Intelligent leakage current detection and alarm method for distribution boxes

Abstract

This invention relates to the field of leakage current detection and discloses a smart leakage current detection and alarm method for distribution boxes. The method includes: continuously acquiring residual current monitoring data to form time series parameters; calculating instantaneous fluctuation characteristics in real time, and confirming the leakage fluctuation trigger moment after interference elimination verification when the value exceeds a preset threshold; triggering a reverse tracing process to extract historical monitoring parameters and load current waveforms; performing reverse differential operations backward from the trigger moment to dynamically eliminate components synchronized with load current changes and determine the leakage start moment; outputting tracing information and determining whether to generate an alarm signal based on the leakage duration, and monitoring reset conditions to automatically reset or maintain the alarm. This invention, through multi-level interference elimination and reverse differential tracing, can accurately identify the leakage start moment, effectively distinguish between real leakage and interference such as load switching and instantaneous pulses, improving the reliability and intelligence level of leakage current detection.

Description

Technical Field

[0001] This invention relates to the field of leakage current detection, and more specifically to a smart leakage current detection and alarm method for distribution boxes. Background Technology

[0002] As a critical node in low-voltage power distribution systems, leakage current monitoring of the outgoing circuits of distribution boxes is an important aspect of ensuring electrical safety. Traditional leakage protection devices mainly rely on residual current operated devices (RCDs), which typically have a fixed operating threshold. When the residual current exceeds the threshold, the device trips.

[0003] However, in actual operation, there are a large number of residual current fluctuations caused by factors other than leakage current, such as normal load switching, capacitor compensation device activation, electrostatic discharge, and harmonic currents generated by nonlinear loads. These disturbances can easily cause traditional protection devices to malfunction or trip frequently, affecting the reliability of normal power supply.

[0004] Meanwhile, traditional methods struggle to accurately capture the true onset of leakage faults, making it impossible to trace the leakage development process and hindering fault location and cause investigation. Although some leakage detection solutions based on waveform analysis have emerged in recent years, most lack systematic interference elimination mechanisms and reverse tracing capabilities, and their detection accuracy in complex electrical environments still needs improvement.

[0005] Therefore, there is an urgent need for a leakage current detection and alarm method that can intelligently identify real leakage events, accurately locate the leakage initiation time, and has strong anti-interference capabilities. Summary of the Invention

[0006] The purpose of this invention is to provide a smart leakage current detection and alarm method for distribution boxes, thereby solving the above-mentioned technical problems.

[0007] The objective of this invention can be achieved through the following technical solutions: The intelligent leakage current detection and alarm method for distribution boxes includes the following steps: Step S1: Continuously acquire residual current monitoring data of the outgoing circuit of the distribution box to form time series monitoring parameters; Step S2: Calculate the instantaneous fluctuation characteristic of the monitoring parameter in real time. When the instantaneous fluctuation characteristic exceeds the preset fluctuation trigger threshold, mark the current time as a candidate fluctuation time, verify the interference elimination of the candidate fluctuation time, and confirm the candidate fluctuation time that passes the verification as the leakage current fluctuation trigger time Tw. Step S3: Trigger the reverse tracing process, extract the historical monitoring parameter sequence within the preset time window before Tw, and synchronously store the load current waveform within that time window; Step S4: Starting from Tw, perform reverse differential operation backwards to dynamically eliminate components that are synchronized with the load current change and determine the leakage current initiation time ta; Step S5: Output traceability information including ta and Tw, and determine whether to generate a leakage alarm signal based on the duration of leakage. After generating a leakage alarm signal, monitor whether the preset reset conditions are met to automatically reset or maintain the alarm state.

[0008] Preferably, the instantaneous fluctuation characteristic includes: the instantaneous difference between the current sampling point and the corresponding point of the previous cycle, the slope change rate of multiple consecutive sampling points, and the symmetry coefficient of the waveform before and after the fluctuation; The fluctuation trigger threshold includes an instantaneous difference threshold and a slope change rate threshold. When the instantaneous difference exceeds the instantaneous difference threshold and the slope change rate exceeds the slope change rate threshold, it is determined that the fluctuation trigger threshold has been exceeded.

[0009] Preferably, the interference elimination verification includes: load current comparison verification, symmetry verification, and multi-cycle consistency verification; The load current comparison verification is used to eliminate interference caused by load switching, the symmetry verification is used to eliminate instantaneous pulse interference, and the multi-cycle consistency verification is used to eliminate periodic load interference.

[0010] Preferably, the load current comparison verification includes: extracting the load current waveforms before and after the candidate fluctuation time, performing correlation analysis with the load current waveforms of the corresponding segment of the previous power frequency cycle at that time, retaining them if the correlation coefficient is higher than the first threshold, and discarding them if the correlation coefficient is lower than the second threshold as load switching interference. The symmetry verification includes: calculating the symmetry coefficients of the rising and falling edges of the wave waveform at the candidate wave moment; if the symmetry coefficients are lower than the first symmetry threshold, it is determined to be capacitor switching or electrostatic discharge interference and discarded; if the symmetry coefficients are higher than the second symmetry threshold, it is retained. The multi-cycle consistency verification includes: observing the positions one and two power frequency cycles backward from the candidate fluctuation moment. If there are fluctuations with waveform similarity higher than the similarity threshold in both cases, they are determined to be periodic load interference and discarded; otherwise, they are confirmed as real leakage current triggering events.

[0011] Preferably, step S4 specifically includes: Starting from Tw, calculate the reverse difference point by point. When the ratio of the rate of change of the load current at the corresponding moment to the rate of change of the residual current is within the preset synchronization range, set the reverse difference to zero to eliminate the component that is synchronized with the change of the load current. Mark the starting point when the absolute value of the inverse difference after all interference is removed first continuously exceeds the steady-state noise threshold, and determine the timestamp corresponding to the starting point as the candidate leakage initiation time; Perform reverse verification on the candidate leakage start time. If the verification passes, the leakage start time ta is finally determined. Otherwise, continue to search for the next candidate point until the verification condition is met or the preset time window is traversed.

[0012] Preferably, the reverse verification includes: the variance of the residual current within a first preset time period preceding the candidate leakage initiation time is less than the steady-state noise threshold, and the residual current within a second preset time period following the initiation time shows a unidirectional increasing or decreasing trend.

[0013] Preferably, in step S5, the conditions for generating a leakage alarm signal are: the duration of leakage exceeds a preset alarm delay threshold, and the integral of leakage energy from ta to Tw exceeds the minimum leakage energy threshold; the reset condition is: within a subsequent preset time window, the residual current waveform recovers to a normal level that matches the load current change, and no new leakage fluctuation trigger time Tw is confirmed.

[0014] Preferably, the symmetry coefficient is calculated as follows: symmetry coefficient = |rise time - fall time| / (rise time + fall time), where the rise time is the time interval from the start point to the peak point of the wave, and the fall time is the time interval from the peak point to the end point; The ascent and descent times are determined in the following manner: Eliminate DC bias from the residual current waveform segment centered on the candidate fluctuation moment; Search for the starting point of the waveform backward from the candidate fluctuation moment, and search for the ending point of the waveform backward; Determine the peak point between the start and end points; The time difference between the peak point and the starting point is taken as the rise time, and the time difference between the ending point and the peak point is taken as the fall time. If there are multiple local extrema between the start point and the peak point of the waveform, the start point is corrected with the last extrema point; if there are multiple local extrema between the peak point and the end point, the end point is corrected with the first extrema point, and the rise time and fall time are recalculated.

[0015] Preferably, the instantaneous difference is obtained in the following way: Calculate the first difference between the current sampling point and the corresponding point of the previous cycle, and the average difference between the current sampling point and the corresponding point of the previous cycle in the neighborhood before and after it; Compare the relationship between the first difference and the average of the differences: If the first difference is greater than a set multiple of the average difference, then the first difference is taken as the instantaneous difference. If the first difference is less than or equal to the average difference, then the instantaneous difference is set to zero; If the difference falls between the two, the weighted average of the first difference and the average of the differences is taken as the instantaneous difference, where the weights are negatively correlated with the current noise level.

[0016] Preferably, the parameters in the reverse tracing process are dynamically adjusted based on the stability of the determined rise and fall times, including: If the rise time or fall time is corrected due to the existence of multiple local extrema, the length of the preset time window is increased and the search step size of the inverse difference operation is increased. If the ratio of rise time to fall time exceeds the set threshold, the preset time window length will remain unchanged, and the search step size of the inverse difference operation will be encrypted. Otherwise, the reverse tracing process will be executed with the default parameters.

[0017] The beneficial effects of this invention are: (1) By setting the comparison between instantaneous fluctuation characteristic quantity and fluctuation trigger threshold, and combining the triple interference elimination mechanism of load current comparison verification, symmetry verification and multi-cycle consistency verification, this invention can effectively distinguish between real leakage current and residual current fluctuations caused by various non-leakage factors such as load switching, capacitor switching, electrostatic discharge, and periodic load interference, thereby significantly reducing the false alarm rate and improving the reliability of leakage current detection. (2) The present invention adopts reverse differential operation from the time of leakage fluctuation triggering and dynamically removes components that are synchronized with the load current change. Combined with reverse verification conditions, the leakage start time is determined. It can accurately trace the true starting point of leakage fault, provide accurate time basis for fault analysis and location and leakage energy assessment, and overcome the defect of traditional methods that can only detect the current leakage state and cannot trace the start time. (3) The present invention determines whether to generate an alarm signal based on the duration of leakage and the integral of leakage energy. After the alarm, it monitors whether the residual current waveform returns to the normal level to determine whether to automatically reset or keep the alarm. This avoids frequent alarms caused by instantaneous leakage fluctuations and keeps the alarm in a state of alert for continuous real leakage faults, thus realizing intelligent and adaptive reset control of leakage protection. Attached Figure Description

[0018] The invention will now be further described with reference to the accompanying drawings.

[0019] Figure 1 This is a diagram illustrating the method steps of the present invention. Detailed Implementation

[0020] 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, and 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.

[0021] Please see Figure 1 As shown, this embodiment provides a smart leakage current detection and alarm method for distribution boxes. This method is applicable to leakage current monitoring and alarm of outgoing circuits of distribution boxes in low-voltage power distribution systems. The distribution box is equipped with a residual current transformer, a current transformer, and corresponding signal conditioning circuits and a microprocessor for real-time acquisition of residual current and load current signals.

[0022] Specifically, the residual current transformer is mounted on the outgoing circuit bus of the distribution box, and its secondary output is filtered, amplified and level shifted by the signal conditioning circuit before being connected to the analog-to-digital conversion interface of the microprocessor.

[0023] The current transformers are respectively installed in each branch circuit for synchronously acquiring load current waveforms. The microprocessor has an integrated or external non-volatile memory for storing time-series monitoring parameters and load current waveform data.

[0024] Step S1: Continuously acquire residual current monitoring data; The microprocessor continuously acquires residual current monitoring data from the outgoing circuits of the distribution box at a preset sampling frequency, forming a time-series sequence of monitoring parameters. In this embodiment, the sampling frequency is set to 128 points per power frequency cycle (corresponding to a sampling frequency of 6400 Hz and a power frequency of 50 Hz). These time-series monitoring parameters are stored in a circular buffer in the order of sampling time for subsequent real-time fluctuation analysis and reverse tracing. Simultaneously, the load current waveform is also synchronously acquired and stored at the same sampling frequency.

[0025] Step S2: Real-time fluctuation detection and interference elimination; The microprocessor calculates the instantaneous fluctuation characteristics of the monitored parameters in real time. In this embodiment, the instantaneous fluctuation characteristics include: the instantaneous difference between the current sampling point and the corresponding point of the previous cycle, the rate of change of the slope of three consecutive sampling points, and the symmetry coefficient of the waveform before and after the fluctuation.

[0026] Methods for obtaining instantaneous differences; First, calculate the first difference between the current sampling point and the corresponding point of the previous cycle. The corresponding point of the previous cycle refers to the sampling value one power frequency cycle (i.e., 128 sampling points) prior to the current sampling time. At the same time, calculate the average difference between the current sampling point and the two adjacent sampling points before and after the corresponding point of the previous cycle (a total of 3 sampling points).

[0027] Then compare the relationship between the first difference and the average of the differences: If the first difference is greater than 4 times the average difference, it is determined to be a true jump, and the first difference is taken as the instantaneous difference.

[0028] If the first difference is less than or equal to the average difference, it is determined to be noise or jitter, and the instantaneous difference is set to zero.

[0029] If the instantaneous difference falls between the two (i.e., greater than the average difference and less than or equal to four times the average difference), then the instantaneous difference is taken as the weighted average of the first difference and the average difference. The weights are negatively correlated with the current noise level. The current noise level is calculated by taking the standard deviation of the residual current over the most recent power frequency cycle. Assuming the normalized noise level is between 0 and 1, the weight of the first difference is equal to 1 minus the noise level, and the weight of the average difference is equal to the noise level. The higher the noise level, the smaller the weight of the first difference and the larger the weight of the average difference.

[0030] The final determined instantaneous difference value is compared with an instantaneous difference threshold. In this embodiment, the instantaneous difference threshold is set to 15 mA.

[0031] Methods for obtaining the rate of change of slope; The slope change rate is obtained by calculating the second-order difference of three consecutive sampling points. Specifically, let the current sampling point be the nth point with a current value of In, the previous point be I{n-1}, and the point before that be I{n-2}. First, the first-order difference is calculated: the first segment difference is In-I{n-1}, and the second segment difference is I{n-1}-I{n-2}. Then, the second-order difference (i.e., the slope change rate) is calculated as: (In-I{n-1})-(I{n-1}-I{n-2}), and then divided by the square of the sampling interval. In this embodiment, the sampling interval is 156.25 microseconds (corresponding to a 6400 Hz sampling rate); the slope change rate threshold is set to 0.01 mA per square microsecond.

[0032] When the instantaneous difference exceeds the instantaneous difference threshold and the slope change rate exceeds the slope change rate threshold, it is determined that the fluctuation trigger threshold has been exceeded. At this time, the current sampling time is marked as a candidate fluctuation time.

[0033] Interference elimination verification; To eliminate residual current fluctuations caused by factors other than leakage current, this implementation method performs a three-level interference elimination verification on candidate fluctuation moments. Only candidate fluctuation moments that pass the verification are confirmed as leakage current fluctuation trigger moments. The specific interference elimination verification includes the following: Load current comparison verification is used to eliminate interference caused by load switching.

[0034] The microprocessor extracts the load current waveform within 1 milliseconds before and after the candidate fluctuation moment (approximately 13 sampling points in total), and simultaneously extracts the load current waveform for the corresponding period of the previous complete power frequency cycle (i.e., the interval from 20 milliseconds to 19 milliseconds before the candidate fluctuation moment) as a reference waveform. Correlation analysis is performed on the two waveforms, and the Pearson correlation coefficient is calculated.

[0035] The correlation coefficient is calculated as follows: the sum of the products of corresponding sampling points of the two waveforms minus the product of their respective means, divided by the product of their respective standard deviations. If the correlation coefficient is higher than the first correlation coefficient threshold (0.95 in this embodiment), it indicates that the waveforms are highly consistent, and is judged as normal load fluctuation, so the candidate fluctuation time is retained; if the correlation coefficient is lower than the second correlation coefficient threshold (0.8 in this embodiment), it indicates that the waveform has changed significantly, and is judged as load switching interference, so the candidate fluctuation time is directly discarded.

[0036] Symmetry verification is used to eliminate transient pulse interference (such as capacitor switching or electrostatic discharge).

[0037] Calculate the symmetry coefficients of the rising and falling edges of the wave waveform at the candidate wave time. The formula for calculating the symmetry coefficient is: the symmetry coefficient equals the absolute value of the difference between the rise time and the fall time divided by the sum of the two.

[0038] The methods for extracting rise and fall times are as follows: Centered on the candidate fluctuation moment, 32 sampling points are extended forward and backward (a total of 64 sampling points) to extract the remaining current waveform segment, so that the segment can completely cover the entire fluctuation process.

[0039] Calculate the median current at all sampling points in this segment, and then shift the waveform as a whole to make the median current zero, thereby eliminating the influence of DC bias.

[0040] Search backward from the candidate fluctuation moment to find the first point. The current values ​​of the three consecutive sampling points after this point are all greater than twice the current noise level, and the current values ​​of these three points increase sequentially (i.e., the later point is greater than the previous point). Record this point as the waveform starting point.

[0041] Search backward from the candidate fluctuation moment and find the first point. The current values ​​of the three consecutive sampling points after this point all drop back to less than twice the current noise level, and the current values ​​of these three points decrease sequentially. Record this point as the end point of the waveform.

[0042] Within the interval between the waveform's start and end points, the sampling point corresponding to the maximum current value is identified as the peak point. The rise time is equal to the time difference between the peak point sampling time and the waveform start point sampling time; the fall time is equal to the time difference between the waveform end point sampling time and the peak point sampling time.

[0043] If the rise time or fall time is less than one sampling period (i.e., 156.25 microseconds), the duration of the fluctuation is determined to be too short, and the candidate fluctuation moment is discarded directly.

[0044] If there are multiple local extrema between the waveform's starting point and peak point, the last extrema is used to correct the waveform's starting point. A local extrema is defined as a current value greater than the current values ​​of its two adjacent sampling points. If there are multiple local extrema between the peak point and waveform's ending point, the first extrema is used to correct the waveform's ending point; the rise and fall times are then recalculated.

[0045] In this embodiment, the first symmetry threshold is set to 0.2, and the second symmetry threshold is set to 0.6. If the symmetry coefficient is lower than 0.2, it indicates that the waveform is close to symmetrical, and it is determined to be capacitor switching or electrostatic discharge interference and discarded; if the symmetry coefficient is higher than 0.6, it indicates that the waveform is asymmetrical, and the candidate fluctuation moment is retained; if it is between the two, it proceeds to the next level of verification.

[0046] Multi-cycle consistency verification is used to eliminate periodic load interference.

[0047] Observe the positions one power frequency cycle (20 milliseconds) and two power frequency cycles (40 milliseconds) backward from the candidate fluctuation moment. Extract the residual current waveform segments one millisecond before and after these two positions, and compare their similarity with the waveform segments one millisecond before and after the candidate fluctuation moment. The similarity is calculated using the waveform correlation coefficient.

[0048] If the correlation coefficients of the two delay locations are both higher than the similarity threshold (0.9 in this implementation), it indicates that the fluctuation has a periodic repetitive characteristic, and it is determined to be a periodic load interference (such as harmonic fluctuations generated by frequency converters or periodic loads), and the candidate fluctuation time is discarded. Otherwise, it is confirmed as a real leakage current trigger event, and the current time is recorded as the leakage current fluctuation trigger time.

[0049] Through the above three-level verification, the residual current fluctuations caused by most non-leakage factors can be effectively eliminated, ensuring that the starting point for subsequent reverse tracing is true and reliable.

[0050] Step S3: Trigger the reverse tracing process; Once the timing of the leakage current fluctuation is confirmed, the microprocessor immediately initiates the reverse tracing process. First, it extracts the historical monitoring parameter sequence within a preset time window preceding the leakage current fluctuation trigger time. The default length of this window is set to six times one power frequency cycle (twenty milliseconds), i.e., one hundred and twenty milliseconds. Simultaneously, the load current waveform data within this time window is stored for subsequent synchronous component removal.

[0051] Step S4: Determine the leakage initiation time using reverse differential calculation; Within the stated time window, reverse differential calculations are performed point-by-point backward from the moment the leakage current fluctuation is triggered, dynamically eliminating components synchronized with changes in load current, and ultimately determining the leakage current initiation moment. The specific process is as follows: First, starting from the moment the leakage current fluctuation is triggered, the reverse difference is calculated point by point backward, which is the residual current at the current sampling point minus the residual current at the previous sampling point. Simultaneously, the load current change rate at the corresponding moment is obtained, which is the load current at the current sampling point minus the load current at the previous sampling point, divided by the sampling interval. When the ratio of the load current change rate to the residual current change rate is within the range of 0.8 to 1.2, and both change in the same direction, it is determined that the current reverse difference is a synchronous component caused by the load current change, rather than a leakage current component. In this case, the reverse difference is set to zero and discarded.

[0052] Then, the starting point where the absolute value of the inverse difference after interference removal first continuously exceeds the steady-state noise threshold is marked, and the timestamp corresponding to the starting point is determined as the candidate leakage initiation time. The steady-state noise threshold is determined as follows: when the system is running normally and there are no leakage events, the residual current is collected at 128 consecutive points (one power frequency cycle), the standard deviation of these sampling points is calculated, and the steady-state noise threshold is set to 4 times the standard deviation. In this embodiment, the typical steady-state noise threshold is 3 mA.

[0053] To prevent false positives due to noise, a reverse verification is performed on the candidate leakage initiation time: if the variance of the residual current within 10 milliseconds preceding the candidate time is less than the square of the steady-state noise threshold (i.e., 9 square milliamperes), it indicates that the system was in a stable state before that time; and if the residual current within 10 milliseconds following the candidate time shows a unidirectional increasing or decreasing trend, the verification method is to calculate whether the signs of the differences between adjacent sampling points within these 10 milliseconds are consistent (all positive or all negative), indicating that leakage has begun to develop. If the verification passes, the leakage initiation time is finally determined; if the verification fails, the search continues to move forward to find the next candidate point that meets the conditions, until the verification conditions are met or the time window has been traversed.

[0054] As a further preferred implementation, the parameters in the reverse tracing process are dynamically adjusted based on the stability of the rise and fall times calculated in step S2 to improve tracing accuracy. Specifically, this includes the following: If the starting or ending point is corrected due to multiple local extrema in the waveform during the extraction of rise and fall times, it indicates that the leakage waveform has oscillations or distortions. In this case, the default length of the time window is increased to twice the original length (i.e., 240 milliseconds), and the search step size of the reverse differential operation is changed from point-by-point to calculation once every two sampling points (i.e., the step size becomes 312.5 microseconds) to smooth out oscillation interference and highlight the overall trend.

[0055] If no correction is made but the ratio of the larger to the smaller of the rise and fall times is greater than 4, it indicates that the leakage waveform has significant fast rise and slow fall or slow rise and fast fall characteristics. In this case, the default window length (120 milliseconds) is maintained, but the reverse search step size is encrypted to every half sampling point (i.e., 78.125 microseconds). The residual current value at the intermediate point is calculated by linear interpolation to finely locate the starting point.

[0056] If no correction is made and the above ratio is not greater than 4, reverse tracing will be performed with the default parameters (window 120 milliseconds, step size is one sampling point).

[0057] The reverse differential operation and candidate starting point search were re-executed using the adjusted window length and search step size to finally determine the leakage initiation time.

[0058] Step S5: Alarm judgment and reset control; After obtaining the leakage initiation time and leakage fluctuation trigger time, the microprocessor outputs traceability information containing these two moments. This information can be stored in local logs or uploaded to a cloud monitoring platform for fault analysis and location. Simultaneously, it determines whether to generate a leakage alarm signal based on the leakage duration.

[0059] The duration of leakage current is defined as the time difference between the trigger time of leakage current fluctuation and the start time of leakage current.

[0060] There are two conditions for generating a leakage current alarm signal, both of which must be met simultaneously: First, the duration of leakage exceeds the preset alarm delay threshold. In this embodiment, the alarm delay threshold is set to 40 milliseconds (i.e., two power frequency cycles) to avoid frequent alarms caused by transient fluctuations.

[0061] Second, the total leakage energy from the start of the leakage to the triggering of the leakage fluctuation exceeds the minimum leakage energy threshold. The leakage energy integral is calculated by integrating the squared value of the residual current within this time window. The integration method is to multiply the squared value of the current at each sampling point by the sampling interval and then sum them up. In this embodiment, the minimum leakage energy threshold is set to 1000 mA square milliseconds (equivalent to the energy integral generated by a leakage lasting 20 milliseconds with an average current of 7 mA).

[0062] When both conditions are met, the microprocessor outputs a leakage current alarm signal, which can notify maintenance personnel or automatically disconnect the circuit through an audible and visual alarm, a remote communication module, or a circuit breaker trip signal.

[0063] After an alarm is triggered, the system continuously monitors the residual current waveform to determine if the reset conditions are met. The reset conditions are: within a subsequent preset time window (100 milliseconds in this implementation, i.e., 5 power frequency cycles), the residual current waveform recovers to a normal level that matches the load current change; that is, the ratio of the residual current change to the load current change remains within the range of 0.8 to 1.2, and no new leakage current fluctuation trigger moment is confirmed within this time window. When the reset conditions are met, the system automatically resets and clears the alarm state; if the reset conditions are not met, the alarm state remains until manual intervention or power outage maintenance.

[0064] Through the above steps, this implementation method achieves accurate identification of actual leakage events in the distribution box outgoing circuit, accurate tracing of the start time, and intelligent control of alarm and reset, effectively solving the problems of traditional leakage protection devices being prone to false activation, unable to trace the source, and having a single reset method.

[0065] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural modifications made based on the content of this specification, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this invention.

[0066] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A smart leakage current detection and alarm method for distribution boxes, characterized in that, Includes the following steps: Step S1: Continuously acquire residual current monitoring data of the outgoing circuit of the distribution box to form time series monitoring parameters; Step S2: Calculate the instantaneous fluctuation characteristic of the monitoring parameter in real time. When the instantaneous fluctuation characteristic exceeds the preset fluctuation trigger threshold, mark the current time as a candidate fluctuation time, verify the interference elimination of the candidate fluctuation time, and confirm the candidate fluctuation time that passes the verification as the leakage current fluctuation trigger time Tw. Step S3: Trigger the reverse tracing process, extract the historical monitoring parameter sequence within the preset time window before Tw, and synchronously store the load current waveform within that time window; Step S4: Starting from Tw, perform reverse differential operation backwards to dynamically eliminate components that are synchronized with the load current change and determine the leakage current initiation time ta; Step S5: Output traceability information including ta and Tw, and determine whether to generate a leakage alarm signal based on the duration of leakage. After generating a leakage alarm signal, monitor whether the preset reset conditions are met to automatically reset or maintain the alarm state.

2. The intelligent leakage current detection and alarm method for a distribution box according to claim 1, characterized in that, The instantaneous fluctuation characteristics include: the instantaneous difference between the current sampling point and the corresponding point of the previous wave, the slope change rate of multiple consecutive sampling points, and the symmetry coefficient of the waveform before and after the fluctuation. The fluctuation trigger threshold includes an instantaneous difference threshold and a slope change rate threshold. When the instantaneous difference exceeds the instantaneous difference threshold and the slope change rate exceeds the slope change rate threshold, it is determined that the fluctuation trigger threshold has been exceeded.

3. The intelligent leakage current detection and alarm method for a distribution box according to claim 2, characterized in that, The interference elimination verification includes: load current comparison verification, symmetry verification, and multi-cycle consistency verification; The load current comparison verification is used to eliminate interference caused by load switching, the symmetry verification is used to eliminate instantaneous pulse interference, and the multi-cycle consistency verification is used to eliminate periodic load interference.

4. The intelligent leakage current detection and alarm method for a distribution box according to claim 3, characterized in that, The load current comparison verification includes: extracting the load current waveforms before and after the candidate fluctuation time, performing correlation analysis with the load current waveforms of the corresponding segment of the previous power frequency cycle at that time, retaining them if the correlation coefficient is higher than the first threshold, and discarding them if the correlation coefficient is lower than the second threshold, which is determined to be load switching interference. The symmetry verification includes: calculating the symmetry coefficients of the rising and falling edges of the wave waveform at the candidate wave moment; if the symmetry coefficients are lower than the first symmetry threshold, it is determined to be capacitor switching or electrostatic discharge interference and discarded; if the symmetry coefficients are higher than the second symmetry threshold, it is retained. The multi-cycle consistency verification includes: observing the positions one and two power frequency cycles backward from the candidate fluctuation moment. If there are fluctuations with waveform similarity higher than the similarity threshold in both cases, they are determined to be periodic load interference and discarded; otherwise, they are confirmed as real leakage current triggering events.

5. The intelligent leakage current detection and alarm method for a distribution box according to claim 4, characterized in that, Step S4 specifically includes: Starting from Tw, calculate the reverse difference point by point. When the ratio of the rate of change of the load current at the corresponding moment to the rate of change of the residual current is within the preset synchronization range, set the reverse difference to zero to eliminate the component that is synchronized with the change of the load current. Mark the starting point when the absolute value of the inverse difference after all interference is removed first continuously exceeds the steady-state noise threshold, and determine the timestamp corresponding to the starting point as the candidate leakage initiation time; Perform reverse verification on the candidate leakage start time. If the verification passes, the leakage start time ta is finally determined. Otherwise, continue to search for the next candidate point until the verification condition is met or the preset time window is traversed.

6. The intelligent leakage current detection and alarm method for a distribution box according to claim 5, characterized in that, The reverse verification includes: the variance of the residual current within a first preset time period preceding the candidate leakage initiation time is less than the steady-state noise threshold, and the residual current within a second preset time period following the initiation time shows a unidirectional increasing or decreasing trend.

7. The intelligent leakage current detection and alarm method for a distribution box according to claim 2, characterized in that, In step S5, the conditions for generating a leakage alarm signal are: the duration of leakage exceeds a preset alarm delay threshold, and the integral of leakage energy from ta to Tw exceeds the minimum leakage energy threshold; the reset condition is: within a subsequent preset time window, the residual current waveform recovers to a normal level that matches the load current change, and no new leakage fluctuation trigger time Tw is confirmed.

8. The intelligent leakage current detection and alarm method for a distribution box according to claim 2, characterized in that, The symmetry coefficient is calculated as follows: symmetry coefficient = |rise time - fall time| / (rise time + fall time), where rise time is the time interval from the start point to the peak point of the wave, and fall time is the time interval from the peak point to the end point. The ascent and descent times are determined in the following manner: Eliminate DC bias from the residual current waveform segment centered on the candidate fluctuation moment; Search for the starting point of the waveform backward from the candidate fluctuation moment, and search for the ending point of the waveform backward; Determine the peak point between the start and end points; The time difference between the peak point and the starting point is taken as the rise time, and the time difference between the ending point and the peak point is taken as the fall time. If there are multiple local extrema between the starting point and the peak point of the waveform, the starting point is corrected based on the last extrema point. If there are multiple local extrema between the peak point and the end point, the end point is corrected based on the first extrema point, and the rise time and fall time are recalculated.

9. The intelligent leakage current detection and alarm method for a distribution box according to claim 8, characterized in that, The instantaneous difference is obtained as follows: Calculate the first difference between the current sampling point and the corresponding point of the previous cycle, and the average difference between the current sampling point and the corresponding point of the previous cycle in the neighborhood before and after it; Compare the relationship between the first difference and the average of the differences: If the first difference is greater than a set multiple of the average difference, then the first difference is taken as the instantaneous difference. If the first difference is less than or equal to the average difference, then the instantaneous difference is set to zero; If the difference falls between the two, the weighted average of the first difference and the average of the differences is taken as the instantaneous difference, where the weights are negatively correlated with the current noise level.

10. The intelligent leakage current detection and alarm method for a distribution box according to claim 9, characterized in that, Based on the stability of the determined rise and fall times, the parameters in the reverse tracing process are dynamically adjusted, including: If the rise time or fall time is corrected due to the existence of multiple local extrema, the length of the preset time window is increased and the search step size of the inverse difference operation is increased. If the ratio of rise time to fall time exceeds the set threshold, the preset time window length will remain unchanged, and the search step size of the inverse difference operation will be encrypted. Otherwise, the reverse tracing process will be executed with the default parameters.

Images