A pole-mounted circuit breaker partial discharge on-line monitoring system and method

Abstract

The application discloses a kind of pole-mounted circuit breaker partial discharge on-line monitoring system and method, it is related to partial discharge monitoring technical field, set up explicit partial discharge monitoring mechanism is used to set high frequency data acquisition mode to carry out partial discharge monitoring, analyze the apparent discharge quantity height, judge whether there is explicit partial discharge;Set up implicit partial discharge monitoring mechanism is used to set periodic low-frequency data acquisition mode to carry out acquisition and monitoring to weak partial discharge original signal data, and weak partial discharge signal is detected using three-level discrimination detection mechanism;Weak partial discharge early warning mechanism is set before being used to generate hierarchical warning according to the apparent discharge quantity analysis result of determined effective weak partial discharge signal;Using two-step layered partial discharge detection scheme, the detection and alarm efficiency of explicit partial discharge is guaranteed;Overcome the problem that weak partial discharge signal and noise signal are difficult to distinguish, greatly reduce false positive rate and false negative rate.

Description

Technical Field

[0001] This invention relates to the field of partial discharge monitoring technology, specifically to an online partial discharge monitoring system and method for pole-mounted circuit breakers. Background Technology

[0002] With the advancement of smart grid construction, higher demands are placed on the status perception and fault early warning capabilities of high-voltage electrical equipment. Partial discharge is an important early sign of insulation degradation within power equipment, and online monitoring of it is an effective means to prevent equipment failures and ensure the safe operation of the power grid. Currently, partial discharge monitoring technologies for large equipment such as GIS and transformers in substations are relatively mature, typically employing ultra-high frequency, ultrasonic, or pulse current methods, and are equipped with wired data transmission and external power supply systems. However, for the large number and wide distribution of outdoor pole-mounted circuit breakers, long-term outdoor service exposes them to temperature differences, salt spray, pollution, and lightning overvoltage, making the insulation epoxy poles and sealed bushings highly susceptible to damage. Micro-level air gap defects can lead to insulation degradation, which is strictly divided into three stages: weak discharge in the air gap, continuous stable discharge, and severe discharge before breakdown. Current online partial discharge monitoring systems often use a single threshold sampling and judgment mode, and only monitor two types of obvious high-intensity partial discharges: continuous stable discharge and severe discharge before breakdown. This has certain shortcomings. It does not distinguish between obvious and latent partial discharge conditions and relies solely on instantaneous discharge amplitude alarms, which leads to a very high rate of missed detection for weak partial discharges. Alarms can only be issued in the middle and late stages of insulation degradation, and it cannot achieve long-term early warning in the nascent stage. The warning is seriously delayed, and maintenance personnel miss the opportunity for low-cost power outage to eliminate defects, which can easily lead to sudden power outages in the distribution network. Summary of the Invention

[0003] The purpose of this invention is to provide an online monitoring system and method for partial discharge of pole-mounted circuit breakers, so as to solve the problems raised in the prior art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for online monitoring of partial discharge in a pole-mounted circuit breaker, the method comprising the following:

[0005] The overt partial discharge monitoring mechanism is set up to monitor partial discharge by setting up a high-frequency data acquisition method, analyzing the apparent discharge level, and determining whether overt partial discharge exists.

[0006] A latent partial discharge monitoring mechanism is set up to collect and monitor the raw signal data of weak partial discharge by setting a periodic low-frequency data acquisition method, and a three-level discrimination detection mechanism is used to detect the weak partial discharge signal.

[0007] The three-level discrimination and detection mechanism uses a three-level progressive detection method—voltage correlation detection, long-term trend change detection, and environmental reversibility detection—to distinguish between effective weak partial discharge signals and noise signals.

[0008] A weak partial discharge early warning mechanism is set up to generate graded early warnings based on the apparent discharge quantity analysis results of the determined effective weak partial discharge signals.

[0009] Furthermore, a two-step layered partial discharge detection scheme is used to adapt to different insulation degradation stages under different operating conditions, prioritizing the monitoring of mid-to-late stage manifest discharges, thus ensuring the detection and alarm efficiency of manifest partial discharges. The manifest partial discharge monitoring mechanism is set to a high sampling frequency of f1 for partial discharge signal data acquisition. The signal acquisition for manifest discharge monitoring typically uses a high-gain, high-frequency sampling range, for example, the sampling frequency f1 is set to 1MHz. The apparent discharge quantity Q is obtained based on the acquired instantaneous discharge pulse signal. The critical threshold for the apparent discharge quantity of manifest partial discharge is set as α, and the critical threshold for the apparent discharge quantity of severe discharge before breakdown is set as β. The apparent discharge quantity is defined as the amount of charge injected into both ends of the circuit breaker under test in a very short time in a specified measurement circuit. If the instrument reading is completely consistent with the reading generated by the actual partial discharge pulse inside the equipment, the injected charge quantity is the apparent discharge quantity. The apparent discharge quantity is not the actual discharge charge at the defect location. The actual charge transferred during the ionization of the air gap inside the insulation cannot be directly measured. The apparent discharge quantity is a circuit equivalent quantitative indicator, used only to uniformly measure the intensity of discharge and the severity of defects.

[0010] The acquired apparent discharge quantity was compared with the set threshold to determine whether there was a visible partial discharge. The analysis results are as follows:

[0011] If Q≥β, it means that the apparent discharge quantity obtained has reached the critical threshold of apparent discharge quantity before breakdown. If it is determined that there is a severe discharge before breakdown in the pole-mounted circuit breaker, the severe discharge warning a1 before breakdown is directly triggered.

[0012] If α≤Q<β, it means that the apparent discharge quantity obtained has reached the critical threshold of apparent discharge quantity before breakdown and violent discharge. It is determined that there is a dominant steady-state discharge in the pole-mounted circuit breaker, and the dominant steady-state discharge warning a2 is directly triggered.

[0013] If Q < α, it means that the apparent discharge quantity obtained has not reached the critical threshold of apparent discharge quantity for manifest partial discharge. It is determined that there is no manifest partial discharge in the pole-mounted circuit breaker, and the latent partial discharge monitoring mechanism is triggered to perform periodic monitoring of latent partial discharge.

[0014] Furthermore, when there is no obvious partial discharge, the system automatically switches to a low-noise periodic data acquisition mode, primarily monitoring latent and weak partial discharges. This balances fault response speed with early defect detection capability, reduces long-term sampling power consumption, and is suitable for outdoor pole-mounted power supply conditions. The latent partial discharge monitoring mechanism uses a low sampling frequency of f2 for partial discharge signal data acquisition. Signal acquisition during latent partial discharge monitoring typically employs a low-noise shielded, long-period, high-fidelity sampling setting, such as setting the sampling frequency f1 to 200kHz, reducing sampling power consumption while preserving the original waveform of the micropulse. The system collects raw signals of weak partial discharge within a data acquisition period T. A three-level discrimination and detection mechanism is employed to distinguish valid weak partial discharge signals from noise signals. This mechanism includes voltage correlation detection, long-term trend change detection, and environmental reversibility detection. This three-level discrimination and detection mechanism, relying on voltage correlation, long-term time-series trend, and environmental reversibility, accurately distinguishes valid weak partial discharge signals from noise signals, overcoming the difficulty in distinguishing between them and significantly reducing false alarm and false negative rates.

[0015] The voltage correlation detection is used to distinguish random noise from effective weak partial discharges and to eliminate the detected random noise that is not coupled to the power frequency voltage.

[0016] The environmental reversibility detection is used to ultimately distinguish between effective weak partial discharge signals and short-term condensation reversible discharge noise, and to eliminate the detected short-term condensation reversible discharge noise signals.

[0017] The long-term trend change detection is used to distinguish between radio frequency interference noise and effective weak partial discharge, and to eliminate the detected radio frequency interference noise.

[0018] Furthermore, the voltage correlation detection acquires the number of pulses of weak partial discharge signals, and simultaneously acquires the line operating voltage value, recording the line operating voltage at m levels, denoted as {U1, U2, ..., Um}, and synchronously recording the number of pulses {X1, X2, ..., Xm} occurring within the same fixed duration under each line operating voltage level; the fitting slope b between the line operating voltage data and the number of pulses is calculated using a univariate linear fitting formula; the calculation of the fitting slope first obtains the mean value of the line operating voltage. and the average number of pulses The fitted slope can be calculated using the following formula: ; where ξ represents the number of the collected line voltage data and pulse quantity data, ξ=1, 2, ...m; the fitted slope is analyzed, and the analysis results are as follows: if b≤0, it indicates that the pulse quantity and the line operating voltage do not meet the linear growth condition, and the monitored weak partial discharge signal is directly determined as random noise; if b>0, it indicates that the pulse quantity rises as the line operating voltage increases, and the linear determination coefficient between the line operating voltage data and the pulse quantity is further analyzed;

[0019] The linear determination coefficient r between the line operating voltage data and the pulse quantity is calculated by using the sum of squared deviations, and the linear determination coefficient is calculated in the form of sum of squared deviations according to the following formula: ; a linear determination coefficient threshold r0 is set, if r≥r0, it indicates that the pulse quantity has a linear growth relationship with the increase of the line operating voltage, and the monitored weak partial discharge signal is judged as an effective weak partial discharge signal; if r<r0, it indicates that there is no linear growth relationship between the pulse quantity and the line operating voltage, the monitored weak partial discharge signal is judged as random noise, and the weak partial discharge signal judged as random noise is eliminated.

[0020] Further, it is analyzed whether the weak partial discharge signal contains reversible discharge noise, a humidity sensor is used to collect the real-time relative humidity St, a detection window duration v is set, z pieces of humidity data and pulse mean data are collected at the same time node within one detection window, and the collected humidity data are arranged in time sequence and denoted as {S t1 , S t2 , ..., S ty}; the collected pulse mean data are arranged in time sequence and denoted as {M t1 , M t2 , ..., M ty}; the maximum value in the collected humidity data is denoted as Smax, and the minimum value is denoted as Smin; then the environmental humidity fluctuation value ΔS within the detection window duration can be calculated; where ΔS=Smax-Smin; a high environmental humidity fluctuation critical threshold ΔS0 is set, first, it is analyzed whether the humidity fluctuation degree is suitable for environmental reversibility detection, and the analysis results are as follows:

[0021] If ΔS<ΔS0, it indicates that the humidity fluctuation degree is low, and there is no obvious humidity drop process within the detection window duration, then the environmental reversibility detection is not triggered temporarily;

[0022] If ΔS≥ΔS0, it indicates that the humidity fluctuation degree is high, and there is an obvious humidity drop process within the detection window duration, then the environmental reversibility detection is triggered.

[0023] Further, after the environmental reversibility detection is triggered, a high humidity critical threshold S high and a low humidity critical threshold Slow ; Set the humidity value S i ≥S high The data is labeled as high humidity range sample data, and the humidity value S is... i ≤S low The data is labeled as low humidity range sample data; where i represents the time sequence number of the humidity data, i=t1, t2, ..., ty; in the pulse mean attenuation rate analysis, the high and low humidity ranges are set to select humidity change ranges with large drops, which are used to further analyze the change rate of the pulse mean during the large humidity drop process from high humidity to low humidity. Because the humidity of the pulse mean collected in the humidity range between the high humidity critical threshold and the low humidity critical threshold has a smaller humidity drop than that in the high humidity range and the low humidity period, it is not included in the pulse mean attenuation rate analysis process; in the pulse mean data, the high humidity range pulse mean and the low humidity range pulse mean are extracted according to the humidity data range, and the high humidity average pulse mean M is calculated respectively. high And the average pulse mean of low humidity after the humidity drops low The pulse mean attenuation rate after humidity decline is calculated using the following formula: Where η represents the pulse mean decay rate after the humidity drops;

[0024] Set a threshold η0 for the pulse mean attenuation rate; compare and analyze the actual pulse mean attenuation rate after humidity drop with the set threshold. The analysis results are as follows:

[0025] If η≥η0, it means that the pulse signal attenuation rate is abnormally high after the humidity drops. The detected partial discharge signal is determined to be reversible discharge noise caused by short-term condensation, and the partial discharge signal that is determined to be reversible discharge noise will be removed.

[0026] If η < η0, it indicates that the pulse signal attenuation rate is low after the humidity drops, and the signal still persists permanently in a low humidity environment. It is determined that the detected partial discharge signal is an effective weak partial discharge signal generated by the permanent micro-air gap inside the circuit breaker.

[0027] Furthermore, the long-term trend change detection is used to analyze the slope of the pulse mean change over time in weak partial discharge signals after removing random noise and reversible discharge noise. A pulse mean data acquisition period of x days is set; x weak pulse means are acquired, and the time series set of the acquired pulse means is denoted as {p1, p2, ..., px}. The time offset is in days, and the acquisition time of the pulse mean is denoted as τ, then τ = 1, 2, ..., x, that is, the first day is denoted as τ=1, the second day as τ=2, the third day as τ=3, and so on. The acquisition time of the pulse mean is used as the independent variable, and the weak pulse mean is used as the dependent variable. The time series growth slope k of the weak pulse mean is calculated using univariate linear fitting. The daily data acquisition time window duration is set to u, and the data... The number of valid pulses detected within the acquisition window is denoted as G. The average weak pulse value under the same operating conditions, pj = G / u, is calculated, where pj represents the average weak pulse value under the same operating conditions on day j. For example, if the daily data acquisition window duration is set to u = 30 minutes and 45 valid pulses are acquired, the average pulse value within the acquisition window is 1.5 pulses / minute. The time-series growth slope k is analyzed, and the results are as follows: If k > 0, it indicates that the average weak pulse value shows an upward trend with time, indicating a unidirectional, slowly increasing trend of insulation degradation, and the weak partial discharge signal is considered a valid weak partial discharge. If k ≤ 0, it indicates that the average pulse value does not have a stable growth pattern with time, and the weak partial discharge signal is identified as wind, rain, bird pecking, or radio frequency interference noise, which is then removed.

[0028] Furthermore, the weak partial discharge early warning mechanism is used to analyze the apparent discharge amount of the effective weak partial discharge signal remaining after noise removal, and the apparent discharge amount of the effective weak partial discharge signal is denoted as q; a weak partial discharge initiation critical threshold ω is set, and the apparent discharge amount of the monitored effective weak partial discharge signal is compared and analyzed with the set threshold to obtain the early warning type and early warning level of the effective weak partial discharge signal:

[0029] If q < ω, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is low, and a warning a3 is issued to prevent potential hazards from emerging. The retesting time h1 is reserved in advance, and it is recommended to continue to use periodic weak partial discharge monitoring for follow-up retesting.

[0030] If ω≤q<α, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is low. Therefore, an early warning of potential hazards a4 is issued, and a maintenance time h2 is reserved in advance. It is recommended that staff conduct inspections and maintenance within the reserved time.

[0031] If α≤q<β, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is too high, and a mid-term fault warning a2 is issued, and a shutdown and maintenance prompt is issued.

[0032] If q≥β, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is abnormally high, and an emergency fault alarm a1 is issued directly, and an immediate isolation shutdown alarm is issued.

[0033] Setting up a four-level pre-warning system can identify potential insulation defects in advance, enabling long-term pre-warning and reserving a window of opportunity for maintenance personnel to plan power outages and perform low-cost insulation remediation, thus preventing equipment from suddenly breaking down and tripping after it has developed into a visible discharge.

[0034] A partial discharge online monitoring system for pole-mounted circuit breakers, comprising a visible partial discharge monitoring module, a hidden partial discharge monitoring module, and a weak partial discharge early warning module;

[0035] The overt partial discharge monitoring module is used to set a high-frequency data acquisition method for partial discharge monitoring, analyze the apparent discharge level, and determine whether overt partial discharge exists.

[0036] The latent partial discharge monitoring module is used to collect and monitor the raw signal data of weak partial discharge by setting a periodic low-frequency data acquisition mode, and to detect the weak partial discharge signal by using a three-level discrimination and detection unit. The three-level discrimination and detection unit includes a three-level progressive detection method that uses voltage correlation detection, long-term trend change detection and environmental reversibility detection to distinguish between effective weak partial discharge signals and noise signals.

[0037] The weak partial discharge early warning module is used to generate graded early warnings based on the apparent discharge quantity analysis results of the determined effective weak partial discharge signals.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] This invention utilizes a two-step, layered partial discharge detection scheme, adaptable to different insulation degradation stages under varying operating conditions. It prioritizes monitoring mid-to-late stage manifest discharges, ensuring efficient detection and alarming of manifest partial discharges. When no manifest partial discharges are present, it automatically switches to a low-noise periodic data acquisition mode, primarily monitoring latent, weak partial discharges. This balances fault response speed with early defect detection capabilities, reduces long-term sampling power consumption, and is suitable for outdoor pole-mounted power supply conditions. During the detection of weak partial discharge signals, a three-level discrimination mechanism is implemented. Based on voltage correlation, long-term time-series trends, and environmental reversibility, it accurately distinguishes between valid weak partial discharges and noise signals, overcoming the difficulty in differentiating between them and significantly reducing false alarm and missed alarm rates. A four-level pre-warning system is established to identify nascent insulation problems in advance, providing long-term pre-warning and reserving a window for planned power outages and low-cost insulation remediation for maintenance personnel, preventing sudden breakdown and tripping after the equipment develops into a manifest discharge. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the process flow for an online monitoring method for partial discharge of a pole-mounted circuit breaker according to the present invention. Detailed Implementation

[0041] 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.

[0042] like Figure 1 As shown, the present invention provides a technical solution: an online monitoring method for partial discharge of a pole-mounted circuit breaker, the method comprising the following:

[0043] The overt partial discharge monitoring mechanism is set up to monitor partial discharge by setting up a high-frequency data acquisition method, analyzing the apparent discharge level, and determining whether overt partial discharge exists.

[0044] A latent partial discharge monitoring mechanism is set up to collect and monitor the raw signal data of weak partial discharge by setting a periodic low-frequency data acquisition method, and a three-level discrimination detection mechanism is used to detect the weak partial discharge signal.

[0045] The three-level discrimination and detection mechanism uses a three-level progressive detection method—voltage correlation detection, long-term trend change detection, and environmental reversibility detection—to distinguish between effective weak partial discharge signals and noise signals.

[0046] A weak partial discharge early warning mechanism is set up to generate graded early warnings based on the apparent discharge quantity analysis results of the determined effective weak partial discharge signals.

[0047] The overt partial discharge monitoring mechanism uses a high sampling frequency of f1 to acquire partial discharge signal data. The apparent discharge quantity Q is obtained based on the acquired instantaneous discharge pulse signal. A critical threshold α is set for the apparent discharge quantity of overt partial discharge, and a critical threshold β is set for the apparent discharge quantity of violent discharge before breakdown. The acquired apparent discharge quantity is compared with the set thresholds to determine whether overt partial discharge exists. The analysis results are as follows:

[0048] If Q≥β, it means that the apparent discharge quantity obtained has reached the critical threshold of apparent discharge quantity before breakdown. If it is determined that there is a severe discharge before breakdown in the pole-mounted circuit breaker, the severe discharge warning a1 before breakdown is directly triggered.

[0049] If α≤Q<β, it means that the apparent discharge quantity obtained has reached the critical threshold of apparent discharge quantity before breakdown and violent discharge. It is determined that there is a dominant steady-state discharge in the pole-mounted circuit breaker, and the dominant steady-state discharge warning a2 is directly triggered.

[0050] If Q < α, it means that the apparent discharge quantity obtained has not reached the critical threshold of apparent discharge quantity for manifest partial discharge. It is determined that there is no manifest partial discharge in the pole-mounted circuit breaker, and the latent partial discharge monitoring mechanism is triggered to perform periodic monitoring of latent partial discharge.

[0051] The latent partial discharge monitoring mechanism uses a low sampling frequency of f2 to collect partial discharge signal data and sets a data acquisition period T. Within one data acquisition period, weak partial discharge raw signals are collected. A three-level discrimination and detection mechanism is used to distinguish effective weak partial discharge signals from noise signals for the collected weak partial discharge raw signals. The three-level discrimination and detection mechanism includes voltage correlation detection, long-term trend change detection, and environmental reversibility detection.

[0052] Voltage correlation detection is used to distinguish random noise from effective weak partial discharges and to eliminate the detected random noise that is not coupled to the power frequency voltage.

[0053] Environmental reversibility detection is used to ultimately distinguish between effective weak partial discharge signals and short-term condensation reversible discharge noise, and to eliminate the detected short-term condensation reversible discharge noise signals.

[0054] Long-term trend change detection is used to distinguish between radio frequency interference noise and effective weak partial discharge, and the detected radio frequency interference noise is eliminated.

[0055] Voltage correlation detection collects the number of pulses of weak partial discharge signals while simultaneously collecting the line operating voltage value. The line operating voltage is recorded in m tiers, denoted as {U1, U2, ..., Um}. The number of pulses {X1, X2, ..., Xm} occurring within the same fixed time period under each line operating voltage tier is also recorded. The fitting slope b between the line operating voltage data and the number of pulses is calculated using a univariate linear fitting formula. Analysis of the fitting slope yields the following results: If b ≤ 0, it indicates that the number of pulses does not meet the linear growth condition of the line operating voltage, and the detected weak partial discharge signal is directly determined to be random noise; if b > 0, it indicates that the number of pulses increases with the increase of the line operating voltage, and the linear determination coefficient between the line operating voltage data and the number of pulses is further analyzed.

[0056] Calculate the linear determination coefficient r between the line operating voltage data and the number of pulses using the combined form of sum of squared deviations, set a threshold r0 for the linear determination coefficient. If r≥r0, it indicates that the number of pulses increases linearly with the rise of the line operating voltage, and the monitored weak partial discharge signal is determined to be an effective weak partial discharge signal; if r<r0, it indicates that there is no linear growth relationship between the number of pulses and the line operating voltage, the monitored weak partial discharge signal is determined to be random noise, and the weak partial discharge signal determined as random noise is eliminated.

[0057] Analyze whether the weak partial discharge signal contains reversible discharge noise, collect the real-time relative humidity St using a humidity sensor, set the detection window duration v, collect z pieces of humidity data and pulse mean data at the same time node within one detection window, and arrange the collected humidity data in chronological order, denoted as {S t1 , S t2 , ..., S ty}; arrange the collected pulse mean data in chronological order, denoted as {M t1 , M t2 , ..., M ty}; denote the maximum value in the collected humidity data as Smax and the minimum value as Smin; then the environmental humidity fluctuation value △S within the detection window duration can be calculated; set the critical threshold △S0 for high environmental humidity fluctuation, first analyze whether the humidity fluctuation degree is suitable for environmental reversibility detection, and the analysis results are as follows:

[0058] If △S<△S0, it indicates that the humidity fluctuation degree is relatively low, and there is no obvious humidity drop process within the detection window duration, then the environmental reversibility detection is not triggered temporarily;

[0059] If △S≥△S0, it indicates that the humidity fluctuation degree is relatively high, and there is an obvious humidity drop process within the detection window duration, then the environmental reversibility detection is triggered.

[0060] After triggering the environmental reversibility detection, set the high humidity critical threshold S high and the low humidity critical threshold S low ; mark the data with the humidity value S i ≥S high as high humidity interval sample data, and mark the data with the humidity value S i ≤S low as low humidity interval sample data; where i represents the chronological number of the humidity data, i=t1, t2, ..., ty; extract the high humidity interval pulse mean and the low humidity interval pulse mean from the pulse mean data according to the humidity data intervals, and respectively calculate the high-humidity average pulse mean M high and the low-humidity average pulse mean M after the humidity drops low , and calculate the attenuation rate of the pulse mean after the humidity drops according to the following formula: Where η represents the pulse mean decay rate after the humidity drops;

[0061] Set a threshold η0 for the pulse mean attenuation rate; compare and analyze the actual pulse mean attenuation rate after humidity drop with the set threshold. The analysis results are as follows:

[0062] If η≥η0, it means that the pulse signal attenuation rate is abnormally high after the humidity drops. The detected partial discharge signal is determined to be reversible discharge noise caused by short-term condensation, and the partial discharge signal that is determined to be reversible discharge noise will be removed.

[0063] If η < η0, it indicates that the pulse signal attenuation rate is low after the humidity drops, and the signal still persists permanently in a low humidity environment. It is determined that the detected partial discharge signal is an effective weak partial discharge signal generated by the permanent micro-air gap inside the circuit breaker.

[0064] Long-term trend change detection is used to analyze the slope of the pulse mean change over time in weak partial discharge signals after removing random noise and reversible discharge noise. A pulse mean data acquisition period of x days is set; x weak pulse means are collected, and the time series set of the collected pulse means is denoted as {p1, p2, ..., px}, with a time offset in days. The acquisition time of the pulse mean is denoted as τ, then τ = 1, 2, ..., x; the acquisition time of the pulse mean is used as the independent variable, and the weak pulse mean is used as the dependent variable. The variable is calculated using univariate linear fitting to determine the temporal growth slope k of the weak pulse mean. Analysis of the temporal growth slope k yields the following results: If k > 0, it indicates that the weak pulse mean shows an upward trend over time, suggesting a slow, unidirectional increase in insulation degradation, thus the weak partial discharge signal is a valid weak partial discharge. If k ≤ 0, it indicates that the pulse mean does not exhibit a stable growth pattern over time, suggesting the weak partial discharge signal is noise from wind, rain, birdsong, or radio frequency interference, thus the noise is eliminated.

[0065] A weak partial discharge early warning mechanism is used to analyze the apparent discharge quantity of the effective weak partial discharge signal remaining after noise removal. The apparent discharge quantity of the effective weak partial discharge signal is denoted as q. A critical threshold ω for the germination of weak partial discharge is set. The apparent discharge quantity of the monitored effective weak partial discharge signal is compared and analyzed with the set threshold to obtain the warning type and warning level of the effective weak partial discharge signal.

[0066] If q < ω, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is low, and a warning a3 is issued to prevent potential hazards from emerging. The retesting time h1 is reserved in advance, and it is recommended to continue to use periodic weak partial discharge monitoring for follow-up retesting.

[0067] If ω≤q<α, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is low. Therefore, an early warning of potential hazards a4 is issued, and a maintenance time h2 is reserved in advance. It is recommended that staff conduct inspections and maintenance within the reserved time.

[0068] If α≤q<β, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is too high, and a mid-term fault warning a2 is issued, and a shutdown and maintenance prompt is issued.

[0069] If q≥β, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is abnormally high, and an emergency fault alarm a1 is issued directly, and an immediate isolation shutdown alarm is issued.

[0070] A partial discharge online monitoring system for pole-mounted circuit breakers, comprising a visible partial discharge monitoring module, a hidden partial discharge monitoring module, and a weak partial discharge early warning module;

[0071] The overt partial discharge monitoring module is used to set up a high-frequency data acquisition method for partial discharge monitoring, analyze the apparent discharge level, and determine whether overt partial discharge exists.

[0072] The latent partial discharge monitoring module is used to collect and monitor the raw signal data of weak partial discharge by setting a periodic low-frequency data acquisition mode. It adopts a three-level discrimination and detection unit to detect weak partial discharge signals. The three-level discrimination and detection unit includes a three-level progressive detection method that uses voltage correlation detection, long-term trend change detection, and environmental reversibility detection to distinguish between effective weak partial discharge signals and noise signals.

[0073] The weak partial discharge early warning module is used to generate graded early warnings based on the apparent discharge quantity analysis results of the determined effective weak partial discharge signals.

[0074] Example 1: A dominant partial discharge monitoring mechanism is used to collect partial discharge signal data at a high sampling frequency of f1. The apparent discharge quantity Q = 4pC is obtained based on the collected instantaneous discharge pulse signal. The critical threshold for the apparent discharge quantity of dominant partial discharge is set as α = 3pC, and the critical threshold for the apparent discharge quantity of violent discharge before breakdown is set as β = 10pC. The acquired apparent discharge quantity is compared with the set thresholds to determine whether dominant partial discharge exists. The analysis results are as follows:

[0075] If α≤Q<β, it means that the apparent discharge quantity obtained has reached the critical threshold of apparent discharge quantity before breakdown and violent discharge. If it is determined that there is a dominant steady-state discharge in the pole-mounted circuit breaker, then the dominant steady-state discharge warning a2 is directly triggered.

[0076] Example 2: Environmental reversibility detection is used to distinguish between effective weak partial discharge signals and short-term condensation reversible discharge noise. A humidity sensor is used to collect real-time relative humidity (St). The detection window duration is set to v=4 days. Within one detection window, y=4 humidity data points and pulse mean data are collected at the same time points. The collected humidity data are arranged chronologically as {80%, 60%, 50%, 40%}; the collected pulse mean data are arranged chronologically as {28, 26, 21, 12}; the maximum value of the collected humidity data is denoted as Smax=80%, and the minimum value as Smin=40%. The environmental humidity fluctuation value ΔS=40% within the detection window duration can then be calculated. A high humidity fluctuation threshold ΔS0=15% is set. First, the suitability of the humidity fluctuation level for environmental reversibility detection is analyzed. The analysis results are as follows:

[0077] If △S≥△S0, it indicates that the humidity fluctuation is too high and there is a significant humidity drop within the detection window, thus triggering environmental reversibility detection.

[0078] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for online monitoring of partial discharge in a pole-mounted circuit breaker, characterized in that: The method includes the following: The overt partial discharge monitoring mechanism is set up to monitor partial discharge by setting up a high-frequency data acquisition method, analyzing the apparent discharge level, and determining whether overt partial discharge exists. A latent partial discharge monitoring mechanism is set up to collect and monitor the raw signal data of weak partial discharge by setting a periodic low-frequency data acquisition method, and a three-level discrimination detection mechanism is used to detect the weak partial discharge signal. The three-level discrimination and detection mechanism uses a three-level progressive detection method—voltage correlation detection, long-term trend change detection, and environmental reversibility detection—to distinguish between effective weak partial discharge signals and noise signals. A weak partial discharge early warning mechanism is set up to generate graded early warnings based on the apparent discharge quantity analysis results of the determined effective weak partial discharge signals.

2. The method for online monitoring of partial discharge in a pole-mounted circuit breaker according to claim 1, characterized in that: The overt partial discharge monitoring mechanism uses a high sampling frequency of f1 to collect partial discharge signal data. Based on the collected instantaneous discharge pulse signal, the apparent discharge quantity Q is obtained. A critical threshold α is set for the apparent discharge quantity of overt partial discharge, and a critical threshold β is set for the apparent discharge quantity of violent discharge before breakdown. The obtained apparent discharge quantity is compared with the set thresholds to determine whether overt partial discharge exists. The analysis results are as follows: If Q≥β, it means that the apparent discharge quantity obtained has reached the critical threshold of apparent discharge quantity before breakdown. If it is determined that there is a severe discharge before breakdown in the pole-mounted circuit breaker, the severe discharge warning a1 before breakdown is directly triggered. If α≤Q<β, it means that the apparent discharge quantity obtained has reached the critical threshold of apparent discharge quantity before breakdown and violent discharge. It is determined that there is a dominant steady-state discharge in the pole-mounted circuit breaker, and the dominant steady-state discharge warning a2 is directly triggered. If Q < α, it means that the apparent discharge quantity obtained has not reached the critical threshold of apparent discharge quantity for manifest partial discharge. It is determined that there is no manifest partial discharge in the pole-mounted circuit breaker, and the latent partial discharge monitoring mechanism is triggered to perform periodic monitoring of latent partial discharge.

3. The method for online monitoring of partial discharge in a pole-mounted circuit breaker according to claim 1, characterized in that: The latent partial discharge monitoring mechanism uses a low sampling frequency of f2 to collect partial discharge signal data and sets a data acquisition period T. Within one data acquisition period, weak partial discharge raw signals are collected. A three-level discrimination and detection mechanism is used to distinguish effective weak partial discharge signals from noise signals. The three-level discrimination and detection mechanism includes voltage correlation detection, long-term trend change detection, and environmental reversibility detection. The voltage correlation detection is used to distinguish random noise from effective weak partial discharges and to eliminate the detected random noise that is not coupled to the power frequency voltage. The environmental reversibility detection is used to ultimately distinguish between effective weak partial discharge signals and short-term condensation reversible discharge noise, and to eliminate the detected short-term condensation reversible discharge noise signals. The long-term trend change detection is used to distinguish between radio frequency interference noise and effective weak partial discharge, and to eliminate the detected radio frequency interference noise.

4. The method for online monitoring of partial discharge in a pole-mounted circuit breaker according to claim 3, characterized in that: The voltage correlation detection collects the number of pulses of weak partial discharge signals, and simultaneously collects the line operating voltage values, records m levels of line operating voltage in a graded manner, the collected line operating voltages are denoted as {U1, U2, ..., Um}, and synchronously records the number of pulses {X1, X2, ..., Xm} appearing within the same fixed duration under each level of line operating voltage; the fitting slope b between the line operating voltage data and the number of pulses is calculated using a unary linear fitting formula; the fitting slope is analyzed, and the analysis results are as follows: if b ≤ 0, it means that the number of pulses and the line operating voltage do not meet the linear growth condition, and the monitored weak partial discharge signal is directly determined as random noise; if b > 0, it means that the number of pulses rises as the line operating voltage increases, and the linear determination coefficient between the line operating voltage data and the number of pulses is further analyzed; The linear determination coefficient r between the line operating voltage data and the number of pulses is calculated using the sum of squared deviations form, a linear determination coefficient threshold r0 is set, if r ≥ r0, it means that the number of pulses has a linear growth relationship with the increase of the line operating voltage, and the monitored weak partial discharge signal is determined to be an effective weak partial discharge signal; if r < r0, it means that there is no linear growth relationship between the number of pulses and the line operating voltage, the monitored weak partial discharge signal is determined as random noise, and the weak partial discharge signal determined as random noise is eliminated.

5. The method for online monitoring of partial discharge in a pole-mounted circuit breaker according to claim 3, characterized in that: To analyze whether the weak partial discharge signal contains reversible discharge noise, a humidity sensor is used to collect real-time relative humidity St. A detection window duration v is set, and within one detection window, z humidity data points and pulse mean data points are collected at the same time points. The collected humidity data are arranged chronologically and denoted as {S}. t1 S t2 S ty The collected pulse mean data are arranged in time sequence and denoted as {M}. t1 M t2 M ty The maximum value of the collected humidity data is denoted as Smax, and the minimum value as Smin. The environmental humidity fluctuation value ΔS within the detection window can then be calculated. A high humidity fluctuation threshold ΔS0 is set, and the suitability of the humidity fluctuation level for environmental reversibility detection is analyzed. The analysis results are as follows: If ΔS < ΔS0, it means that the humidity fluctuation degree is low, there is no obvious humidity drop process within the detection window duration, and the environmental reversibility detection is not triggered temporarily; If ΔS ≥ ΔS0, it means that the humidity fluctuation degree is high, there is an obvious humidity drop process within the detection window duration, and the environmental reversibility detection is triggered.

6. The method for online monitoring of partial discharge in a pole-mounted circuit breaker according to claim 5, characterized in that: After triggering the environmental reversibility detection, set the high humidity critical threshold S. high and low humidity critical threshold S low ; Set the humidity value S i ≥S high The data is labeled as high humidity range sample data, and the humidity value S is... i ≤S low The data is labeled as low humidity interval sample data; where i represents the time sequence number of the humidity data, i=t1, t2, ..., ty; the pulse mean of the high humidity interval and the pulse mean of the low humidity interval are extracted from the pulse mean data according to the humidity data interval, and the high humidity average pulse mean M is calculated respectively. high And the average pulse mean of low humidity after the humidity drops low The pulse mean attenuation rate after humidity decline is calculated using the following formula: Where η represents the pulse mean decay rate after the humidity drops; A pulse mean attenuation rate threshold η0 is set; the actually obtained pulse mean attenuation rate after the humidity drops is compared and analyzed with the set threshold, and the analysis results are as follows: If η ≥ η0, it means that the attenuation rate of the pulse signal after the humidity drops is abnormally high, the detected partial discharge signal is determined to be reversible discharge noise caused by short-term condensation, and the partial discharge signal determined as reversible discharge noise is eliminated; If η < η0, it means that the attenuation rate of the pulse signal after the humidity drops is low, the signal still persists permanently in a low humidity environment, and the detected partial discharge signal is determined to be an effective weak partial discharge signal generated by the permanent micro-air gap inside the circuit breaker.

7. The method for online monitoring of partial discharge in a pole-mounted circuit breaker according to claim 3, characterized in that: The long-term trend change detection is used to analyze the slope of the pulse mean change over time in weak partial discharge signals after removing random noise and reversible discharge noise. A pulse mean data acquisition period of x days is set; x weak pulse means are acquired, and the time series set of the acquired pulse means is denoted as {p1, p2, ..., px}, with a time offset in days. The acquisition time of the pulse mean is denoted as τ, then τ = 1, 2, ..., x; the acquisition time of the pulse mean is used as the independent variable, and the weak pulse mean is used as... The dependent variable is calculated using univariate linear fitting to determine the temporal growth slope k of the weak pulse mean. Analysis of the temporal growth slope k yields the following results: If k > 0, it indicates that the weak pulse mean shows an upward trend over time, suggesting a slow, unidirectional increase in insulation degradation, thus the weak partial discharge signal is considered a valid weak partial discharge. If k ≤ 0, it indicates that the pulse mean does not exhibit a stable growth pattern over time, suggesting the weak partial discharge signal is noise from wind, rain, birdsong, or radio frequency interference, thus the noise is eliminated.

8. The method for online monitoring of partial discharge in a pole-mounted circuit breaker according to claim 2, characterized in that: The weak partial discharge early warning mechanism is used to analyze the apparent discharge amount of the effective weak partial discharge signal remaining after noise removal, and the apparent discharge amount of the effective weak partial discharge signal is denoted as q; a weak partial discharge initiation critical threshold ω is set, and the apparent discharge amount of the monitored effective weak partial discharge signal is compared and analyzed with the set threshold to obtain the early warning type and early warning level of the effective weak partial discharge signal: If q < ω, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is low, and a warning a3 is issued to prevent potential hazards from emerging. The retesting time h1 is reserved in advance, and it is recommended to continue to use periodic weak partial discharge monitoring for follow-up retesting. If ω≤q<α, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is low. Therefore, an early warning of potential hazards a4 is issued, and a maintenance time h2 is reserved in advance. It is recommended that staff conduct inspections and maintenance within the reserved time. If α≤q<β, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is too high, and a mid-term fault warning a2 is issued, and a shutdown and maintenance prompt is issued. If q≥β, it indicates that the apparent discharge quantity of the effective weak partial discharge signal is abnormally high, and an emergency fault alarm a1 is issued directly, and an immediate isolation shutdown alarm is issued.

9. A partial discharge online monitoring system for pole-mounted circuit breakers, applied to the partial discharge online monitoring method for pole-mounted circuit breakers as described in any one of claims 1-8; characterized in that: The system includes a visible partial discharge monitoring module, a hidden partial discharge monitoring module, and a weak partial discharge early warning module; The overt partial discharge monitoring module is used to set a high-frequency data acquisition method for partial discharge monitoring, analyze the apparent discharge level, and determine whether overt partial discharge exists. The latent partial discharge monitoring module is used to collect and monitor the raw signal data of weak partial discharge by setting a periodic low-frequency data acquisition mode, and to detect the weak partial discharge signal by using a three-level discrimination and detection unit. The three-level discrimination and detection unit includes a three-level progressive detection method that uses voltage correlation detection, long-term trend change detection and environmental reversibility detection to distinguish between effective weak partial discharge signals and noise signals. The weak partial discharge early warning module is used to generate graded early warnings based on the apparent discharge quantity analysis results of the determined effective weak partial discharge signals.

Images