An intelligent circuit breaker health online diagnosis system, device and method

Abstract

The application is particularly a kind of intelligent circuit breaker health degree online diagnosis system, equipment and method, relates to the technical field of power equipment monitoring and diagnosis, and comprises: an environmental stress perception and quantification module; an electrical response synchronous monitoring module; a feature extraction and correlation modeling module; and a tolerance degradation evaluation module.In the application, multiple types of high-precision sensors capture environmental parameters such as contamination, temperature and humidity in real time, and combine GPS / Beidou dual-mode synchronization technology to synchronously collect electrical signals such as leakage current and partial discharge, thereby achieving spatio-temporal alignment of environmental and electrical data.Compared with traditional offline detection, the application does not need to interrupt the operation of the equipment and can dynamically capture the insulation degradation process.

Description

Technical Field

[0001] This invention relates to the field of power equipment monitoring and diagnosis technology, and in particular to an online health diagnosis system, device and method for intelligent circuit breakers. Background Technology

[0002] As the core equipment for circuit switching control and fault protection in power systems, the health status of intelligent circuit breakers directly determines the safety, reliability, and stability of power system operation. With the development of power systems towards ultra-high voltage and intelligent operation, the operating environment of circuit breakers is becoming increasingly complex. Environmental stresses such as pollution and drastic changes in temperature and humidity continuously act on the insulation structure of circuit breakers, easily leading to insulation aging and performance degradation, which in turn causes equipment failure and serious consequences such as large-scale power outages.

[0003] Existing circuit breaker health diagnostic technologies mostly rely on periodic offline testing or monitoring of single electrical parameters, which has significant limitations:

[0004] Offline detection requires interrupting device operation, resulting in high maintenance costs and an inability to capture real-time degradation processes in dynamic environments;

[0005] Single-parameter monitoring ignores the dynamic correlation between environmental stress and electrical response, and is prone to misjudgment or omission due to environmental interference;

[0006] Therefore, there is an urgent need to build an online diagnostic system that can sense environmental stress in real time, monitor electrical response synchronously, establish dynamic correlation models, and realize multi-dimensional degradation assessment, so as to solve the problems of diagnostic lag, insufficient accuracy, and poor targeting in existing technologies. Summary of the Invention

[0007] The purpose of this invention is to provide an online health diagnosis system, device, and method for intelligent circuit breakers in order to solve the above-mentioned problems.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] An online health diagnostic system for intelligent circuit breakers includes:

[0010] The environmental stress sensing and quantification module is configured to monitor environmental parameters by deploying multiple types of sensors, calculate the comprehensive pollution and wet index through calibration and feature engineering, and quantify environmental stress and determine high-stress events.

[0011] The electrical response synchronous monitoring module is configured to use GPS / BeiDou dual-mode synchronous technology to synchronously collect leakage current, partial discharge, and insulation temperature signals during environmental stress changes. After preprocessing, key electrical features are extracted to ensure spatiotemporal alignment of environmental and electrical data.

[0012] The feature extraction and correlation modeling module is configured to filter core environmental and electrical features, construct a dynamic correlation model of environmental stress and electrical response, quantify feature sensitivity, and identify early degradation indicators.

[0013] The tolerance degradation assessment module is configured to obtain degradation assessment coefficients from three dimensions: self-recovery ability, safety margin, and aging rate, and match the corresponding health status level based on the degradation assessment coefficients.

[0014] Preferably, the environmental stress sensing and quantification module specifically includes:

[0015] A laser scattering type pollution sensor is selected and installed on the windward side of the three-phase insulator of the circuit breaker.

[0016] Digital temperature and humidity sensors are installed inside the mechanism housing; the outer casing of the arc-extinguishing chamber; and inside the busbar chamber.

[0017] To obtain the relevant data and calibrate the humidity sensor data;

[0018] And calculate the comprehensive index of sewage and wetness. : ;

[0019] in,

[0020] For equivalent salt density;

[0021] The preset maximum permissible equivalent salt density;

[0022] It is a piecewise nonlinear humidity function;

[0023] The rate of temperature change over a preset duration;

[0024] , , These are the preset weighting factors;

[0025] A high-stress event is defined as the occurrence of any of the following conditions:

[0026] Comprehensive index of sewage and wetland The value exceeds the preset threshold and continues for the preset allowed duration;

[0027] When the condensation warning signal is triggered;

[0028] The equivalent salt density reaches the preset percentage of the maximum permissible equivalent salt density.

[0029] Preferably, the electrical response synchronization monitoring module specifically includes:

[0030] When a high-stress event occurs, the following data acquisition is triggered:

[0031] Leakage current data: A wideband Rogowski coil micro-current sensor is installed at the grounding lead of the circuit breaker insulator to obtain the data.

[0032] Partial discharge data: UHF sensor and HFCT sensor are used; the UHF sensor is adsorbed on the outside of the arc-extinguishing chamber shell, and the HFCT sensor is sleeved on the grounding lead; the sensor shell is grounded;

[0033] Insulation temperature data: Distributed fiber optic temperature sensors are attached to the surface of the basin insulator and both ends of the insulating tie rod;

[0034] Synchronize the acquired data;

[0035] Perform appropriate preprocessing on each type of data;

[0036] Leakage current processing: Wavelet denoising is used to eliminate white noise, and the power frequency component and the effective values ​​of each harmonic are extracted, where:

[0037] Partial discharge processing: pulse classification, elimination of external interference, and generation of phase-resolved spectra;

[0038] Outliers in the temperature data are removed, and the temperature characteristics of each monitoring point on the insulation surface are calculated.

[0039] Preferably, the feature extraction and association modeling module specifically includes:

[0040] Environmental feature set, including the combined pollutant and wet index and its 10-minute rate of change Duration of continuous high humidity, number of temperature cycles, rate of dirt accumulation, and cumulative condensation time.

[0041] The electrical characteristic set includes the fundamental RMS value, the ratio of the third harmonic to the fundamental value, the pulse peak value, the total harmonic distortion rate, the correlation dimension, and the maximum Lyapunov exponent in the leakage current characteristics.

[0042] It also includes the discharge initiation voltage, average discharge quantity, PRPD spectrum skewness, kurtosis, discharge phase concentration, maximum discharge quantity, and discharge frequency in the partial discharge characteristics;

[0043] It also includes temperature characteristics such as insulation surface temperature rise, temperature gradient, and temperature response hysteresis time.

[0044] Preferably, the method further includes:

[0045] A mapping from environmental characteristics to electrical characteristics is established using time series analysis and machine learning models.

[0046] Align time series of environmental and electrical characteristics through dynamic time warping;

[0047] A global sensitivity analysis method is used to quantify the impact of various environmental characteristics on electrical characteristics:

[0048] Calculate the sensitivity coefficients; sum the absolute values ​​of the calculated sensitivity coefficients for each electrical feature, and use the top 3 electrical features as the core indicators of early degradation.

[0049] Finally, we obtain the trained LSTM model, core feature set, sensitivity coefficient ranking table, similarity matrix, and electrical feature prediction values ​​under given environmental stress.

[0050] Preferably, the tolerance degradation assessment module specifically includes:

[0051] Acquire historical data on the health status of the equipment within a preset time period after it is put into operation, and calculate the mean value of the target electrical characteristics under the corresponding environmental stress.

[0052] When environmental stress decreases, the average value of the target's electrical characteristics is collected within the corresponding time period;

[0053] The difference between the mean value under environmental stress and the mean value of each target electrical characteristic during the period of environmental stress reduction is obtained. The absolute value is then divided by the mean value under environmental stress to obtain the recovery deviation value.

[0054] A preset recovery deviation threshold is set, and recovery deviation values ​​that exceed the recovery deviation threshold are recorded as abnormal recovery deviation values.

[0055] Obtain the recovery anomaly deviation values ​​corresponding to each monitoring duration during the period of environmental stress reduction. Sort the obtained recovery anomaly deviation values ​​in descending order of numerical value, extract the three largest recovery anomaly deviation values, take the average value, and record it as the recovery anomaly quantification value. Obtain the recovery anomaly quantification value corresponding to each electrical feature, and after normalizing each recovery anomaly quantification value, record the largest value as the anomaly determination quantification value.

[0056] Preferably, the method further includes:

[0057] Take the device's health status, different The corresponding target electrical characteristic values ​​are used to estimate the upper and lower confidence limits of the fitted electrical characteristics using kernel density estimation, thus forming a normal range;

[0058] Calculate unsafe values: ,in, The upper limit of confidence is 95%. This is the 95% confidence lower limit; This is the actual measured value of the electrical characteristic at the current moment; A function that takes the minimum value;

[0059] Obtain the unsafe values ​​of each electrical feature, preset the unsafe value thresholds for each electrical feature, mark the unsafe values ​​of electrical features that are greater than the preset unsafe value thresholds, and calculate the difference between the unsafe value and the unsafe value threshold to obtain the unsafe difference.

[0060] After normalizing the unsafe differences corresponding to each electrical characteristic, the values ​​are sorted in descending order to obtain a normalized set. The time points corresponding to each value and the time differences between each adjacent value are extracted from the set. A time difference threshold is preset, and time differences less than the time difference threshold are recorded as abnormal time differences.

[0061] Obtain all outlier time differences, extract the smallest outlier time difference, and take its reciprocal to obtain the insecurity quantification value.

[0062] Preferably, the method further includes:

[0063] Obtain the monthly average values ​​of electrical characteristics within a preset time period, based on the exponential law of insulation aging: Taking the logarithm of both sides, we get: The aging slope was fitted using linear regression. ,in, Let be the target electrical characteristic value at time t; The initial values ​​of the target electrical characteristics at the initial moment;

[0064] Obtain the aging slope of each electrical feature. , and the various aging slopes After normalization, calculate each aging slope. The difference between the corresponding maximum permissible aging slope is taken as the absolute value to obtain the aging difference value corresponding to each electrical characteristic, and the largest aging difference value is recorded as the aging quantification value.

[0065] After normalizing the abnormality determination quantification value, the unsafety quantification value, and the aging quantification value, a weighted summation is performed to obtain the degradation assessment coefficient.

[0066] A method for online health diagnosis of intelligent circuit breakers includes:

[0067] Deploy multiple types of sensors to monitor environmental parameters, calculate the comprehensive pollution and wet index through calibration and feature engineering, and complete the quantification of environmental stress and the determination of high-stress events;

[0068] During high-stress events, leakage current, partial discharge, and insulation temperature signals are simultaneously acquired, and key electrical characteristics are extracted after preprocessing.

[0069] We screened the core environmental and electrical features, constructed a dynamic correlation model using DTW time series alignment and LSTM deep learning, quantified feature sensitivity, and identified early degradation indicators.

[0070] Using the equipment health baseline as a reference, the quantitative values ​​of the three dimensions of self-recovery capability, safety margin, and aging rate are calculated respectively to complete the tolerance assessment of each dimension;

[0071] The degradation evaluation coefficient is obtained by normalizing the three-dimensional quantified values ​​and then weighting and summing them. Based on this coefficient, the health status level of the smart circuit breaker is matched.

[0072] An online health diagnostic device for intelligent circuit breakers includes: a housing, and an online health diagnostic system for intelligent circuit breakers integrated within the housing.

[0073] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0074] 1. This invention uses multiple types of high-precision sensors to capture environmental parameters such as dirt, temperature and humidity in real time, and combines GPS / BeiDou dual-mode synchronization technology to synchronously collect electrical signals such as leakage current and partial discharge, so as to achieve spatiotemporal alignment of environmental and electrical data; compared with traditional offline detection, it does not require interruption of equipment operation and can dynamically capture the insulation degradation process.

[0075] 2. This invention provides a quantitative assessment through three dimensions: self-recovery capability, safety margin, and aging rate, taking into account both individual equipment differences and the cumulative effect of aging. It employs kernel density estimation to fit the normal range of electrical characteristics and the Morris method to screen core degradation indicators, making the assessment more consistent with the actual operating patterns of the equipment. The final degradation assessment coefficient can accurately match the health level, clarify the timing and focus of maintenance, avoid blind maintenance, reduce operation and maintenance costs, and improve the operational reliability of intelligent circuit breakers. Attached Figure Description

[0076] Further details, features, and advantages of this application are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which:

[0077] Figure 1 This is a system structure diagram of the present invention;

[0078] Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0079] Several embodiments of this application will now be described in more detail with reference to the accompanying drawings to enable those skilled in the art to implement this application. This application may be embodied in many different forms and for various purposes and should not be limited to the embodiments set forth herein. These embodiments are provided to make this application thorough and complete, and to fully convey the scope of this application to those skilled in the art. The embodiments described do not limit this application.

[0080] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It will be further understood that terms such as those defined in commonly used dictionaries shall be interpreted as having a meaning consistent with their meaning in the relevant field and / or the context of this specification, and shall not be interpreted in an idealized or overly formal sense unless expressly defined herein.

[0081] Example 1

[0082] Its specific implementation method is combined with the appendix Figure 1 and attached Figure 2 Please provide a detailed explanation.

[0083] Appendix Figure 1 The present invention provides a structural block diagram of an online health diagnosis system for intelligent circuit breakers, which shows the connection relationship between the environmental stress sensing and quantification module and the tolerance degradation assessment module, and marks the main functional interaction flow of each module.

[0084] Appendix Figure 2 The flowchart of an online health diagnosis method for intelligent circuit breakers provided in this embodiment of the invention illustrates the complete steps from deploying multiple types of sensors to monitor environmental parameters to obtaining degradation assessment coefficients.

[0085] In this embodiment, it includes:

[0086] The environmental stress sensing and quantification module is configured to monitor environmental parameters such as pollution, temperature and humidity by deploying multiple types of high-precision sensors, and calculate the comprehensive pollution and humidity index through calibration and feature engineering to quantify environmental stress and determine high-stress events.

[0087] Specifically, it includes:

[0088] Pollution monitoring unit: A laser scattering pollution sensor or a high-definition macro camera (1920×1080 resolution, 30fps frame rate, with anti-fog coating) is selected and installed on the upper-middle part of the windward side of the three-phase insulator of the circuit breaker (1 / 3 of the way from the top), avoiding direct sunlight and rain. The angle between the sensor probe and the insulator surface is ≤30°. The laser sensor emits a fixed-intensity laser and receives the scattering coefficient of the reflected light from the insulator surface. Combined with a preset pollution-scattering coefficient calibration curve, the equivalent salt density is calculated. This is a direct reference to existing technology and will not be elaborated upon here; the camera acquires grayscale images of the insulator surface, segments the dirty area and the clean area, calculates the dirt coverage rate, and then converts it into gray density using an empirical formula ( ). ): , , For on-site calibration coefficients, Pollution coverage rate, which is the ratio of the area of ​​the polluted region on the insulator surface to the total surface area;

[0089] The digital temperature and humidity sensors are installed inside the mechanism box (to avoid vibration); on the outer shell of the arc-extinguishing chamber (near the connection of the insulating tie rod); and inside the busbar room (≤50cm from the basin insulator).

[0090] To obtain the relevant data, temperature compensation is applied to the humidity sensor data, and sliding window mean filtering (window size 5 min) is used to eliminate random sensor drift; zero-point calibration is performed monthly using standard calibration equipment.

[0091] And calculate the comprehensive index of sewage and wetness. : ;

[0092] in,

[0093] To obtain the equivalent salt density, the laser sensor emits a laser of fixed intensity, receives the scattering coefficient of the light reflected from the insulator surface, and then calculates it by combining it with a preset pollution-scattering coefficient calibration curve.

[0094] The preset maximum permissible equivalent salt density;

[0095] Piecewise nonlinear humidity function: relative humidity When <70%, Relative humidity < 70% When ≤90%, relative humidity When >90%, ;

[0096] The rate of temperature change for a preset duration (10 minutes);

[0097] , , These are the preset weighting factors;

[0098] A high-stress event is defined as the occurrence of any of the following conditions:

[0099] Comprehensive index of sewage and wetness The value exceeds the preset threshold and continues for the preset allowed duration;

[0100] When the measured temperature is close to the dew point temperature and the humidity is ≥85%. When the condensation warning signal is triggered;

[0101] The equivalent salt density reaches the preset percentage (80%) of the maximum permissible equivalent salt density.

[0102] The electrical response synchronous monitoring module is configured to use GPS / BeiDou dual-mode synchronous technology to synchronously collect leakage current, partial discharge, and insulation temperature signals during environmental stress changes. After preprocessing, key electrical features are extracted to ensure spatiotemporal alignment of environmental and electrical data.

[0103] Specifically, it includes:

[0104] When a high-stress event occurs, the following data acquisition is triggered:

[0105] Leakage current data: The wideband Rogowski coil micro-current sensor is installed at the grounding lead of the circuit breaker insulator. The through-type installation is used to ensure that the lead is aligned with the center of the sensor to reduce electromagnetic interference.

[0106] Partial discharge data: A UHF sensor (center frequency 500MHz~1.5GHz, gain ≥20dB) and an HFCT sensor (bandwidth 1MHz~100MHz, sensitivity 1mV / mA) are used, with a dual-sensor redundancy design; the UHF sensor is adsorbed on the outside of the arc-extinguishing chamber shell (near the insulating tie rod), and the HFCT sensor is sleeved on the grounding lead; the sensor shell is grounded, and the cable uses a shielded grounded coaxial cable to resist interference;

[0107] Insulation temperature data: High-temperature adhesive is used to attach distributed optical fiber temperature sensors to the surface of the basin insulator and both ends of the insulating rod. The optical fiber route avoids the high-voltage conductor and the bending radius is ≥20mm.

[0108] The acquired data is synchronized using a GPS / BeiDou dual-mode synchronization module combined with a local crystal oscillator; and it is synchronized with the satellite clock once every hour to complete the synchronization calibration.

[0109] Each type of data is preprocessed accordingly (the preprocessing process is a direct reference to existing technology and will not be elaborated here).

[0110] Leakage current processing: Wavelet denoising is used to eliminate white noise, and the power frequency component and the effective values ​​of each harmonic are extracted, where:

[0111] Extract the fundamental RMS value, third harmonic (150Hz) RMS value, and fifth harmonic (250Hz) RMS value from the leakage current data, and calculate the total harmonic distortion rate.

[0112] Partial discharge processing: Pulse classification is performed, external interference is eliminated, and a phase-resolved spectrum (PRPD) is generated, in which:

[0113] Based on partial discharge data, with the power frequency voltage phase (0°~360°) as the horizontal axis and the discharge quantity as the vertical axis, the number of discharges in each phase interval is counted to generate a PRPD spectrum with a resolution of 256×256. The discharge initiation voltage, average discharge quantity (average amplitude of all discharge pulses), skewness (reflecting the asymmetry of the spectrum distribution), and kurtosis are extracted from the PRPD spectrum.

[0114] Insulation temperature processing: The 3σ criterion is used to remove outliers in the temperature data, and the temperature characteristics of each monitoring point on the insulation surface are calculated, including average temperature, maximum temperature rise, and temperature gradient.

[0115] The feature extraction and correlation modeling module is configured to filter core environmental and electrical features, construct a dynamic correlation model of environmental stress-electrical response through DTW time series alignment and LSTM deep learning, quantify feature sensitivity and identify early degradation indicators.

[0116] Specifically, it includes:

[0117] Environmental feature set, including the combined pollutant and wet index and its 10-minute rate of change , ;

[0118] Derived features:

[0119] Duration of continuous high humidity: humidity ≥ 80% The continuous duration (in minutes), exceeding 2 hours is counted as 2 hours;

[0120] Temperature cycle count: Count once when the daily temperature difference (highest temperature - lowest temperature) is >15℃, and reset to zero daily;

[0121] Rate of dirt accumulation: A negative value indicates the loss of filth;

[0122] Cumulative condensation duration: The cumulative duration of condensation signals within the current day;

[0123] The electrical feature set includes the fundamental RMS value, the ratio of the third harmonic to the fundamental value, the pulse peak value, the total harmonic distortion rate, the correlation dimension (reflecting the complexity of the current signal, which increases significantly with insulation degradation), and the maximum Lyapunov exponent (reflecting the chaotic characteristics of the signal, with a positive exponent indicating signal chaos) calculated through phase space reconstruction (embedding dimension m=5, delay time τ=2).

[0124] It also includes the discharge initiation voltage, average discharge quantity, PRPD spectrum skewness, kurtosis, discharge phase concentration (discharge phase concentration = number of discharges in the 10° phase interval with the most discharges / total number of discharges (C_phase will increase when insulation degrades)), maximum discharge quantity, and discharge frequency in the partial discharge characteristics;

[0125] It also includes temperature characteristics such as insulation surface temperature rise, temperature gradient, and temperature response hysteresis time: the time it takes for the insulation surface temperature to reach a stable value after a change in ambient temperature.

[0126] Time series analysis methods (such as Dynamic Time Warping (DTW) or Transfer Entropy) and machine learning models (such as LSTM networks) are used to establish a mapping from environmental features to electrical features.

[0127] The time series of environmental and electrical features are aligned using Dynamic Time Warping (DTW) (to address the issue of response lag between the two), and the similarity distance between the two series is calculated. The smaller the distance, the stronger the correlation. The similarity distance is Euclidean distance, which is a direct reference to existing technology and will not be elaborated here.

[0128] Machine learning models specifically include:

[0129] Input characteristics: comprehensive pollution and wetness index, relative humidity, ambient temperature, duration of continuous high humidity, and electrical characteristics at the first two monitoring times;

[0130] Output characteristics: Predicted values ​​of the target electrical characteristics;

[0131] Loss function: Mean Squared Error (MSE); Optimizer: Adam;

[0132] Training process:

[0133] Dataset partitioning: Historical data is divided into training set, validation set, and test set in a 7:2:1 ratio (the historical data must contain at least 100 high-stress events).

[0134] Data normalization: Min-Max normalization is used to map the feature values ​​to the [0,1] interval;

[0135] Model training: 100 iterations, using an early stopping strategy (training stops if the validation set loss does not decrease for 10 consecutive iterations).

[0136] Model evaluation: A mean absolute error (MAE) of ≤5% on the test set is considered acceptable;

[0137] Model Example: Predicting the ratio of the third harmonic to the fundamental frequency in leakage current The LSTM model;

[0138] ;in, The composite environmental stress index at the current time t; The ambient relative humidity at the current time t; The ambient temperature at the current time t; The duration of continuous high humidity at the current time t; This represents the ratio of the third harmonic to the fundamental frequency of the leakage current at the previous time point (t-1). The ratio of the third harmonic to the fundamental frequency of the leakage current at the first two time points (t-2);

[0139] The global sensitivity analysis method (Morris screening method) is used to quantify the influence of various environmental characteristics on electrical characteristics:

[0140] Calculate the sensitivity coefficient: ,in, Let be the sensitivity coefficient of the i-th environmental feature to the j-th electrical feature; The change in environmental characteristics; This corresponds to the change in electrical characteristics;

[0141] The absolute values ​​of the sensitivity coefficients of each electrical feature obtained from the calculation are summed, and the top 3 electrical features are used as the core indicators of early degradation.

[0142] A positive sensitivity coefficient indicates that the electrical characteristic value increases with increasing environmental stress, while a negative coefficient indicates a decrease, which is used to explain the insulation response mechanism;

[0143] Finally, we obtain the trained LSTM model, core feature set, sensitivity coefficient ranking table, dynamic time warping similarity matrix, electrical feature prediction values ​​under given environmental stress, and 95% confidence interval.

[0144] The tolerance degradation assessment module is configured to obtain degradation assessment coefficients from three dimensions: self-recovery ability, safety margin, and aging rate, and match the corresponding health status level based on the degradation assessment coefficients.

[0145] Specifically, it includes:

[0146] Acquire historical data on the health status of the equipment within a preset period (first 6 months) after it is put into operation, and calculate the average value of the target electrical characteristics under the corresponding environmental stress (as a health benchmark).

[0147] When environmental stress decreases, the average value of the target's electrical characteristics is collected during that period;

[0148] The difference between the mean value under environmental stress and the mean value of each target electrical characteristic during the period of environmental stress reduction is obtained. The absolute value is then divided by the mean value under environmental stress to obtain the recovery deviation value.

[0149] A preset recovery deviation threshold is set, and recovery deviation values ​​that exceed the recovery deviation threshold are recorded as abnormal recovery deviation values.

[0150] Obtain the recovery anomaly deviation values ​​corresponding to each monitoring duration during the period of environmental stress reduction. Sort the obtained recovery anomaly deviation values ​​in descending order of numerical value, extract the three largest recovery anomaly deviation values, take the average value, and record it as the recovery anomaly quantification value. Obtain the recovery anomaly quantification value corresponding to each electrical feature, and after normalizing each recovery anomaly quantification value, record the largest value as the anomaly determination quantification value.

[0151] A scientific self-recovery capability assessment system that fits the actual operating rules of equipment has been constructed, providing a precise degradation benchmark for health diagnosis;

[0152] Using the health data of the equipment in the six months before it is put into operation as a reference, it can objectively quantify the self-healing ability of the equipment after being subjected to environmental stress by calculating the recovery deviation of electrical characteristics after the environmental stress is reduced.

[0153] This comparison method based on its own health baseline avoids misjudgments caused by relying on general thresholds in traditional diagnosis. The logic of averaging the three largest abnormal deviation values ​​and normalizing to take the extreme value highlights the degradation impact of key electrical characteristics (in line with the equipment health bottleneck effect) and eliminates the evaluation interference caused by the difference in the dimensions of different characteristics, making the abnormal judgment of self-recovery capability more targeted.

[0154] It can be achieved without adding additional sensors or complex detection equipment, simply by comparing the time series of existing monitoring data, thus reducing implementation costs. It accurately captures the latent degradation signals that cannot be restored to a healthy state after stress impact by restoring abnormal quantification values. These signals often precede the obvious manifestation of equipment failure, providing maintenance personnel with a basis for early intervention, avoiding the expansion of faults due to the continuous decline in self-recovery capability, significantly improving the reliability of smart circuit breaker operation, and reducing losses caused by unplanned downtime.

[0155] Take the device's health status, different The corresponding target electrical characteristic values ​​are fitted with the 95% confidence upper limit and 95% confidence lower limit of the electrical characteristics using kernel density estimation to form a normal range;

[0156] Calculate unsafe values: ,in, The upper limit of confidence is 95%. This is the 95% confidence lower limit; This is the actual measured value of the electrical characteristic at the current moment; A function that takes the minimum value;

[0157] Obtain the unsafe values ​​of each electrical feature, preset the unsafe value thresholds for each electrical feature, mark the unsafe values ​​of electrical features that are greater than the preset unsafe value thresholds, and calculate the difference between the unsafe value and the unsafe value threshold to obtain the unsafe difference.

[0158] After normalizing the unsafe differences corresponding to each electrical characteristic, the values ​​are sorted in descending order to obtain a normalized set. The time points corresponding to each value and the time differences between each adjacent value are extracted from the set. A time difference threshold is preset, and time differences less than the time difference threshold are recorded as abnormal time differences.

[0159] Obtain all outlier time differences, extract the smallest outlier time difference, and take its reciprocal to obtain the insecurity quantification value.

[0160] Kernel density estimation is used to fit the 95% confidence interval of electrical characteristics under the health state of equipment, replacing the traditional fixed threshold standard. This can fully adapt to the individual differences of different equipment and the special characteristics of the operating environment. The deviation of the actual measured value from the normal range is accurately quantified by the unsafe value formula. Then, the normalization process is used to eliminate the dimensional differences of different electrical characteristics, ensuring that the safety risks of each characteristic can be compared horizontally. The logic of screening out-of-threshold unsafe differences and analyzing adjacent time differences can effectively focus on key risk points and avoid redundant data from interfering with the assessment results.

[0161] By extracting the minimum abnormal time difference and taking its reciprocal, an unsafe quantitative value can be obtained, which can intuitively reflect the degree of concentrated outbreak of safety hazards.

[0162] The smaller the time difference, the more concentrated the risk; the larger the reciprocal, the higher the level of insecurity. This allows maintenance personnel to quickly determine the urgency of the risk. The entire assessment process relies solely on existing monitoring data and algorithm optimization, without the need for additional sensors or detection equipment. It has low implementation costs, strong compatibility, and can accurately capture hidden risks where electrical characteristics are close to the safety boundary but have not exceeded the limit. This provides a reliable basis for taking protective measures in advance and preventing failures caused by a continuous decline in safety margin.

[0163] Obtain the monthly average values ​​of electrical characteristics within a preset time period, based on the exponential law of insulation aging: Taking the logarithm of both sides, we get: The aging slope was fitted using linear regression. ,in, The target electrical characteristic value at time t; The initial values ​​of the target electrical characteristics at the initial moment;

[0164] Obtain the aging slope of each electrical feature. , and the various aging slopes After normalization, calculate each aging slope. The difference between the corresponding maximum permissible aging slope is taken as the absolute value to obtain the aging difference value corresponding to each electrical characteristic, and the largest aging difference value is recorded as the aging quantification value.

[0165] After normalizing the anomaly determination quantification value, the unsafe quantification value, and the aging quantification value, a weighted summation is performed to obtain the degradation assessment coefficient.

[0166] The weighted summation process includes:

[0167] The weighting factors for the quantification values ​​of anomaly determination, insecurity, and aging are preset. The quantification values ​​of anomaly determination, insecurity, and aging are multiplied by their corresponding weighting factors, and the sum is used to obtain the degradation evaluation coefficient.

[0168] Based on the exponential law of insulation aging, by fitting the aging slope through the monthly average of electrical characteristics, the gradual change trend of equipment aging can be objectively captured.

[0169] This quantitative method, which closely aligns with the physical nature of insulation degradation, can identify hidden risks of accelerated aging even when the threshold is not exceeded, compared to the crude judgment of traditional threshold exceeding the standard and triggering an alarm.

[0170] Meanwhile, by normalizing the dimensional differences of different electrical characteristics, the severity of aging is quantified by the absolute value of the difference between the value and the maximum allowable aging slope, and the maximum value is taken as the aging quantification value. This not only highlights the dominant role of key characteristics in aging, but also ensures the pertinence and objectivity of the assessment results.

[0171] By using a weighted fusion of three-dimensional quantitative values, a comprehensive assessment of health status was achieved, significantly improving the comprehensiveness and practicality of diagnostic results.

[0172] By performing a second normalization and weighted summation on self-recovery capability (quantified value of anomaly determination), safety margin (quantified value of unsafety), and aging rate (quantified value of aging), the system takes into account both the current operational risks of the equipment and the long-term degradation trend and recovery capability after stress impact, avoiding misjudgments or omissions caused by single-dimensional assessment. Furthermore, the weighting factors can be flexibly adjusted according to actual scenarios such as circuit breaker operating environment and voltage level, making it highly adaptable. The final degradation assessment coefficient can accurately match the health status level, providing maintenance personnel with a clear basis for when to maintain and what the maintenance focus should be, effectively reducing the risk of unplanned downtime and the cost of blind maintenance.

[0173] Example 2

[0174] Please see Figure 2 An online health diagnosis method for intelligent circuit breakers includes the following components:

[0175] Deploy multiple types of high-precision sensors to monitor environmental parameters such as pollution, temperature, and humidity. After calibration and feature engineering, calculate the comprehensive pollution and humidity index to complete the quantification of environmental stress and the determination of high-stress events.

[0176] Based on GPS / BeiDou dual-mode synchronization technology, leakage current, partial discharge, and insulation temperature signals are simultaneously collected during high-stress events, and key electrical features are extracted after preprocessing.

[0177] We screened the core environmental and electrical features, constructed a dynamic correlation model using DTW time series alignment and LSTM deep learning, quantified feature sensitivity, and identified early degradation indicators.

[0178] Using the equipment health baseline as a reference, the quantitative values ​​of the three dimensions of self-recovery capability, safety margin, and aging rate are calculated respectively to complete the tolerance assessment of each dimension;

[0179] The degradation evaluation coefficient is obtained by normalizing the three-dimensional quantified values ​​and then weighting and summing them. Based on this coefficient, the health status level of the smart circuit breaker is matched.

[0180] Example 3

[0181] An online health diagnostic device for intelligent circuit breakers includes: a housing, and an online health diagnostic system for intelligent circuit breakers integrated within the housing.

[0182] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0183] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0184] It should be noted that, in this document, the use of relational terms such as "first" and "second" is merely for distinguishing one entity or operation from another, and does not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0185] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0186] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0187] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0188] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0189] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0190] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0191] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. An online health diagnostic system for intelligent circuit breakers, characterized in that, include: The environmental stress sensing and quantification module is configured to monitor environmental parameters by deploying multiple types of sensors, calculate the comprehensive pollution and wet index through calibration and feature engineering, and quantify environmental stress and determine high-stress events. The electrical response synchronous monitoring module is configured to use GPS / BeiDou dual-mode synchronous technology to synchronously collect leakage current, partial discharge, and insulation temperature signals during environmental stress changes. After preprocessing, key electrical features are extracted to ensure spatiotemporal alignment of environmental and electrical data. The feature extraction and correlation modeling module is configured to filter core environmental and electrical features, construct a dynamic correlation model of environmental stress and electrical response, quantify feature sensitivity, and identify early degradation indicators. The tolerance degradation assessment module is configured to obtain degradation assessment coefficients from three dimensions: self-recovery ability, safety margin, and aging rate, and match corresponding health status levels based on these coefficients. Specifically, this includes: Acquire historical data on the health status of the equipment within a preset time period after it is put into operation, and calculate the mean value of the target electrical characteristics under the corresponding environmental stress. When environmental stress decreases, the average value of the target's electrical characteristics is collected within the corresponding time period; The difference between the mean value of the target electrical feature under the environmental stress corresponding to the healthy state and the mean value of each target electrical feature during the period of reduced environmental stress is obtained. The absolute value is then divided by the mean value under the environmental stress corresponding to the healthy state to obtain the recovery deviation value. A preset recovery deviation threshold is set, and recovery deviation values ​​that exceed the recovery deviation threshold are recorded as abnormal recovery deviation values. Obtain the recovery anomaly deviation value corresponding to each monitoring duration during the period of environmental stress reduction. Sort the obtained recovery anomaly deviation values ​​in descending order of numerical value, extract the three largest recovery anomaly deviation values, take the average value and record it as the recovery anomaly quantification value. Obtain the recovery anomaly quantification value corresponding to each electrical feature, and after normalizing each recovery anomaly quantification value, record the largest value as the anomaly determination quantification value. Also includes: Under the condition of equipment health, different comprehensive pollution and wetness indices The corresponding target electrical characteristic values ​​are used to estimate the upper and lower confidence limits of the fitted electrical characteristics using kernel density estimation, thus forming a normal range; Calculate unsafe values: ,in, The upper limit of confidence is 95%. This is the 95% confidence lower limit; This is the actual measured value of the electrical characteristic at the current moment; A function that takes the minimum value; Obtain the unsafe values ​​of each electrical feature, preset the unsafe value thresholds for each electrical feature, mark the unsafe values ​​of electrical features that are greater than the preset unsafe value thresholds, and calculate the difference between the unsafe value and the unsafe value threshold to obtain the unsafe difference. After normalizing the unsafe differences corresponding to each electrical characteristic, the values ​​are sorted in descending order to obtain a normalized set. The time points corresponding to each value and the time differences between each adjacent value are extracted from the set. A time difference threshold is preset, and time differences less than the time difference threshold are recorded as abnormal time differences. Obtain all outlier time differences, extract the smallest outlier time difference from them, and take its reciprocal to obtain the insecurity quantification value; Also includes: Obtain the monthly average values ​​of electrical characteristics within a preset time period, based on the exponential law of insulation aging: Taking the logarithm of both sides, we get: The aging slope was fitted using linear regression. ,in, The target electrical characteristic value at time t; The initial values ​​of the target electrical characteristics at the initial moment; Obtain the aging slope of each electrical feature. , and the various aging slopes After normalization, calculate each aging slope. The difference between the corresponding maximum permissible aging slope is taken as the absolute value to obtain the aging difference value corresponding to each electrical characteristic, and the largest aging difference value is recorded as the aging quantification value. After normalizing the abnormality determination quantification value, the unsafety quantification value, and the aging quantification value, a weighted summation is performed to obtain the degradation assessment coefficient.

2. The intelligent circuit breaker health online diagnostic system according to claim 1, characterized in that, The environmental stress sensing and quantification module specifically includes: A laser scattering type pollution sensor is selected and installed on the windward side of the three-phase insulator of the circuit breaker. Digital temperature and humidity sensors are installed inside the mechanism housing; the outer casing of the arc-extinguishing chamber; and inside the busbar chamber. To obtain the relevant data and calibrate the humidity sensor data; And calculate the comprehensive index of sewage and wetness. : ; in, For equivalent salt density; The preset maximum permissible equivalent salt density; It is a piecewise nonlinear humidity function; The rate of temperature change over a preset duration; , , These are the preset weighting factors; A high-stress event is defined as the occurrence of any of the following conditions: Comprehensive index of sewage and wetland The value exceeds the preset threshold and continues for the preset allowed duration; When the condensation warning signal is triggered; The equivalent salt density reaches the preset percentage of the maximum permissible equivalent salt density.

3. The intelligent circuit breaker health online diagnostic system according to claim 1, characterized in that, The electrical response synchronous monitoring module specifically includes: When a high-stress event occurs, the following data acquisition is triggered: Leakage current data: A wideband Rogowski coil micro-current sensor is installed at the grounding lead of the circuit breaker insulator to obtain the data. Partial discharge data: UHF sensor and HFCT sensor are used; the UHF sensor is adsorbed on the outside of the arc-extinguishing chamber shell, and the HFCT sensor is sleeved on the grounding lead; the sensor shell is grounded; Insulation temperature data: Distributed fiber optic temperature sensors are attached to the surface of the basin insulator and both ends of the insulating tie rod; The acquired data is synchronized, and each type of data is preprocessed accordingly. Leakage current processing: Wavelet denoising is used to eliminate white noise, and the power frequency component and the effective values ​​of each harmonic are extracted, where: Partial discharge processing: pulse classification, elimination of external interference, and generation of phase-resolved spectra; Outliers in the temperature data are removed, and the temperature characteristics of each monitoring point on the insulation surface are calculated.

4. The intelligent circuit breaker health online diagnostic system according to claim 1, characterized in that, The feature extraction and association modeling module specifically includes: Environmental feature set, including the combined pollutant and wet index and its 10-minute rate of change Duration of continuous high humidity, number of temperature cycles, rate of dirt accumulation, and cumulative condensation time. The electrical characteristic set includes the fundamental RMS value, the ratio of the third harmonic to the fundamental value, the pulse peak value, the total harmonic distortion rate, the correlation dimension, and the maximum Lyapunov exponent in the leakage current characteristics. It also includes the discharge initiation voltage, average discharge quantity, PRPD spectrum skewness, kurtosis, discharge phase concentration, maximum discharge quantity, and discharge frequency in the partial discharge characteristics; It also includes temperature characteristics such as insulation surface temperature rise, temperature gradient, and temperature response hysteresis time.

5. The intelligent circuit breaker health online diagnostic system according to claim 4, characterized in that, Also includes: A mapping from environmental characteristics to electrical characteristics is established using time series analysis and machine learning models. Align time series of environmental and electrical characteristics through dynamic time warping; A global sensitivity analysis method is used to quantify the impact of various environmental characteristics on electrical characteristics: Calculate the sensitivity coefficients; sum the absolute values ​​of the calculated sensitivity coefficients for each electrical feature, and use the top 3 electrical features as the core indicators of early degradation. Finally, we obtain the trained LSTM model, core feature set, sensitivity coefficient ranking table, similarity matrix, and electrical feature prediction values ​​under given environmental stress.

6. A method for online health diagnosis of an intelligent circuit breaker, comprising the online health diagnosis system for an intelligent circuit breaker according to claim 1, characterized in that, include: Deploy multiple types of sensors to monitor environmental parameters, calculate the comprehensive pollution and wet index through calibration and feature engineering, and complete the quantification of environmental stress and the determination of high-stress events; During high-stress events, leakage current, partial discharge, and insulation temperature signals are simultaneously acquired, and key electrical characteristics are extracted after preprocessing. We screened the core environmental and electrical features, constructed a dynamic correlation model using DTW time series alignment and LSTM deep learning, quantified feature sensitivity, and identified early degradation indicators. Using the equipment health baseline as a reference, the quantitative values ​​of the three dimensions of self-recovery capability, safety margin, and aging rate are calculated respectively to complete the tolerance assessment of each dimension; The degradation evaluation coefficient is obtained by normalizing the three-dimensional quantified values ​​and then weighting and summing them. Based on this coefficient, the health status level of the smart circuit breaker is matched.

7. An online health diagnostic device for intelligent circuit breakers, characterized in that, It includes: a housing, and an online health diagnostic system for a smart circuit breaker as described in any one of claims 1-5 integrated within the housing.

Images