A high-break high-voltage current-limiting fuse fault diagnosis method

Abstract

The application discloses a high-break high-voltage current-limiting fuse fault diagnosis method, and relates to the technical field of high-voltage fuse fault diagnosis. The method comprises the following steps: collecting high-voltage end and low-voltage end current signals and recording fault indicator output states; constructing a current signal phase difference sequence, and determining an interception interval according to the fault indicator output states; performing adaptive sliding interception in the interception interval to generate multiple groups of phase difference sub-sequences; calculating the phase signal distortion rate of each group of phase difference sub-sequences to determine a fuse fault discrimination factor; acquiring the high-voltage end and low-voltage end voltage falling edge time and calculating a falling edge time difference; and correcting the discrimination factor by using the falling edge time difference to obtain a corrected fuse fault discrimination factor, comparing the corrected fuse fault discrimination factor with a fault discrimination threshold to generate a discrimination result. The method can improve the accuracy of fuse fault diagnosis and reduce the misjudgment rate.

Description

Technical Field

[0001] This invention relates to the field of high-voltage fuse fault diagnosis technology, specifically a method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse. Background Technology

[0002] High-breaking capacity, high-voltage current-limiting fuses play a crucial role in power systems by limiting short-circuit current and isolating faults. Accurate diagnosis of their blown state directly impacts the reliability of protection coordination and the efficiency of power restoration. When a fuse blows, the internal fusible element melts and generates an electric arc, causing a current interruption and a sudden change in voltage across the fuse terminals. Consequently, the phase relationship between the current signals at the high-voltage and low-voltage ends becomes distorted. Existing diagnostic methods generally collect current signals from both sides of the fuse, calculate the power frequency phase difference, and compare it to a fixed threshold. When the phase difference exceeds a set limit, it is determined to be blown. Some solutions incorporate the action signal of a fault indicator as an auxiliary measure, but this is only used to initiate waveform recording or trigger an alarm, failing to precisely define the analysis period in conjunction with the fault indicator's status.

[0003] Existing technical solutions have significant shortcomings. Phase difference comparison methods based on fixed thresholds cannot adapt to the dynamic evolution of current phase difference during the fusing process. In the early stages of fusing, the phase difference fluctuation amplitude may not yet reach the threshold, while instantaneous phase disturbances caused by external load switching or system oscillations may exceed it, leading to both missed and false fault detections. The status indicators output by the fault indicator record the time information of the fault's start and end. Existing methods do not convert this information into interval constraints for phase difference analysis, resulting in a large amount of data from non-fault periods being mixed into the calculation, obscuring fault characteristics. Voltage signals exhibit significant falling edge characteristics during fusing. The time difference between the voltage drop at the high-voltage and low-voltage ends contains information about the fuse's operating characteristics; existing methods generally ignore this parameter and do not incorporate it into the discrimination logic. The assessment of phase distortion lacks an adaptive mechanism, and the fixed-length analysis window cannot simultaneously accommodate the feature extraction requirements of long-term continuous fluctuations and short-term sudden jumps.

[0004] Diagnosing fuse failures requires addressing how to extract current phase difference distortion features with clear fault indication from the time period defined by the fault indicator status, and how to use the time interval between the falling edges of the high and low voltage terminals to correct the judgment results in order to eliminate the interference of phase changes caused by external disturbances on the judgment conclusions. Summary of the Invention

[0005] A method for diagnosing fuse failure in high-breaking capacity, high-voltage current-limiting fuses is provided. Within the interception interval defined by the output state of the fault indicator, the current phase difference sequence is adaptively intercepted and the phase signal distortion rate is calculated to generate a fuse failure discrimination factor. The discrimination factor is then corrected by using the time difference between the voltage falling edges of the high-voltage end and the low-voltage end, so as to achieve accurate identification of fuse failure.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides a method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse. The method includes: performing a preparatory operation for fuse fault diagnosis, wherein the preparatory operation includes acquiring a high-voltage end current signal and a low-voltage end current signal, and recording the output state of a fault indicator; constructing a current signal phase difference sequence based on the high-voltage end current signal and the low-voltage end current signal, and determining a truncation interval of the phase difference sequence according to the output state of the fault indicator; and performing adaptive testing on the current signal phase difference sequence within the truncation interval. Multiple phase difference sub-sequences are generated by sliding and truncating; the phase signal distortion rate of each phase difference sub-sequence is calculated, and the fuse failure discrimination factor is determined based on the phase signal distortion rate; the falling edge time of the high-voltage end voltage and the falling edge time of the low-voltage end voltage are obtained, and the falling edge time difference between the falling edge time of the high-voltage end voltage and the falling edge time of the low-voltage end voltage is calculated; the fuse failure discrimination factor is corrected based on the falling edge time difference to obtain the corrected fuse failure discrimination factor; the corrected fuse failure discrimination factor is compared with the fault discrimination threshold to generate the fuse failure discrimination result.

[0007] This method constructs a sequence using the phase difference between the high-voltage and low-voltage current signals and defines an effective intercept interval using the output status of a fault indicator. This allows for precise pinpointing of the fault occurrence time and eliminates interference from the steady-state operating range. An adaptive sliding intercept method processes the current signal phase difference sequence, generating multiple sets of phase difference sub-sequences. The window width dynamically adjusts with the sequence variance, ensuring that the intercepted sub-sequences closely match the signal's inherent fluctuation characteristics, thereby improving the sensitivity of phase distortion feature extraction. Based on the phase signal distortion rate, a fuse failure discrimination factor is determined, which quantifies the severity of current phase abrupt changes during the fusing process, providing a stable basis for fault identification.

[0008] As a technical solution of the present invention, when constructing the current signal phase difference sequence, zero-crossing detection is performed on the high-voltage end current signal and the low-voltage end current signal respectively, and the corresponding zero-crossing time is extracted. Based on the difference between the zero-crossing time of the high-voltage end current and the zero-crossing time of the low-voltage end current within the same power frequency cycle, the current signal phase difference sequence is constructed, so that the phase difference information can accurately reflect the asynchronous characteristics of the current at both ends of the fuse.

[0009] As a technical solution of the present invention, when determining the truncation interval of the phase difference sequence based on the output state of the fault indicator, the fault start flag and fault end flag in the output state of the fault indicator are parsed, the time point corresponding to the fault start flag is taken as the truncation start point, and the time point corresponding to the fault end flag is taken as the truncation end point. This interval is set as the truncation interval of the phase difference sequence, thereby realizing the automatic calibration of the fault time period and avoiding the uncertainty caused by manually selecting the interval.

[0010] As a technical solution of the present invention, the specific process of adaptive sliding interception includes: setting the initial window width and sliding step size of the sliding window; moving the sliding window within the interception interval by the sliding step size, and after each movement, intercepting the elements in the current signal phase difference sequence that fall within the window to form a set of phase difference subsequences; calculating the sequence variance of each set of phase difference subsequences, and adjusting the window width for the next round of interception based on the sequence variance. Preferably, the window width is reduced when the sequence variance is greater than the upper limit threshold, and increased when the sequence variance is less than the lower limit threshold, so that the window width can be adaptively adjusted according to the degree of local fluctuation of the signal. A narrow window is used to capture details in areas with dense phase abrupt changes, and a wide window is used to smooth noise in areas with gentle phase changes, thereby preserving fault characteristics more completely.

[0011] As a technical solution of the present invention, the process of calculating the phase signal distortion rate and determining the fuse failure discrimination factor includes: performing a phase expansion operation on each group of phase difference subsequences to obtain an expanded phase sequence; calculating the phase jump value between adjacent elements in the expanded phase sequence, and counting the number of jumps where the phase jump value is greater than the jump threshold; using the ratio of the number of jumps to the total number of elements in the expanded phase sequence as the phase signal distortion rate of that group of phase difference subsequences; arranging the phase signal distortion rates corresponding to all groups of phase difference subsequences in the truncation order to form a distortion rate sequence, and using the median of this distortion rate sequence as the fuse failure discrimination factor. The distortion rate obtained by using phase expansion and jump statistics can effectively enhance the expression of the degree of local phase distortion, while using the median as the discrimination factor can resist the interference caused by individual abnormal truncation intervals and improve the robustness of the discrimination result. In the specific implementation of the phase unrolling operation, starting from the second element of each phase difference subsequence, the difference between the current element and the previous element is calculated sequentially. When the difference is greater than π, 2π is subtracted from the current element and all subsequent elements. When the difference is less than -π, 2π is added to the current element and all subsequent elements, so that the phase sequence restores its continuous change characteristics, providing conditions for accurate detection of phase jumps.

[0012] As a technical solution of the present invention, the calculation process of the falling edge time difference includes: performing falling edge detection on the high-voltage side voltage signal, and recording the moment when the amplitude of the high-voltage side voltage signal drops below the falling edge determination ratio of the high-voltage side rated voltage as the falling edge moment of the high-voltage side voltage; performing falling edge detection on the low-voltage side voltage signal, and recording the moment when the amplitude of the low-voltage side voltage signal drops below the falling edge determination ratio of the low-voltage side rated voltage as the falling edge moment of the low-voltage side voltage; and calculating the absolute value of the difference between the two moments as the falling edge time difference. This falling edge time difference reflects the difference in the timing of voltage drops on the high-voltage side and the low-voltage side during the fuse blowing process, and can provide auxiliary fault identification information from the voltage change dimension.

[0013] When correcting the fuse failure discrimination factor, the value of the correction coefficient is determined based on the falling edge time difference. This correction coefficient is negatively correlated with the falling edge time difference. The fuse failure discrimination factor is then multiplied by the correction coefficient to obtain the corrected fuse failure discrimination factor. By introducing falling edge time difference correction, a higher correction coefficient obtained when the voltages on both sides drop almost simultaneously can amplify the phase distortion characteristics. Conversely, when the falling edge time difference is long, the weight of the discrimination factor is reduced, avoiding misjudgments caused by non-fusing voltage fluctuations.

[0014] As a technical solution of the present invention, the process of generating a fuse failure discrimination result includes: subtracting the modified fuse failure discrimination factor from the fault discrimination threshold to obtain a discrimination difference value; when the discrimination difference value is greater than the positive judgment threshold, generating a discrimination result that the fuse has blown; when the discrimination difference value is less than the negative judgment threshold, generating a discrimination result that the fuse has not blown; when the discrimination difference value is between the negative judgment threshold and the positive judgment threshold, generating a discrimination result that the fuse status is pending confirmation. This three-state discrimination method can not only quickly distinguish between blown and non-blown states, but also provide a confirmation prompt for intermediate working conditions with ambiguous features, facilitating further inspection and maintenance, and improving the engineering practicality of the diagnostic method.

[0015] The technical effects and advantages provided by the present invention in the above technical solution are as follows: Within the defined interception interval of the fault indicator output state, adaptive sliding interception is performed on the current signal phase difference sequence to generate multiple sets of phase difference sub-sequences. After setting the initial width and sliding step size of the sliding window, the window moves within the interception interval by the sliding step size and intercepts successively to form multiple sets of phase difference sub-sequences. The sequence variance of each sub-sequence is calculated, and the window width for the next round of interception is dynamically adjusted according to the variance magnitude. When the sequence variance is greater than the upper limit threshold, the window width is reduced to focus the window more on the concentrated fluctuation area and capture the local features of rapid phase difference changes more precisely. When the sequence variance is less than the lower limit threshold, the window width is increased to smooth random noise and extract the overall trend. The multiple sets of sub-sequences obtained by the adaptive window cover phase difference fluctuation patterns at different time granularities. A phase expansion operation is performed on each set of phase difference sub-sequences to obtain an expanded phase sequence. The phase jump value between adjacent elements in the expanded phase sequence is calculated, and the ratio of the number of jumps exceeding the jump threshold to the total number of elements is used as the phase signal distortion rate. All phase signal distortion rates are arranged in the order of truncation to form a distortion rate sequence, and the median is taken as the fuse fault discrimination factor. This method uses the start and end markers of the fault indicator as boundaries to eliminate invalid data during non-fault periods. It adaptively matches the local statistical characteristics of the phase difference sequence through a window, enabling the discrimination factor to focus on reflecting the abrupt changes and drastic fluctuations in the current phase difference during the fusing process, rather than the stable differences during uniform periods. The median calculation remains stable even when there are individual extreme values ​​in the distortion rate sequence, suppressing outlier interference introduced by accidental truncation segments. This ensures that the discrimination factor maintains a sensitive representation of the fault state during both the initial fusing stage and the continuous arcing stage.

[0016] Obtain the falling edge times of the high-voltage and low-voltage terminals of the fuse, and calculate the falling edge time difference between them. Perform falling edge detection on the high-voltage terminal voltage signal, and record the time when the amplitude drops below the falling edge judgment ratio of the high-voltage terminal rated voltage; similarly, detect and record the falling edge time of the low-voltage terminal voltage signal. When the fuse body blows, the voltage drop at the high-voltage and low-voltage terminals occurs almost synchronously with a very small time difference; voltage changes caused by external load disturbances or system voltage fluctuations exhibit significant time delays at the high-voltage and low-voltage terminals, with a larger time difference. Establish a negative correlation between the correction coefficient and the falling edge time difference, and multiply the fuse failure discrimination factor by this correction coefficient to obtain the corrected fuse failure discrimination factor. The smaller the falling edge time difference, the closer the correction coefficient is to the upper limit, maintaining or slightly increasing the value of the discrimination factor; the larger the falling edge time difference, the faster the correction coefficient decreases, weakening the artificially high discrimination factor caused by disturbances. The modified discrimination factor is compared with the fault discrimination threshold. Based on whether the difference falls within the positive, negative, or in-between range of the discrimination threshold, a discrimination result of "fused," "not fused," or "pending confirmation" is generated. The introduction of the time difference between the high and low voltage drop edges adjusts the amplitude of the fault discrimination factor, providing a positive response only to phase distortions consistent with the fuse's operating timing characteristics. It effectively suppresses phase difference changes caused by disturbances with inconsistent timing characteristics, reducing misjudgments caused by non-fuse events such as load fluctuations and reclosing, and ensuring that the discrimination result better corresponds to the actual physical process of fuse fusion. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 This is a flowchart of a method for diagnosing high-breaking capacity, high-voltage current-limiting fuse failures; Figure 2 This is a flowchart for constructing a current signal phase difference sequence; Figure 3 This is a flowchart of the fault indicator output status analysis and phase difference sequence extraction. Figure 4 It is a graph showing the change in the phase difference sequence of the current signal; Figure 5 This is a schematic diagram of the unfolded phase sequence and transition events. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] See Figure 1 This invention provides a method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse. The method includes the following steps: performing a preparatory operation for fuse fault diagnosis, acquiring high-voltage and low-voltage current signals, and recording the output status of a fault indicator; constructing a current signal phase difference sequence based on the high-voltage and low-voltage current signals, and determining the truncation interval of the phase difference sequence according to the fault indicator output status; performing adaptive sliding truncation on the current signal phase difference sequence within the truncation interval to generate multiple sets of phase difference sub-sequences; calculating the phase signal distortion rate of each set of phase difference sub-sequences, and determining the fuse failure discrimination factor based on the phase signal distortion rate; acquiring the falling edge time of the high-voltage and low-voltage voltages of the fuse, and calculating the falling edge time difference between the two times; correcting the fuse failure discrimination factor based on the falling edge time difference to obtain a corrected fuse failure discrimination factor; comparing the corrected fuse failure discrimination factor with a fault discrimination threshold to generate a fuse failure discrimination result.

[0021] Example 1: In specific implementation, please refer to Figure 2 Zero-crossing detection is performed on the high-voltage current signal to extract the zero-crossing moment. After analog-to-digital conversion, the high-voltage current signal forms a discrete sequence of high-voltage current sample values, containing time and amplitude information for each sample point. Iterating through temporally adjacent sample points in the high-voltage current sample value sequence, it is determined whether the product of the amplitudes of the preceding and following sample points is less than zero. If the product is less than zero, a sign change interval is defined between these two sample points. Within the sign change interval, a linear interpolation method is used to calculate the zero-crossing moment of the high-voltage current. Linear interpolation uses the time of the preceding sample point, the current amplitude of the preceding sample point, the time of the following sample point, and the current amplitude of the following sample point within the sign change interval to determine the moment when the current amplitude is zero. The zero-crossing moments of the high-voltage current extracted from all sign change intervals in chronological order constitute a high-voltage current zero-crossing moment sequence.

[0022] Zero-crossing detection is performed on the low-voltage side current signal to extract the zero-crossing time. After analog-to-digital conversion, the low-voltage side current signal forms a discrete sequence of sampled values, containing time and amplitude information for each sampling point. Following the same method as zero-crossing detection for the high-voltage side current signal, the amplitude products of adjacent sampling points in the low-voltage side current sampled value sequence are traversed to determine the sign change interval. Within each sign change interval, the zero-crossing time of the low-voltage side current is calculated using linear interpolation. All zero-crossing times of the low-voltage side current are arranged chronologically to form a low-voltage side zero-crossing time sequence.

[0023] Based on the difference between the zero-crossing times of the high-voltage end current and the low-voltage end current within the same power frequency cycle, a current signal phase difference sequence is constructed. The power frequency cycle is determined as follows: two consecutive rising zero-crossing points are selected from the high-voltage end zero-crossing time sequence, and the time interval between these two rising zero-crossing points is taken as the cycle length of one power frequency cycle. For each power frequency cycle, the zero-crossing time of the high-voltage end current located at the beginning of that power frequency cycle is extracted from the high-voltage end zero-crossing time sequence and denoted as... Where the subscript k represents the k-th power frequency cycle; extract the data from the zero-crossing time sequence of the low-voltage side. The zero-crossing moment of the low-voltage end current at the beginning of the same power frequency cycle is denoted as . Calculate the time difference between the zero-crossing moments of the high-voltage end current and the low-voltage end current within the k-th power frequency cycle, and then use the corresponding cycle length of that power frequency cycle. Convert the time difference to a phase difference. The formula for calculating the phase difference is: In the formula, This represents the k-th phase difference value in the current signal phase difference sequence. The moment when the high-voltage end current crosses zero during the k-th power frequency cycle. The moment when the low-voltage end current crosses zero during the k-th power frequency cycle. Let be the duration of the kth power frequency cycle. The value is the difference in time between the rising zero points of two adjacent high-voltage end currents within the k-th power frequency cycle. This is calculated for all power frequency cycles. Arrange the current signals in chronological order according to the power frequency cycle to obtain the current signal phase difference sequence.

[0024] See Figure 4 In the figure, the horizontal axis represents the power frequency cycle number, and the vertical axis represents the calculated phase difference of the current signal within the corresponding power frequency cycle, in radians (rad). The curve is labeled "Current Signal Phase Difference Sequence," reflecting the time difference between the high-voltage and low-voltage currents at the zero-crossing point in each power frequency cycle, and the trend of phase difference changes obtained by converting the cycle duration.

[0025] As can be seen from the figure, within the range of power frequency cycle number 0 to approximately 600, the phase difference value remains within a small range, with minimal and stable fluctuations, and the value is basically maintained at around 0.02 rad. This indicates that the phase relationship between the high-voltage and low-voltage current signals at the zero-crossing moment is in a steady-state normal operating condition during this stage, and no abnormal phase difference shift occurs.

[0026] Starting from approximately 600 cycles in the power frequency cycle, the phase difference sequence shows a clear upward trend, with the curve fluctuation amplitude increasing significantly. The value gradually climbs from near zero to around 0.6 rad, and there are many abrupt jumps. This change reflects the abnormal change in current phase caused by the fuse blowing process. The fuse blows, causing a sudden change in the phase relationship between the high and low voltage sides, which in turn manifests as local distortion and large fluctuations in the phase difference sequence.

[0027] Example 2: In specific implementation, please refer to Figure 3 The fault indicator output status is generated and provided by the fault indicator device equipped with the fuse. The fault indicator device continuously monitors changes in the electrical parameters of the circuit containing the fuse. When a sudden change in electrical quantity matching the fault characteristics is detected, the fault indicator device switches the fault indicator output status from one initial logic level to another. When the fault condition ends, the fault indicator output status returns to the initial logic level. The fault indicator output status is read as a digital signal input, and its logic value changes over time to form a status sequence.

[0028] The implementation of the fault start flag and fault end flag in the fault indicator output status is as follows: The logic value of the fault indicator output status is scanned in real time. When a transition from the initial logic level to another logic level is detected between two adjacent sampling times, the moment of the transition is identified as the fault start flag. The time point corresponding to the fault start flag is extracted from the sampling timestamp of the fault indicator output status. The sampling timestamp is determined by the product of the sampling clock count and the sampling period. The sampling period is determined by the sampling frequency of the analog-to-digital converter, and the sampling frequency is set to an integer multiple of the power grid frequency. After identifying the fault start flag, the logic value of the fault indicator output status is continuously monitored. When a transition from another logic level back to the initial logic level is detected, the moment of the transition is identified as the fault end flag. The time point corresponding to the fault end flag is also extracted from the sampling timestamp of the fault indicator output status.

[0029] In some implementations, the direction of the logic level transition in the fault indicator output state is related to the internal design of the fault indicator device. The identification of the fault initiation flag does not depend on the transition direction, but rather on the event of the logic level change relative to the steady-state state. The fault indicator device outputs a continuous initial logic level during normal operation and a continuous other logic level during a fault. Therefore, the moment when the initial logic level is first detected as the fault initiation flag, and the moment when the initial logic level is first detected as the fault termination flag, is designated as the fault termination flag.

[0030] The interception interval for the phase difference sequence is defined by taking the time point corresponding to the fault start flag as the starting point and the time point corresponding to the fault end flag as the ending point. The left boundary of the interception interval is the index position corresponding to the time point corresponding to the fault start flag, and the right boundary is the index position corresponding to the time point corresponding to the fault end flag. The current signal phase difference sequence is a time sequence arranged in order of the power frequency cycle, and each phase difference value element in the sequence has a time label that corresponds one-to-one with the power frequency cycle. When determining the interception interval, the time label of the current signal phase difference sequence is compared with the time of the interception start point, and the first phase difference value element with a time label greater than or equal to the time of the interception start point is taken as the starting element of the interception interval; the time label of the current signal phase difference sequence is compared with the time of the interception end point, and the last phase difference value element with a time label less than or equal to the time of the interception end point is taken as the ending element of the interception interval. All phase difference elements contained between the starting element and the ending element, together with the starting element and the ending element, constitute the set of elements within the truncation interval of the phase difference sequence.

[0031] It is understandable that when the fault indicator output status does not detect a complete pairing of the fault start flag and fault end flag, the truncation interval setting operation will not be performed. If only the fault start flag appears within the preset maximum analysis time but not the fault end flag, the truncation termination point will be set to the time corresponding to the maximum analysis time. The maximum analysis time is an integer multiple of the power grid frequency cycle, and this multiple is determined according to the high voltage level of the system where the fuse is located.

[0032] Example 3: In practical implementation, the initial window width and sliding step size of the sliding window are set. The initial window width is determined based on the length of the intercept interval, which is the number of elements in the current signal phase difference sequence within the intercept interval. The number of elements in the intercept interval is denoted as... Initial window width according to Calculation, where This is the initial window width scaling factor. The value range is from 0.1 to 0.3. The value is determined based on: when When the value is less than 0.1, the phase difference subsequence extracted by the sliding window contains too few elements, making it difficult to form a statistically effective characterization of phase difference fluctuations within the phase difference subsequence; when When the value is greater than 0.3, the sliding window covers an excessively wide interval, which will overly smooth the phase difference change characteristics of different stages within the intercepted interval, masking the local phase abrupt change information caused by the fuse-breaking process. Sliding step size according to set up, This is the sliding step size proportionality coefficient. The value is set to 0.5. The sliding step size is set to half the initial window width, so that there is a half-overlapping area between adjacent sliding windows.

[0033] Within the intercept interval, a sliding window moves with a sliding step size. After each move, elements from the current signal phase difference sequence that fall within the window are intercepted at the current window position, forming a set of phase difference subsequences. The sliding window movement begins with the starting element of the intercept interval, and the left boundary of the sliding window is aligned with the first element of the intercept interval. At each window position, all current signal phase difference sequence elements within the coverage area of ​​the sliding window are sequentially extracted according to their original time order, forming a set of phase difference subsequences. After interception at a window position, the sliding window is translated by a sliding step size element along the time-increasing direction of the current signal phase difference sequence. After reaching the new position, the sliding window continues to intercept the next set of phase difference subsequences. The translation and interception operations are repeated until the right boundary of the sliding window exceeds the position of the ending element of the intercept interval.

[0034] Calculate the sequence variance for each phase difference subsequence. For any phase difference subsequence, first calculate the arithmetic mean of all elements in the phase difference subsequence, and denot the arithmetic mean as... The elements of the phase difference subsequence are denoted in order as follows: ,in This represents the total number of elements contained in the phase difference subsequence. Sequence variance. The calculation method is as follows: in, The sequence variance of the phase difference subsequence. This represents the total number of elements in the phase difference subsequence. Indicates the index of the phase difference subsequence. elements, This represents the arithmetic mean of all elements in the phase difference subsequence. Sequence variance. The degree of dispersion between elements within the phase difference subsequence was quantified.

[0035] The window width for the next round of truncation is adjusted based on the sequence variance. An upper and lower variance threshold are set. The upper variance threshold is set to three times the phase difference variance of the current signal phase difference sequence during non-fault steady-state operation. The phase difference variance during non-fault steady-state operation is obtained by calculating the variance of the historical phase difference sequence recorded under normal operating conditions for the circuit containing the fuse. The lower variance threshold is set to 0.5 times the phase difference variance during non-fault steady-state operation.

[0036] In some implementations, when the calculated sequence variance When the variance exceeds the upper limit threshold, it indicates that the phase difference fluctuations within the current phase difference subsequence are severe, requiring a smaller window to focus on local distortion details. In this case, the window width is reduced by multiplying the current window width by a first adjustment factor (0.8), and the product is used as the window width for the next round of truncation. When the sequence variance... When the variance is less than the lower limit threshold, it indicates that the phase difference changes within the current phase difference subsequence are gradual. The window can be expanded to capture a wider phase evolution trend. In this case, the window width is increased by multiplying the current window width by a second adjustment factor (1.25). The product is used as the window width for the next round of truncation. When the sequence variance... When the variance falls between the lower and upper bounds, the window width for the next round of truncation remains consistent with the current window width. The adjusted window width is constrained by the minimum and maximum window widths: the minimum window width is set to 0.5 times the initial window width, and the maximum window width is set to twice the initial window width.

[0037] In the next round of cropping, the cropping operation will be performed using the adjusted window width, and the sliding step size will be updated synchronously to the adjusted window width multiplied by [the adjusted window width]. , Keeping the value at 0.5, continue generating phase difference subsequences within the truncated interval.

[0038] Example 4: In practice, a phase expansion operation is performed on each set of phase difference subsequences to obtain an expanded phase sequence. The phase expansion operation processes each element in the phase difference subsequence sequentially, eliminating the limitations imposed by the phase value. The phase entanglement phenomenon occurs within the interval. The phase unrolling operation is performed as follows: starting from the second element of each phase difference subsequence, the difference between the current element and the previous element is calculated sequentially. The elements in the phase difference subsequence are denoted in order as follows: ,in This represents the total number of elements contained in the phase difference subsequence. For an index of... The current element ,in Increasing from 2 to Calculate the difference When the difference Greater than When a positive phase transition causes entanglement, the current element is considered. and all subsequent elements after the current element in the phase difference subsequence minus When the difference Less than When a negative phase transition causes entanglement, the current element is considered. and all subsequent elements after the current element in the phase difference subsequence plus When the difference The absolute value is less than or equal to When adding or subtracting elements, no addition or subtraction operations are performed on the current element or subsequent elements. (Based on index) After traversing all elements in ascending order, the values ​​of each element in the phase difference subsequence are adjusted to continuously changing expanded phase values, and the resulting sequence is the expanded phase sequence.

[0039] Calculate the phase transition value between adjacent elements in the expanded phase sequence, and count the number of transitions where the phase transition value is greater than a threshold. For adjacent elements in the expanded phase sequence, the phase transition value is defined as the absolute value of the difference between the values ​​of two adjacent elements. Compare the phase transition values ​​of adjacent elements in the expanded phase sequence one by one with the threshold. For each case where a phase transition value is greater than the threshold, increment the transition count by one. The threshold value is set to [value to be filled in]. 1.5 times. The jump threshold is set to... The basis for the 1.5 times figure is that, under the normal conducting state of the fuse, the change in phase difference between the current signals at the high-voltage and low-voltage ends during the power frequency cycle is much smaller than that between the high-voltage and low-voltage ends. When a fuse blows, the current path is broken, causing a sharp change in the voltage-current phase relationship. The phase difference between adjacent power frequency cycles approaches or exceeds the phase difference between them. The jump. Set the jump threshold to... This system can effectively distinguish between minor phase drifts caused by normal fluctuations in grid frequency and phase abrupt changes caused by fuse failures, thus avoiding misjudging normal phase fluctuations as jump events.

[0040] The ratio of the number of transitions to the total number of elements in the expanded phase sequence is used as the phase signal distortion rate. The statistically obtained number of transitions is denoted as... The total number of elements in the expanded phase sequence is denoted as The phase signal distortion rate is calculated as follows: in, Indicates the phase signal distortion rate. This represents the number of phase transitions in the expanded phase sequence where the phase transition value is greater than the transition threshold. This represents the total number of elements in the expanded phase sequence. The denominator is taken as... This is because adjacent elements in the expanded phase sequence co-generate Each phase transition value, the number of transitions is counted based on Next-nearest comparison, phase signal distortion rate This represents the proportion of abrupt transitions that occur among all adjacent phase transitions. The value ranges from 0 to 1.

[0041] In some implementations, when the total number of elements of the expanded phase sequence When the value is 1, the phase jump value between adjacent elements cannot be calculated, and the phase signal distortion rate is directly taken as 0.

[0042] The phase signal distortion rates corresponding to all groups of phase difference subsequences are arranged in the truncation order to form a distortion rate sequence. The truncation order refers to the order in which each phase difference subsequence is generated during the adaptive sliding truncation process, and the truncation order corresponds to the process of the sliding window moving along the direction of time increase of the current signal phase difference sequence. The phase signal distortion rates calculated for each group of phase difference subsequences are then placed into a one-dimensional array in the truncation order to form a distortion rate sequence.

[0043] The median of the distortion rate sequence is used as the fault detection factor for fuse failure. The median of the distortion rate sequence is determined by sorting and taking the median: all elements in the distortion rate sequence are arranged in ascending order; when the number of elements in the distortion rate sequence is odd, the value in the middle position after sorting is taken as the median; when the number of elements in the distortion rate sequence is even, the average of the values ​​in the two middle positions after sorting is taken as the median. This median is set as the fault detection factor for fuse failure.

[0044] See Figure 5 In the graph, the horizontal axis represents the element index of the unfolded phase sequence, and the vertical axis represents the unfolded phase value, in radians (rad). The blue solid line curve represents the unfolded phase sequence, and the curve shows a gradual upward trend, indicating that the phase unfolds and changes continuously over time. There are several obvious abrupt change points on the curve, corresponding to the orange dots marked in the graph, representing phase jump events that occur between adjacent elements in the unfolded phase sequence. The red dashed line represents the jump threshold, which is set to ±4.71 radians (i.e., 1.5 times π). This threshold is used to determine whether adjacent phase jumps are abrupt jump events.

[0045] The expanded phase sequence shows significant upward or downward transitions at indexes approximately 130, 240, and 390, respectively. These transitions exceed the transition threshold, triggering transition event markers. The occurrence of these transition events reflects a sharp change in the phase of the fuse's current signal within the corresponding time period, consistent with the definition of phase transitions and their association with fuse failures described in Example 4. The relatively small number of transition events in the overall expanded phase sequence, with their clear intervals, indicates that the phase transitions exhibit a certain degree of staged and sudden occurrence.

[0046] Example 5: In practice, falling edge detection is performed on the high-voltage side voltage signal. The moment when the amplitude of the high-voltage side voltage signal drops below the falling edge determination ratio of the high-voltage side rated voltage is recorded as the high-voltage side voltage falling edge moment. The high-voltage side voltage signal is converted from analog to digital to form a discrete high-voltage side voltage sampling value sequence. Each sampling point in the high-voltage side voltage sampling value sequence includes the sampling time and voltage amplitude. The high-voltage side rated voltage is the nominal effective voltage value of the high-voltage side circuit where the fuse is located, and the high-voltage side rated voltage is obtained from the rated parameters of the power system where the fuse is located. The falling edge determination ratio is set to 0.9 times the high-voltage side rated voltage. The basis for setting the falling edge determination ratio to 0.9 times the high-voltage side rated voltage is as follows: when the fuse is normally conducting, the high-voltage side voltage amplitude is maintained near the high-voltage side rated voltage, and the amplitude fluctuation range does not exceed ±5% of the rated voltage; when the fuse blows, the high-voltage side voltage begins to drop from the rated voltage level, and the moment when the voltage amplitude drops below 0.9 times the rated voltage clearly marks the beginning of the voltage drop process, which can be distinguished from the small voltage fluctuations during normal operation.

[0047] The specific process for detecting the falling edge of the high-voltage side voltage signal is as follows: Starting from the first sampling point of the high-voltage side voltage sampling value sequence, the voltage amplitude of each sampling point is traversed in chronological order, continuously determining whether the voltage amplitude of the current sampling point is less than the falling edge determination ratio. When the first sampling point with a voltage amplitude less than the falling edge determination ratio is detected, the sampling time corresponding to that sampling point is recorded, and this sampling time is set as the falling edge time of the high-voltage side voltage. Before traversing the high-voltage side voltage sampling value sequence, a steady-state judgment is performed on the initial few sampling points of the high-voltage side voltage sampling value sequence to confirm that the voltage amplitude of the initial sampling points is greater than the falling edge determination ratio.

[0048] Falling edge detection is performed on the low-voltage side voltage signal. The moment when the amplitude of the low-voltage side voltage signal drops to below the falling edge determination ratio of the low-voltage side rated voltage is recorded as the falling edge moment of the low-voltage side voltage. The low-voltage side voltage signal is converted into a discrete sequence of low-voltage side voltage sample values ​​through analog-to-digital conversion. Each sampling point in the low-voltage side voltage sample value sequence includes the sampling time and voltage amplitude. The low-voltage side rated voltage is the nominal effective voltage value of the low-voltage side circuit where the fuse is located, and the low-voltage side rated voltage is obtained from the rated parameters of the power system where the fuse is located. The falling edge determination ratio of the low-voltage side voltage signal is set to 0.9 times the low-voltage side rated voltage, and the setting basis is the same as that of the high-voltage side voltage signal. The detection process of the falling edge moment of the low-voltage side voltage is the same as that of the high-voltage side voltage. The low-voltage side voltage sample value sequence is traversed, and the sampling time corresponding to the first sampling point where the voltage amplitude is less than 0.9 times the low-voltage side rated voltage is recorded as the falling edge moment of the low-voltage side voltage.

[0049] Calculate the absolute value of the difference between the falling edge time of the high-voltage side voltage and the falling edge time of the low-voltage side voltage, and set this absolute value as the falling edge time difference. Denote the falling edge time of the high-voltage side voltage as... The falling edge of the low-voltage side voltage is denoted as The falling edge time difference is calculated as follows: in, Indicates the time difference of the falling edge. Indicates the time of the falling edge of the high-voltage terminal voltage. This indicates the falling edge time of the low-voltage side voltage; the two vertical lines represent the absolute value operation. Falling edge time difference. This reflects the degree of asynchronous timing between the voltage drop at the high-voltage end and the voltage drop at the low-voltage end. The greater the time difference, the more significant the timing misalignment of the voltage response between the high-voltage end and the low-voltage end during the fuse-breaking process.

[0050] The correction factor is determined based on the fall-edge time difference, and it is negatively correlated with the fall-edge time difference. The correction factor ranges from 0.3 to 1. The relationship between the correction factor and the fall-edge time difference is determined using a piecewise linear function: a time difference reference value is set, which is one-quarter of the power system's operating frequency period. The operating frequency period is determined based on the grid's rated frequency; when the grid's rated frequency is 50Hz, the operating frequency period is 20 milliseconds, and the time difference reference value is 5 milliseconds. When the fall-edge time difference is less than or equal to the time difference reference value, the correction factor is 1. When the fall-edge time difference is greater than the time difference reference value but less than four times the time difference reference value, the correction factor linearly decreases from 1 to 0.3. When the fall-edge time difference is greater than or equal to four times the time difference reference value, the correction factor is 0.3. The basis for setting the lower limit of the correction factor to 0.3 is that even if the voltage drop time of the high-voltage end and the low-voltage end is greatly misaligned, a certain proportion of the fuse failure discrimination factor contribution still needs to be retained to avoid the fuse failure discrimination factor becoming completely ineffective due to the correction factor being too small.

[0051] The corrected fuse fault discrimination factor is obtained by multiplying the fuse fault discrimination factor by the correction coefficient. The fuse fault discrimination factor is the median value determined based on the phase signal distortion rate, and the correction coefficient is determined based on the falling edge time difference according to the above segmentation rule. The product of the two is the corrected fuse fault discrimination factor.

[0052] In some implementations, the negative correlation between the correction factor and the fall-edge time difference can also be achieved through a lookup table. The correction factor value corresponding to the fall-edge time difference increasing in fixed steps from zero is pre-calculated, generating a correction factor lookup table. The fixed step size is taken as the sampling period, which is the sampling interval of the analog-to-digital converter. When actually calculating the corrected fuse fault discrimination factor, the correction factor corresponding to the closest fall-edge time difference value is found in the correction factor lookup table based on the calculated fall-edge time difference. Then, the fuse fault discrimination factor is multiplied by the correction factor obtained from the lookup table.

[0053] The difference between the corrected fuse fault discrimination factor and the fault discrimination threshold is used to obtain the discrimination difference value. The fault discrimination threshold is set to 0.5. The basis for setting the fault discrimination threshold to 0.5 is as follows: In the phase difference subsequence, if more than half of the adjacent phase jumps are judged as abrupt jumps, it indicates that there is a continuous and significant distortion of the phase difference within the fault interval, and the probability that the fuse has blown is high; if the proportion of abrupt jumps is much less than half, it indicates that the degree of phase difference distortion is weak, and the probability that the fuse has not blown is high.

[0054] The discrimination difference is calculated by subtracting the fault discrimination threshold from the corrected fuse fault discrimination factor. When the discrimination difference is greater than the positive discrimination threshold, a discrimination result indicating that the fuse has blown is generated. The positive discrimination threshold is 0.1. When the discrimination difference is less than the negative discrimination threshold, a discrimination result indicating that the fuse has not blown is generated. The negative discrimination threshold is -0.1. When the discrimination difference is between the negative and positive discrimination thresholds, a discrimination result indicating that the fuse status is pending confirmation is generated. The positive and negative judgment thresholds are set based on the following: a confirmation interval with a width of 0.2 is set near the fault judgment threshold of 0.5. When the corrected fuse fault judgment factor falls between 0.4 and 0.6, it indicates that the fuse fault judgment factor is close to the fault judgment threshold. At this time, the judgment criteria derived from the phase difference distortion characteristics and voltage drop edge synchronization are insufficient to give a definitive conclusion. It is necessary to generate a judgment result with pending confirmation to prompt further manual review or supplementary testing.

[0055] The above description is merely a specific embodiment of this application, but the scope of protection of this application 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 this application should be included within the scope of protection of this application.

Claims

1. A method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse, characterized in that, The method includes: Perform preparatory operations for fuse fault diagnosis, including acquiring high-voltage end current signals and low-voltage end current signals, and recording the output status of the fault indicator; Based on the high-voltage end current signal and the low-voltage end current signal, a current signal phase difference sequence is constructed, and the truncation interval of the phase difference sequence is determined according to the output state of the fault indicator. Within the interception interval, an adaptive sliding interception is performed on the current signal phase difference sequence to generate multiple sets of phase difference sub-sequences; Calculate the phase signal distortion rate of each phase difference subsequence, and determine the fuse failure discrimination factor based on the phase signal distortion rate; Obtain the falling edge times of the high-voltage and low-voltage terminals of the fuse, and calculate the time difference between the falling edge times of the high-voltage and low-voltage terminals. The fuse failure discrimination factor is corrected based on the falling edge time difference to obtain the corrected fuse failure discrimination factor; The modified fuse fault discrimination factor is compared with the fault discrimination threshold to generate a fuse fault discrimination result.

2. The method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse according to claim 1, characterized in that, The step of constructing a current signal phase difference sequence based on the high-voltage end current signal and the low-voltage end current signal includes: Zero-crossing detection is performed on the high-voltage end current signal to extract the zero-crossing time of the high-voltage end current, and zero-crossing detection is also performed on the low-voltage end current signal to extract the zero-crossing time of the low-voltage end current. Based on the difference between the zero-crossing times of the high-voltage end current and the low-voltage end current within the same power frequency cycle, a current signal phase difference sequence is constructed.

3. The method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse according to claim 1, characterized in that, Determining the truncation interval of the phase difference sequence based on the output state of the fault indicator includes: Analyze the fault start flag and fault end flag in the output status of the fault indicator; Using the time point corresponding to the fault start flag as the starting point and the time point corresponding to the fault end flag as the ending point, the interval between the starting point and the ending point is set as the truncation interval of the phase difference sequence.

4. The method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse according to claim 1, characterized in that, The adaptive sliding truncation of the current signal phase difference sequence within the truncation interval generates multiple sets of phase difference sub-sequences, including: Set the initial window width and sliding step size of the sliding window; Within the interception interval, the sliding window is moved with the sliding step size. After each movement, the elements in the current signal phase difference sequence that fall into the window are intercepted at the current window position to form a set of phase difference subsequences. Calculate the sequence variance of each phase difference subsequence, and adjust the window width for the next round of truncation based on the sequence variance.

5. The method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse according to claim 4, characterized in that, The calculation of the phase signal distortion rate of each group of phase difference subsequences, and the determination of the fuse failure discrimination factor based on the phase signal distortion rate, includes: Perform a phase unrolling operation on each set of phase difference subsequences to obtain the unrolled phase sequence; Calculate the phase transition value between adjacent elements in the expanded phase sequence, and count the number of transitions where the phase transition value is greater than the transition threshold; The ratio of the number of transitions to the total number of elements in the expanded phase sequence is used as the phase signal distortion rate; Arrange the phase signal distortion rates corresponding to all the phase difference subsequences in the order of truncation to form a distortion rate sequence, and use the median of the distortion rate sequence as the fuse failure discrimination factor.

6. The method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse according to claim 1, characterized in that, The step of obtaining the falling edge times of the high-voltage and low-voltage terminals of the fuse, and calculating the falling edge time difference between the high-voltage and low-voltage terminals, includes: Falling edge detection is performed on the high-voltage terminal voltage signal, and the moment when the amplitude of the high-voltage terminal voltage signal drops to below the falling edge determination ratio of the high-voltage terminal rated voltage is recorded as the high-voltage terminal voltage falling edge moment; Falling edge detection is performed on the low-voltage side voltage signal, and the moment when the amplitude of the low-voltage side voltage signal drops to below the falling edge determination ratio of the low-voltage side rated voltage is recorded as the falling edge moment of the low-voltage side voltage; Calculate the absolute value of the difference between the falling edge time of the high-voltage side voltage and the falling edge time of the low-voltage side voltage, and set the absolute value of the difference as the falling edge time difference.

7. The method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse according to claim 1, characterized in that, The step of correcting the fuse failure discrimination factor based on the falling edge time difference to obtain the corrected fuse failure discrimination factor includes: The value of the correction coefficient is determined based on the falling edge time difference, and the correction coefficient is negatively correlated with the falling edge time difference; The corrected fuse fault discrimination factor is obtained by multiplying the fuse fault discrimination factor by the correction coefficient.

8. The method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse according to claim 1, characterized in that, The step of comparing the modified fuse fault discrimination factor with the fault discrimination threshold to generate a fuse fault discrimination result includes: The difference between the modified fuse fault discrimination factor and the fault discrimination threshold is used to obtain the discrimination difference value; When the discrimination difference is greater than the positive judgment threshold, a judgment result indicating that the fuse has blown is generated; When the discrimination difference is less than the negative discrimination threshold, a discrimination result indicating that the fuse has not blown is generated; When the discrimination difference is between the negative discrimination threshold and the positive discrimination threshold, a discrimination result for the fuse status to be confirmed is generated.

9. The method for diagnosing the failure of a high-breaking capacity high-voltage current-limiting fuse according to claim 4, wherein adjusting the window width of the next round of interception according to the sequence variance includes reducing the window width when the sequence variance is greater than the upper limit threshold of variance, and increasing the window width when the sequence variance is less than the lower limit threshold of variance.

10. A method for diagnosing the failure of a high-breaking capacity, high-voltage current-limiting fuse according to claim 5, characterized in that, The step of performing a phase expansion operation on each group of phase difference subsequences to obtain an expanded phase sequence includes calculating the difference between the current element and the previous element sequentially, starting from the second element of each group of phase difference subsequences. When the difference is greater than π, 2π is subtracted from the current element and all subsequent elements. When the difference is less than -π, 2π is added to the current element and all subsequent elements.

Images