein, j is an imaginary unit; According to the thickness of the encapsulating material sample, the sample thickness d is equally divided into N sampling units, corresponding to discrete radii r n = j·Δr, Δr = d / N; Selecting a center frequency f c Based on the frequency domain attenuation kernel function, the N×N diagonal form attenuation dispersion compensation matrix is constructed as: where the matrix elements are denoted by ; By introducing time-varying characteristics, the sound wave attenuation coefficient is updated based on real-time temperature field, and a time-varying attenuation dispersion compensation matrix is obtained M atten (t i ) ; The geometric diffusion compensation constructs a corresponding diffusion attenuation compensation matrix according to the actual structure of the packaging material sample, which comprises: If the packaging material sample is a coaxial structure, the compensation matrix is as follows: wherein, a is the inner radius of the encapsulating material sample, r i is the radius of the discrete sampling points, and N is the total number of sampling points of the original electrical signal; If the packaging material sample is a flat plate structure, the compensation matrix is as follows: wherein, d is the sampling distance in the thickness direction of the flat specimen; The high-fidelity space charge distribution waveform obtained by the waveform recovery processing is subjected to average noise reduction processing to output the space charge distribution data of the packaging material.

2. The method of claim 1, wherein the burst circuit-based encapsulant space charge sampling method is characterized by, Before collecting the original space charge waveform generated by the packaging material sample under the excitation of the test pulse, the packaging material sample is pre-calibrated to obtain sound wave propagation characteristic parameters, which comprise a sound wave attenuation coefficient α and a dispersion coefficient β.

3. The method of claim 1, wherein the burst circuit-based encapsulant space charge sampling method is characterized by, The specific operation of the temperature-sound velocity dynamic mapping compensation is as follows: An ANSYS thermal simulation model matched with the structure of the packaging material sample is constructed, and the thermodynamic parameters of the packaging material of the ANSYS thermal simulation model are set, including thermal conductivity, specific heat capacity, and electrical conductivity; Based on the real-time collected polarization voltage and polarization current, the ANSYS thermal simulation model is inverted by using a Kalman filtering algorithm to obtain an instantaneous temperature field of the packaging material; The inverted instantaneous temperature field is substituted into a sound velocity-temperature linear equation of the packaging material to map the temperature field into a sound velocity field to obtain a sound velocity distribution vector corresponding to the test pulse.

4. The method of claim 1, wherein the method is a method of sampling space charge in a Burst circuit based encapsulant, and The time-varying recovery matrix formula is as follows: where M decov represents the deconvolution compensation matrix; M atten is the attenuation dispersion compensation matrix; M diff is the geometric diffusion compensation matrix; v i represents the sound wave propagation speed of the encapsulating material at the i-th test pulse moment; v0 is the reference sound speed of the encapsulating material.

5. The method of claim 1, wherein the method is a method of sampling space charge in a Burst circuit based encapsulant, and wherein the method further comprises: applying a voltage to the Burst circuit based encapsulant; and measuring the current through the Burst circuit based encapsulant. The high-frequency test pulse sequence has a number of pulses not less than 100 within 0.5 seconds, and the width of a single pulse is in the nanosecond level.

6. A Burst circuit-based encapsulant space charge sampling system implementing the Burst circuit-based encapsulant space charge sampling method of claim 1, characterized by, The burst pulse power supply, the polarization voltage power supply, the charge test unit, the oscilloscope and the waveform processing module, the output end of the burst pulse power supply and the output end of the polarization voltage power supply are coupled by the value matched capacitor and the resistance and accessed to the high voltage electrode of the charge test unit, the charge test unit is placed with the encapsulating material sample, the output end is electrically connected with the oscilloscope, and the oscilloscope and the waveform processing module are electrically connected; The burst pulse power supply is used for sending a trigger signal according to a preset timing logic, forming an initial pulse, power amplifying and waveform shaping the initial pulse, and outputting a high-voltage burst pulse of 0-100kV. The polarization voltage power supply is used for outputting a polarization voltage to provide an electric field environment simulating an actual operation condition for the encapsulating material sample. The charge test unit is used for placing the encapsulating material sample and applying a high-voltage burst pulse and a polarization voltage, collecting an acoustic signal generated by the space charge of the encapsulating material sample under the excitation of the high-voltage burst pulse, and converting the acoustic signal into an electric signal. The oscilloscope is used for receiving the electric signal output by the charge test unit to form an original space charge waveform. The waveform processing module is used for performing waveform recovery processing on the original space charge waveform formed by the oscilloscope to obtain the space charge distribution data of the encapsulating material sample.

7. The Burst circuit based package material space charge sampling system of claim 6, wherein, The burst pulse power supply comprises a direct current power supply, a plurality of high-power discharge modules, a high-power pulse transformer, a Marx power source and an FPGA timing control unit; the output end of the direct current power supply is electrically connected with the input ends of the plurality of high-power discharge modules, the output ends of the high-power discharge modules are electrically connected with the input end of the high-power pulse transformer, the output end of the high-power pulse transformer is electrically connected with the Marx power source, and the FPGA timing control unit is respectively electrically connected with the direct current power supply, the Marx power source and the high-power discharge modules.

8. The Burst circuit based package material space charge sampling system of claim 6, wherein, The high-power discharge module comprises an isolation silicon stack, a high-voltage energy storage pulse capacitor, a high-power thyristor, a trigger control unit and an isolation power supply unit; the isolation silicon stack and the high-voltage energy storage pulse capacitor are electrically connected, and the high-power thyristor is respectively electrically connected with the high-voltage energy storage pulse capacitor, the trigger control unit and the isolation power supply unit.

Images