A surge pulse response test method for high-voltage varistors

Abstract

The application relates to the technical field of surge testing, in particular to a surge pulse response test method for a high-voltage voltage-dependent resistor, which comprises the following steps: obtaining voltage data of the high-voltage voltage-dependent resistor in the whole surge pulse response test process; dividing the whole test process into multiple time intervals; obtaining voltage disturbance factors of the time intervals according to voltage variation characteristics when the time intervals are disturbed by parasitic inductance and ringing phenomena and voltage variation characteristics when the time intervals are disturbed by thermal drift and aging effects, so as to obtain filtering thresholds of the time intervals, and then filtering voltage data of the time intervals. The application adaptively obtains the filtering thresholds of the time intervals, improves the filtering effect of the voltage data in the surge pulse response test process, and further improves the accuracy of the surge pulse response test result of the high-voltage voltage-dependent resistor.

Description

Technical Field

[0001] This application relates to the field of surge testing technology, specifically to a surge pulse response testing method for a high-voltage varistor. Background Technology

[0002] A varistor, as a surge protection device, possesses nonlinear current-voltage characteristics and is therefore also known as a variable resistor. Its core function is to rapidly transition from a high-resistance state to a low-resistance state when encountering transient surges caused by lightning strikes, switching operations, etc., diverting the surge current to ground and maintaining the voltage across the equipment at a safe level. This safe voltage level is known as the clamping voltage or residual voltage. Therefore, accurately measuring the varistor's voltage under surge pulse response conditions is crucial for evaluating its protective performance and ensuring equipment safety.

[0003] In actual measurements, the response voltage of a varistor is subject to various noises and interferences. For example, the grounding lead of the voltage probe, the pins of the device under test (DUT), and the connecting wires all have non-negligible parasitic inductances. Furthermore, the second-order system formed by parasitic inductance and capacitance experiences damped free oscillations under step excitation. This results in the measured voltage waveform being a superposition of the actual response voltage and various noise signal components. Wavelet threshold filtering algorithms are typically used to process this data, reducing the impact of noise interference on the accuracy of the voltage data and ensuring the accuracy of the measurement results. However, traditional wavelet threshold filtering algorithms usually use a global threshold to process the acquired data, failing to consider the actual interference situation of the varistor's voltage data under different time states. This leads to insufficient filtering strength for voltage data with strong noise interference, or over-filtering for voltage data with low noise interference, resulting in inaccurate filtered voltage data and consequently errors in the surge pulse response test results. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a surge pulse response test method for high-voltage varistors, thereby resolving the existing issues.

[0005] The surge pulse response test method for a high-voltage varistor disclosed in this application adopts the following technical solution:

[0006] One embodiment of this application provides a surge pulse response test method for a high-voltage varistor, the method comprising the following steps:

[0007] Acquire voltage data of the high-voltage varistor throughout the surge pulse response test; divide the entire test process into multiple time intervals;

[0008] Each time interval is divided into multiple high-level periods and low-level periods. Based on the voltage change rate between the first voltage peak and the previous voltage valley in the first high-level period in each time interval, the voltage fluctuation degree in each high and low level period, and the frequency and bandwidth of each peak in the frequency domain of the voltage data in each time interval, the voltage interference coefficient of each time interval is obtained.

[0009] Based on the voltage difference between each time interval and the previous time interval during the high-level period, as well as the increasing rate of voltage rise during the high-level period in each time interval and all time intervals before it, the voltage normality coefficient of each time interval is obtained. Combined with the voltage interference coefficient of each time interval, the voltage disturbance factor of each time interval is obtained. In this way, the filtering threshold of each time interval is obtained, and the voltage data of each time interval is filtered.

[0010] Preferably, each time interval refers to the time period between the n sampling times before the application of each surge pulse and the n sampling times before the application of the next surge pulse, where n is a preset number.

[0011] Preferably, the specific process of dividing each time interval into multiple high-level periods and low-level periods is as follows:

[0012] The sampling time when the voltage value is greater than the preset voltage threshold is recorded as the high-level time; otherwise, it is recorded as the low-level time.

[0013] A high-level period is defined as a time interval consisting of all consecutive high-level moments within any time interval; similarly, low-level periods within each time interval are obtained.

[0014] Preferably, the method for obtaining the voltage interference coefficient for each time interval is as follows:

[0015] The peak slope of each time interval is obtained by measuring the rate of change between the first voltage peak and the previous voltage trough in the first high-level period within each time interval.

[0016] Calculate the variance of all voltage peak values ​​during all high-level periods in each time interval, and the variance of all voltage trough values ​​during all low-level periods in each time interval;

[0017] Obtain the frequency domain response of the voltage data for each time interval; calculate the product between the mean frequency of all peaks except the fundamental frequency in each frequency domain response and the mean signal bandwidth.

[0018] The voltage interference coefficient of each time interval is positively correlated with the peak slope, the two variances mentioned above, and the product.

[0019] Preferably, the method for obtaining the peak slope of each time interval is as follows:

[0020] The first voltage peak in the first high-level period within each time interval is taken as the peak point of each time interval.

[0021] The voltage trough closest to the sampling time before the peak point in each time interval is taken as the valley point of each time interval.

[0022] The slope of the line connecting the valley point and the peak point within each time interval is taken as the peak slope of each time interval.

[0023] Preferably, the method for obtaining the voltage normality coefficient for each time interval is as follows:

[0024] Calculate the difference between the average voltage at all high-level moments in each time interval and the previous time interval;

[0025] The voltage change coefficient for each time interval is obtained by measuring the rate of increase of voltage rise during the high-level period in each time interval and all time intervals preceding it.

[0026] The voltage normality coefficient for each time interval is negatively correlated with the difference and the voltage change coefficient.

[0027] Preferably, the method for obtaining the voltage variation coefficient for each time interval is as follows:

[0028] Calculate the voltage data difference value of all adjacent high-level moments in each time interval, and use the ratio between the sum of all difference values ​​and the sum of the time lengths of all high-level periods in each time interval as the voltage rise rate of each time interval.

[0029] The sequence formed by arranging the voltage rise rate of each time interval and all time intervals before it, and the sampling time of the first voltage data in chronological order, is denoted as the voltage rise rate sequence and time sequence of each time interval.

[0030] The rank correlation coefficient between the voltage rise rate sequence and the time series in each time interval is denoted as the voltage change coefficient in each time interval.

[0031] Preferably, the voltage disturbance factor in each time interval is positively correlated with the voltage disturbance coefficient and negatively correlated with the voltage normal coefficient.

[0032] Preferably, the filtering threshold for each time interval is the product of the normalized result of the voltage disturbance factor in each time interval and the preset maximum threshold of the wavelet threshold filtering algorithm.

[0033] Preferably, the specific process of filtering the voltage data in each time interval is as follows: the voltage data in each time interval is used as the input of the wavelet threshold filtering algorithm, and the threshold of each time interval is set as the calculated filtering threshold of the corresponding time interval, and the output is the filtered voltage data of each time interval.

[0034] This application has at least the following beneficial effects:

[0035] This application addresses the problem that traditional wavelet threshold filtering algorithms typically use a global threshold to process acquired voltage data, leading to inaccurate filtered voltage data. By analyzing the voltage change characteristics of high-voltage varistors during surge pulse response testing under the interference of parasitic inductance and ringing, a voltage interference coefficient is constructed. This coefficient can assess the combined interference level of parasitic inductance and ringing in voltage data across different time intervals. Furthermore, by analyzing the impact of thermal drift and aging effects of the high-voltage varistor under multiple surge impacts on voltage data, a voltage normality coefficient is constructed. This coefficient can assess the normality of voltage data across different time intervals. Combined with the voltage interference coefficient, the degree of noise interference in voltage data across different time intervals is comprehensively evaluated, allowing for adaptive adjustment of the filtering threshold for each time interval. Ultimately, while ensuring filtering effectiveness, this improves the usability and accuracy of voltage data, thereby enhancing the accuracy of surge pulse response test results. Attached Figure Description

[0036] To more clearly illustrate the technical solutions and advantages in the embodiments of this application 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 this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 A flowchart illustrating the steps of a surge pulse response test method for a high-voltage varistor provided in this application;

[0038] Figure 2 A flowchart for obtaining the voltage disturbance factor for each time interval provided in this application. Detailed Implementation

[0039] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a surge pulse response test method for a high-voltage varistor proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0041] The following description, in conjunction with the accompanying drawings, details the specific scheme of the surge pulse response test method for a high-voltage varistor provided in this application.

[0042] This application provides an embodiment of a surge pulse response test method for a high-voltage varistor. Specifically, it provides the following surge pulse response test method for a high-voltage varistor. Please refer to [link to relevant documentation]. Figure 1 The method includes the following steps:

[0043] Step 1: Obtain the voltage data of the high-voltage varistor throughout the surge pulse response test; divide the entire test process into multiple time intervals.

[0044] Securely mount the high-voltage varistor sample to be tested onto a dedicated, well-insulated test fixture. Connect the two input terminals of the high-voltage differential probe to the two ends of the varistor to measure the voltage across them. Connect the output terminal of the high-voltage differential probe to the corresponding channel of the oscilloscope. Connect the surge generator to the varistor's test circuit and initialize it. The initialization process is as follows: according to the test requirements (e.g., IEC61000-4-5), select a suitable surge waveform (e.g., 1.2 / 50μs voltage or 8 / 20μs current) for the surge generator, gradually setting the surge voltage / current level from small to large. In this embodiment, a 10kA 8 / 20μs impulse current is selected as the surge waveform.

[0045] After confirming that all connections and settings are correct, start the surge generator to apply H surge pulses to the high-voltage varistor. The time interval between each surge pulse is Tmin. The oscilloscope will automatically capture and display the voltage waveform across the high-voltage varistor. In this embodiment, H is 100 and T is 1.

[0046] Furthermore, data is sampled from the voltage waveform captured by the oscilloscope at a sampling frequency of UkHz to obtain the voltage data of the high-voltage varistor during the entire surge pulse response test; where U is the preset sampling frequency, and the sampling frequency setting must satisfy the Nyquist sampling theorem; in this embodiment, U is set to 2. The time interval between the n sampling moments before each surge pulse application and the n sampling moments before the next surge pulse application is recorded as a time interval, where n is a preset number; in this embodiment, it is set to 200.

[0047] Step 2: Divide each time interval into multiple high-level periods and low-level periods; based on the voltage change rate between the first voltage peak and the previous voltage valley in the first high-level period in each time interval, the voltage fluctuation degree in each high and low level period, and the frequency and bandwidth of each peak in the frequency domain of the voltage data in each time interval, obtain the voltage interference coefficient of each time interval.

[0048] During surge testing of varistors, the grounding wire of the voltage probe, the pins of the device under test, and the connecting wires all have non-negligible parasitic inductance. When a rapidly changing surge current flows through these inductors, an induced voltage is generated. This voltage is superimposed on the clamping voltage of the varistor, resulting in a higher clamping voltage and forming a clamping voltage spike overshoot, which affects the accuracy of the measurement data. Furthermore, parasitic capacitance exists between the probe itself, the oscilloscope input, and the test fixture. These capacitances, together with the parasitic inductance in the circuit, form an LC oscillation circuit. After the clamping voltage spike overshoot, the LC oscillation circuit is excited, generating a high-frequency, exponentially decaying sinusoidal oscillation, i.e., "ringing." This ringing phenomenon is superimposed on the plateau period and falling edge of the clamping voltage, making the plateau region of the varistor response voltage no longer flat. In the time domain, the plateau region of the voltage data exhibits a gradually decaying ripple vibration; in the frequency domain, this interference manifests as one or more resonant peaks with a wide bandwidth in the high-frequency band. All of these situations can interfere with the collected voltage data. Therefore, it is necessary to analyze the degree of impact of these situations on the voltage data in order to adjust the processing parameters of the filtering algorithm.

[0049] Specifically, under normal circumstances, when a surge occurs, the detection instrument itself has a certain capacitive load (equivalent to a capacitor connected in parallel with the circuit under test, which filters the measured signal, affects the rise and fall times of the measured signal, causes transmission delay, and affects the bandwidth of the transmission interconnection channel). This results in the voltage waveform within a single time interval rising relatively slowly from its normal value to the clamping voltage when subjected to a surge pulse. However, when a surge occurs due to parasitic inductance interference, at the rapid rising edge of the surge pulse, due to the high di / dt effect, the induced voltage generated by the parasitic inductance is superimposed on the response voltage of the varistor, causing the voltage within a single time interval to rise rapidly when subjected to a surge pulse.

[0050] To characterize the above features, all voltage values ​​within each time interval are used as input to the Otsu thresholding method to obtain a segmentation threshold, denoted as the preset voltage threshold. Sampling times where the voltage value is greater than the preset voltage threshold are recorded as high-level moments, and vice versa. A time interval consisting of all consecutive high-level moments within any time interval is recorded as a high-level period. Similarly, low-level periods within each time interval are obtained. The Otsu thresholding method is a well-known technique, and its specific process will not be elaborated further. A peak and trough detection algorithm is used to obtain all peaks and troughs of the voltage data within each time interval. The first voltage peak in the first high-level period within each time interval is taken as the peak point of each time interval, and the voltage trough closest to its sampling time before the peak point is taken as the trough point of each time interval. The slope of the line connecting the trough point and the peak point within each time interval is taken as the peak slope of each time interval. The larger the peak slope, the faster the voltage rises when impacted by a surge pulse within the corresponding time interval, and the greater the degree of interference from parasitic inductance on the voltage data of the corresponding time interval. It should be noted that the first peak in the first high-level period within each time interval represents the moment when the surge pulse is first applied within that time interval.

[0051] Furthermore, due to the ringing phenomenon, the voltage waveform exhibits decaying ripple vibrations on both the rising and falling edges. Overall, this manifests as significant amplitude fluctuations during both the high and low level periods of the voltage data waveform, and the presence of numerous high-frequency, wide-bandwidth peaks in the frequency domain.

[0052] To characterize the above features, the variance of all voltage peak values ​​during all high-level periods in each time interval is taken as the high-level fluctuation of the voltage data in each time interval; the variance of all voltage trough values ​​during all low-level periods in each time interval is taken as the low-level fluctuation of the voltage data in each time interval. The larger the obtained high-level and low-level fluctuations, the more significant the amplitude fluctuation of the voltage data in the corresponding time interval, and the more likely the corresponding time interval is to be disturbed by ringing. Simultaneously, the frequency domain response of the voltage data in each time interval is obtained using a Fourier transform algorithm, and the product of the mean frequency of all peaks (excluding the fundamental frequency) in the frequency domain response and the mean signal bandwidth is taken as the frequency domain characteristic coefficient of each time interval. The larger the obtained frequency domain characteristic coefficient, the higher the frequency and wider the bandwidth of the voltage data peaks in the frequency domain in the corresponding time interval, and the more likely the corresponding time interval is to be disturbed by ringing.

[0053] To avoid the impact of inconsistent dimensions on subsequent calculations, the min-max normalization method is used to normalize the peak slope, frequency domain characteristic coefficients, high-level fluctuations, and low-level fluctuations for all time intervals throughout the entire test process, resulting in normalized peak slopes, frequency domain characteristic coefficients, high-level fluctuations, and low-level fluctuations for each time interval. The min-max normalization method is a well-known technique, and its specific process will not be elaborated further.

[0054] In a preferred embodiment, the voltage interference coefficient for each time interval is obtained based on the rate of voltage change between the first voltage peak and the preceding voltage trough in the first high-level period within each time interval, the voltage fluctuation degree in each high and low-level period, and the frequency and bandwidth of each peak in the frequency domain of the voltage data in each time interval. This coefficient characterizes the combined interference degree of parasitic inductance and ringing phenomena on the voltage data in each time interval. The method for obtaining the voltage interference coefficient for each time interval is as follows: obtaining the peak slope of each time interval; statistically analyzing the high-level and low-level fluctuations of the voltage data in each time interval; obtaining the frequency domain characteristic coefficient of each time interval; the voltage interference coefficient of each time interval is positively correlated with the peak slope, the high-level and low-level fluctuations, and the frequency domain characteristic coefficient. The positive correlation means that the dependent variable increases (decreases) as the independent variable increases (decreases).

[0055] Preferably, in this embodiment, the voltage interference coefficient of the i-th time interval is denoted as... Its specific expression is: In the formula, Let be the voltage interference coefficient for the i-th time interval; This is the normalized value of the peak slope in the i-th time interval; is the normalized value of the frequency domain characteristic coefficients for the i-th time interval; This is the mean of the normalized values ​​of the high-level fluctuation and low-level fluctuation of the voltage data within the i-th time interval; This is a preset constant to prevent When the value is 0, other parameters become meaningless. The value range is [0.01, 0.1]. In this embodiment, the value is 0.01.

[0056] income The larger the value, the stronger the interference from parasitic inductance and ringing in the circuit is to the voltage data in the i-th time interval. Therefore, when filtering the voltage data in this time interval in the future, a larger filtering parameter should be used to ensure the noise reduction effect.

[0057] Step 3: Based on the voltage difference between each time interval and the previous time interval during the high-level period, and the increasing rate of voltage rise during the high-level period in each time interval and all time intervals before it, obtain the voltage normality coefficient of each time interval. Combined with the voltage interference coefficient of each time interval, obtain the voltage disturbance factor of each time interval. In this way, obtain the filtering threshold of each time interval, and then filter the voltage data of each time interval.

[0058] To test the actual surge resistance performance of high-voltage varistors, multiple surge impact tests are typically performed. During continuous and repeated surge pulse impacts, the varistor absorbs a large amount of energy and generates heat, leading to thermal drift in the varistor's voltage. Furthermore, as the test progresses, the varistor may also exhibit an aging effect, resulting in a systematic increase in voltage data while having a relatively small impact on changes in the voltage waveform. Therefore, analysis based solely on step two may overlook the interference of thermal drift and aging effects on the data during the test process, requiring further analysis.

[0059] Specifically, as the surge test proceeds, the temperature of the varistor continuously rises, primarily manifested in a gradual upward trend in voltage data across different time intervals over time. Furthermore, due to the aging effect, the more surge impacts the varistor experiences during the test, the more pronounced the aging effect becomes, exhibiting an overall accelerated aging process. This is mainly reflected in the fact that the rate of rise of the high-level voltage data in each time interval gradually increases with the number of tests, showing a clear positive correlation with the number of tests.

[0060] To characterize the above features, for voltage data within a single time interval, the voltage data difference values ​​of all adjacent high-level moments within that time interval are calculated. The ratio between the sum of all difference values ​​and the sum of the durations of all high-level periods within that time interval is taken as the voltage rise rate of that time interval. Further, the voltage rise rates of all time intervals throughout the entire test process are obtained, and the sequence formed by arranging the voltage rise rates of each time interval and all preceding time intervals in chronological order is recorded as the voltage rise rate sequence of each time interval. Then, the sampling time of the first voltage data in each time interval is taken as the time factor of that time interval, and the sequence formed by arranging the time factors of each time interval and all preceding time intervals in chronological order is recorded as the time series of each time interval. Finally, the voltage rise rate sequence and the time series of each time interval are taken as inputs, and the rank correlation coefficient algorithm is used to output the rank correlation coefficient between the two sequences, which is recorded as the voltage change coefficient of each time interval. When the rank correlation coefficient is greater than 0, it indicates that the rate of voltage change before each time interval increases with time. The larger the value, the more significant the interference of thermal drift and aging on the test as the test progresses. When the rank correlation coefficient is 0 or negative, it indicates that the rate of voltage change before each time interval does not increase with time, meaning that the interference of thermal drift and aging on the test as the test progresses is less significant. The rank correlation coefficient algorithm is a well-known technique, and the specific process will not be elaborated further.

[0061] Based on the above analysis, as a preferred implementation, a voltage normality coefficient for each time interval is obtained based on the voltage difference between each time interval and the previous time interval during the high-level period, as well as the increasing rate of voltage rise during the high-level period in each time interval and all time intervals prior to it. This coefficient is used to characterize the normality of the voltage data in each time interval. The method for obtaining the voltage normality coefficient for each time interval is as follows: Calculate the difference between the average voltage values ​​of all high-level moments in each time interval and the previous time interval; obtain the voltage change coefficient for each time interval; the voltage normality coefficient for each time interval is negatively correlated with both the difference and the voltage change coefficient. This negative correlation means that the dependent variable decreases (increases) as the independent variable increases (decreases).

[0062] Preferably, in this embodiment, the voltage normal coefficient of the i-th time interval is denoted as... Its specific expression is: In the formula, Let be the voltage normality coefficient for the i-th time interval; , These are the average voltage values ​​at all high-level moments in the i-th and (i-1)-th time intervals, respectively. Let be the voltage change coefficient for the i-th time interval; sig() is the sigmoid function used to normalize the input data. It should be noted that for the first three time intervals in the entire test process, since sufficient data cannot be collected, the voltage normal coefficients for the first three time intervals cannot be accurately calculated. Therefore, the voltage normal coefficients for the first three time intervals are set equal to the voltage normal coefficients for the fourth time interval.

[0063] income The larger the value, the less the voltage data in each time interval is affected by thermal drift and aging effects, and the more normal the change characteristics of the voltage data. Therefore, when filtering the voltage data in this time interval, a smaller filtering parameter should be used.

[0064] Furthermore, based on the voltage interference coefficient and voltage normality coefficient of each time interval, a voltage disturbance factor for each time interval is obtained to characterize the degree of interference present in the voltage data of each time interval. The voltage disturbance factor for each time interval is positively correlated with the voltage interference coefficient and negatively correlated with the voltage normality coefficient, respectively. The flowchart for obtaining the voltage disturbance factor for each time interval is shown below. Figure 2 As shown.

[0065] Preferably, in this embodiment, the voltage disturbance factor of the i-th time interval is denoted as... Its specific expression is: In the formula, Let be the voltage disturbance factor for the i-th time interval; Let be the voltage interference coefficient for the i-th time interval; Let be the voltage normality coefficient for the i-th time interval.

[0066] Thus, the voltage disturbance factor for all time intervals during the entire test process was calculated.

[0067] Furthermore, this application employs a wavelet threshold filtering algorithm to filter the voltage data throughout the entire testing process. During the filtering process, the size of the filtering threshold has a crucial impact on the filtering effect. A larger filtering threshold results in stronger noise suppression; conversely, a smaller filtering threshold allows for higher noise tolerance and the retention of more detailed information. Therefore, the larger the voltage disturbance factor in each time interval, the larger the filtering threshold should be set for that time interval.

[0068] Based on the above characteristics, the filtering threshold for each time interval is calculated according to the voltage disturbance factor of each time interval. The specific expression is as follows: In the formula, Let be the filtering threshold for the i-th time interval; The voltage disturbance factor in the i-th time interval; This is the hyperbolic tangent function, used to normalize input data; This is the preset maximum threshold for the wavelet threshold filtering algorithm. The value range is 90~150V, and 100V is used in this embodiment. The implementer should adjust it according to the actual measured voltage data.

[0069] The more significant the noise interference in the voltage data of the i-th time interval, the larger the threshold needs to be set to remove the noise interference; conversely, the smaller the threshold should be set to preserve the characteristics of the signal as much as possible.

[0070] The filtering thresholds for all time intervals during the entire test can be calculated using the above method. The voltage data for each time interval is used as input to the wavelet threshold filtering algorithm. During filtering, the threshold for each time interval is set to the calculated filtering threshold for that time interval. The decomposition level is set to 5 levels, and the output is the filtered voltage data for each time interval.

[0071] The above method adaptively filters the voltage data throughout the test process, enabling the filtered voltage data to more accurately represent the surge pulse response process of the varistor, thereby making the surge pulse response test results of the varistor more accurate.

[0072] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, specific embodiments of this specification have been described above. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0073] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0074] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them; modifications to the technical solutions described in the foregoing embodiments, or equivalent substitutions of some of the technical features, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method of testing the response of a high voltage varistor to a surge pulse, characterized by, The method comprises the following steps: obtaining voltage data of the high-voltage pressure-sensitive resistor during the entire surge impulse response test process; dividing the entire test process into multiple time intervals; dividing each time interval into multiple high-level time periods and low-level time periods; obtaining a voltage disturbance coefficient of each time interval according to a voltage change rate between a first voltage peak in the first high-level time period in each time interval and a previous voltage valley, a voltage fluctuation degree of each high-level time period and low-level time period, and a frequency and bandwidth of each wave peak in a frequency domain of the voltage data of each time interval; obtaining a voltage normal coefficient of each time interval according to a high-level time period voltage difference between each time interval and a previous time interval, and an incremental degree of a voltage rising rate of the high-level time period in each time interval and all previous time intervals, and obtaining a voltage disturbed factor of each time interval in combination with the voltage disturbance coefficient of each time interval, so as to obtain a filtering threshold of each time interval, and then filtering the voltage data of each time interval; the voltage disturbed factor of each time interval is in positive correlation with the voltage disturbance coefficient and in negative correlation with the voltage normal coefficient; the filtering threshold of each time interval is a product between a normalized result of the voltage disturbed factor of each time interval and a preset maximum threshold of a wavelet threshold filtering algorithm.

2. The surge pulse response test method of a high voltage varistor according to claim 1, wherein each time interval refers to a time period between the first n sampling time points when each surge impulse is applied and the first n sampling time points when the next surge impulse is applied, and n is a preset number.

3. The surge pulse response test method for a high voltage varistor according to claim 1, wherein The specific process of dividing each time interval into multiple high-level time periods and low-level time periods is as follows: sampling time points with voltage values greater than a preset voltage threshold are recorded as high-level time points; otherwise, as low-level time points; all high-level time points in any time interval are recorded as a high-level time period; similarly, low-level time periods in each time interval are obtained.

4. The surge pulse response test method for a high voltage varistor according to claim 1, wherein The method for obtaining the voltage disturbance coefficient of each time interval is as follows: obtaining a peak slope of each time interval according to a voltage change rate between a first voltage peak in the first high-level time period in each time interval and a previous voltage valley; statistically obtaining a variance of all voltage peak values of all high-level time periods in each time interval, and a variance of all voltage valley values of all low-level time periods in each time interval; respectively obtaining a frequency domain response of the voltage data of each time interval; and statistically obtaining a product between a frequency mean value of all wave peaks except a fundamental frequency and a signal bandwidth mean value in each frequency domain response; the voltage disturbance coefficient of each time interval is in positive correlation with the peak slope, the variance of the voltage peak value, the variance of the voltage valley value, and the product.

5. A surge pulse response test method for a high voltage varistor as defined in claim 4, wherein The method for obtaining the peak slope of each time interval is as follows: taking a first voltage wave peak in the first high-level time period in each time interval as a peak point of each time interval; taking a voltage valley closest to a sampling time point of each time interval before the peak point as a valley point of each time interval; taking a slope of a line between the valley point and the peak point in each time interval as the peak slope of each time interval.

6. The surge pulse response test method for a high voltage varistor of claim 3, wherein, The method for obtaining the voltage normal coefficient of each time interval is as follows: The difference between the voltage average of all high level time points of each time interval and the previous time interval is calculated; The voltage change coefficient of each time interval is obtained according to the increasing degree of the voltage rising speed of the high level time period in each time interval and all previous time intervals; The voltage normal coefficient of each time interval is negatively correlated with the difference and the voltage change coefficient.

7. A surge pulse response test method for a high voltage varistor as defined in claim 6, wherein The method for obtaining the voltage change coefficient of each time interval is as follows: The voltage data difference value of all adjacent high level time points in each time interval is calculated, and the ratio between the sum of all difference values and the sum of time lengths of all high level time periods in each time interval is taken as the voltage rising speed of each time interval; The sequence composed of the voltage rising speed of each time interval and all previous time intervals and the sampling time of the first voltage data arranged in time sequence order is taken as the voltage rising speed sequence and time sequence of each time interval; The rank correlation coefficient between the voltage rising speed sequence and the time sequence of each time interval is taken as the voltage change coefficient of each time interval.

8. The surge pulse response test method for a high voltage varistor of claim 1, wherein, The specific process of filtering the voltage data of each time interval is as follows: the voltage data of each time interval is taken as the input of the wavelet threshold filtering algorithm, the threshold value of each time interval is set as the filtering threshold value of the corresponding time interval calculated, and the output is the filtered voltage data of each time interval.

Images