Self-diagnosis high-voltage fuse and running state online monitoring method thereof

Abstract

The invention discloses a self-diagnosis high-voltage fuse and a running state online monitoring method thereof, and belongs to the technical field of high-voltage power equipment.The method comprises the following steps that S1, running parameters and environmental parameters of a fuse body are collected, and collected data are corrected in real time based on calibration factors; s2, filtering and denoising the calibrated data, and extracting core feature parameters; s3, calculating a state evaluation index based on a multi-parameter coupling state evaluation formula, constructing a dynamic threshold self-optimization model in combination with a simulation result of the digital twin model, adjusting a state judgment threshold in real time, and adaptively matching a typical fault mode; s4, calculating the residual life of the fuse link body through a dynamic aging prediction formula, and quantifying the fault risk level; and S5, performing graded response and self-diagnosis feedback based on a calculation result. According to the invention, the technical defects of single monitoring parameter, low diagnosis precision and lack of self-repairing linkage in the prior art are overcome, and the operation safety and reliability of a power system are improved.

Description

Technical Field

[0001] This invention relates to the field of high-voltage power equipment technology, and in particular to a self-diagnostic high-voltage fuse and a method for online monitoring of its operating status. Background Technology

[0002] High-voltage fuses are critical short-circuit and overload protection devices in power systems, and their operating status directly affects the safety and stability of the power system. Most existing high-voltage fuses are of traditional structure, lacking effective online monitoring and self-diagnostic functions. Faults can only be detected through manual inspection after a fuse has blown, resulting in problems such as delayed fault identification, high maintenance costs, and a significant risk of unplanned power outages.

[0003] Currently, although some improved high-voltage fuses have introduced simple monitoring functions, such as monitoring operating current through current transformers and monitoring surface temperature through temperature sensors, they still have the following technical shortcomings: First, the monitoring parameters are singular, focusing only on electrical or thermal parameters and ignoring the comprehensive impact of multiple factors such as strain, harmonics, and environment on the operating status of fuses, resulting in low diagnostic accuracy and a tendency to make misjudgments or omissions. Secondly, the lack of an adaptive calibration mechanism makes the sensor accuracy susceptible to environmental factors, leading to a decrease in the reliability of monitoring data after long-term operation. Third, the linkage between monitoring and self-repair has not been realized; it can only passively alarm and cannot proactively pre-process potential faults. Fourth, existing condition assessment formulas mostly use fixed weights or single parameters for judgment, which cannot adapt to complex and ever-changing operating conditions and make it difficult to accurately quantify the aging degree and remaining life of fuses.

[0004] Therefore, developing a high-voltage fuse and online monitoring method that integrates multi-dimensional monitoring, adaptive self-diagnosis, dynamic life prediction and self-repair linkage functions has become an urgent need in the current power equipment field. Summary of the Invention

[0005] The purpose of this invention is to provide a self-diagnostic high-voltage fuse and its online monitoring method for operating status, which can realize synchronous monitoring of multi-dimensional parameters of high-voltage fuses, adaptive and accurate diagnosis, dynamic remaining life prediction and graded response, so as to solve the technical defects of existing technologies such as single monitoring parameters, low diagnostic accuracy and lack of self-repair linkage, and improve the safety and reliability of power system operation.

[0006] To achieve the above objectives, the present invention provides a self-diagnostic high-voltage fuse, comprising a fuse body made of a novel self-healing alloy material, with a micro-nano level conductive repair medium embedded inside; The self-diagnostic integrated module is integrated into the end caps at both ends of the fuse body. It includes a physical field sensor unit, a data processing unit, a digital twin virtual modeling unit, a self-calibration unit, a wireless communication unit, and a self-repair trigger unit. The physical field sensor unit synchronously collects real-time operating current voltages, body temperature, surface strain, and environmental parameters of the fuse. The data processing unit preprocesses and extracts features from the collected data. The digital twin virtual modeling unit constructs a dynamic simulation model of the fuse based on real-time data. The self-calibration unit dynamically corrects the sensor accuracy and diagnostic threshold using a backpropagation algorithm. Auxiliary execution module: electrically connected to the self-diagnostic integrated module, including a miniature heat dissipation module, an arc suppression device, and a status indication unit.

[0007] Preferably, the novel self-healing alloy material used in the fuse body is based on a copper-silver alloy and reinforced with nano-alumina; the micro-nano conductive repair medium is a micro-nano bismuth-tin alloy medium with a melting point of 120-150℃ and a solid insulating state below 80℃.

[0008] Preferably, the physical field sensor unit includes a Rogowski coil current sensor, a fiber optic grating temperature sensor, a miniature strain sensor, a voltage sensor, and an environmental sensor.

[0009] This invention also provides an online monitoring method for the operating status of a self-diagnostic high-voltage fuse, comprising the following steps: S1. Multi-dimensional data acquisition and synchronous calibration: The physical field sensor unit acquires the operating parameters and environmental parameters of the fuse body at the sampling frequency, and the self-calibration unit corrects the acquired data in real time based on the calibration factor K. S2. Data preprocessing and multi-factor feature extraction: The calibrated data is filtered and denoised to extract the current change rate, temperature change rate, voltage harmonic components, equivalent resistance and strain accumulation. S3. Adaptive Operation Status Diagnosis: Based on the multi-parameter coupled status assessment formula, the status assessment index S is calculated. Combined with the simulation results of the digital twin model, the full life cycle data of the fuse body is introduced to construct a dynamic threshold self-optimization model, which adjusts the status judgment threshold in real time and adaptively matches typical fault modes to perform accurate status diagnosis and fault type identification. S4. Remaining life dynamic prediction: The remaining life of the fuse body is calculated by using a dynamic aging prediction formula to quantify the failure risk level. S5. Graded Response and Self-Diagnosis Feedback: Based on the status diagnosis results and remaining lifetime prediction values, trigger corresponding early warning, self-repair preprocessing, or fault isolation actions, and feed back the diagnosis results to the power system monitoring platform.

[0010] Preferably, in S1, an adaptive calibration algorithm is used to correct the collected data in real time, and the formula for calculating the calibration factor K is: ; Wherein, K0 is the standard calibration coefficient, with a value ranging from 0.98 to 1.02; This is the temperature influence coefficient. Humidity influence coefficient The dust influence coefficient is given by T, where T is the fuse body temperature, T0 is the standard calibration temperature, H is the ambient relative humidity, and D is the ambient dust concentration; the calibrated data X... cal =X raw ×K, X raw For the raw data collected by the sensor, X cal This is the calibrated data.

[0011] Preferably, in S3, the multi-parameter coupling state evaluation formula is: ; Where S is the state assessment index. All are dynamic weighting coefficients, satisfying It is obtained through optimization using a genetic algorithm, and responds to parameter changes in real time; I is the real-time operating current, I N T is the rated current of the fuse, and T is the temperature of the fuse body. env For ambient temperature, T max R is the maximum allowable operating temperature of the fuse. eq R0 is the equivalent resistance of the fuse body, and R0 is the initial resistance of the fuse body. For cumulative strain, U represents the maximum allowable strain of the fuse body. h U is the total effective value of voltage harmonics. N This is the rated voltage of the fuse.

[0012] Preferably, the optimization formula for the dynamic threshold self-optimization model is: ; Among them, T std As the standard threshold, k life This is the lifespan degradation impact coefficient. k represents the percentage of cumulative runtime. work The severity coefficient of the working conditions. Rate the severity of the working conditions; The optimized normal threshold T is finally calculated based on the dynamic threshold self-optimization model. opt1 Optimized potential fault threshold T opt2 And the optimized failure threshold T opt3 ; The dynamic threshold determination rule is: S≤Topt1 is in the normal state, T opt1 <S≤T opt2 is in the potential fault state, T opt2 <S≤T opt3 is in the critical fusing state, S>T opt3 is in the failure state; at the same time, the fault type is adaptively matched based on the fault mode library.

[0013] Preferably, in S4, the dynamic aging prediction formula is: ; where, L r is the remaining life, L0 is the rated service life of the fuse, is the fault type correction coefficient, k1 is the current aging coefficient, k2 is the temperature aging coefficient, k3 is the self-repair times influence coefficient, t is the cumulative operation time, n is the self-repair trigger times, is the cumulative effect of current overload, is the cumulative effect of temperature rise.

[0014] Preferably, the state evaluation index S in S3 and the digital twin model simulation life L sim are introduced for error correction, and the correction formula is: ; where, L is the finally corrected remaining life.

[0015] Preferably, the determination rule of the fault risk level in S5 is: S≤0.3 or L>2000h is determined as the normal state: the self-diagnosis module enters the low-power monitoring mode, reduces the sampling frequency to 100Hz, uploads the status data to the monitoring platform through the wireless communication unit every 30 minutes, and the auxiliary execution module is in the standby state; 0.3<S≤0.6 or 1000h<L≤2000h is determined as the potential fault state: trigger a first-level warning, the LED light of the status indication unit flashes yellow, start the micro heat dissipation module to reduce the temperature of the fuse body, increase the sampling frequency to 500Hz, and upload the status data every 5 minutes to track the change trend of parameters in real time; 0.6<S≤0.9 or 500h<L≤1000h is determined as the critical fusing state: trigger a second-level warning, the LED light of the status indication unit flashes red, start the arc suppression device to prevent the arc from spreading, synchronously send a fault warning message to the monitoring platform, increase the sampling frequency to 1kHz, and upload the status data every 1 minute to remind the operation and maintenance personnel to deal with it in time; If S>0.9 or L≤500h, it is considered a failure state: triggering a level three warning, the status indicator unit LED light turns red and stays on, the self-repair trigger unit starts self-repair, the micro-nano level conductive repair medium is melted by heating with an electric heating wire, filling the fused gap to achieve temporary conductive repair, and at the same time sending an emergency trip request to the monitoring platform to cooperate with the power system to achieve fault isolation.

[0016] Therefore, the beneficial effects of the above-mentioned self-diagnostic high-voltage fuse and its online monitoring method for operating status are as follows: (1) It realizes the integrated design of monitoring, diagnosis, prediction and repair, solves the technical defects of traditional fuses with simple structure and single function, and significantly improves the intelligence level of fuses; (2) It breaks through the limitations of existing technologies that monitor only a single parameter, provides comprehensive data support for accurate diagnosis, and reduces the risk of misjudgment and missed judgment; (3) It solves the shortcomings of traditional fixed threshold diagnosis, such as poor adaptability, ambiguous fault identification, and large life prediction error. The diagnostic accuracy is greatly improved and the remaining life prediction error is significantly reduced. (4) A graded response mechanism has been established, realizing proactive response throughout the entire process from early warning and pre-processing to self-repair and fault isolation. This avoids the limitation that traditional fuses can only passively blow out, buys time for power system fault handling, and reduces the risk of unplanned power outages.

[0017] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the structure of a self-diagnostic high-voltage fuse according to the present invention; Figure 2 This is a schematic diagram of the composition of the self-diagnostic integrated module of the present invention; Figure 3 This is a schematic diagram illustrating the steps of an online monitoring method for the operating status of a self-diagnostic high-voltage fuse according to the present invention.

[0019] Figure Labels 1. Fuse body; 2. Self-diagnostic integrated module; 31. Miniature heat dissipation module; 32. Arc suppression device; 33. Status indication unit. Detailed Implementation

[0020] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0021] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0022] Example 1: like Figure 1 and Figure 2 As shown, this embodiment provides a self-diagnostic high-voltage fuse, including: Fuse body 1: Made of copper-silver alloy as the base material, with added nano-alumina reinforcing phase to improve the mechanical strength and high temperature resistance of the fuse body; embedded with micro-nano bismuth-tin alloy repair medium, the repair medium has a melting point of 120-150℃ and is in a solid insulating state at the rated operating temperature (≤80℃), which does not affect the normal conductivity of the fuse; when an overload or short circuit occurs, the high temperature of the arc causes the repair medium to melt rapidly, filling the fuse gap to form a temporary conductive path, realizing self-repair and buying time for fault handling.

[0023] The self-diagnostic integrated module 2, as the core control unit, is integrated into the end caps at both ends of the fuse body 1. It adopts a waterproof, dustproof, and high-temperature resistant sealed structure design, including: The physical field sensor unit includes a Rogowski coil current sensor (measurement range 0-I×2, accuracy ±0.5%), a fiber optic grating temperature sensor (measurement range -40℃-200℃, accuracy ±0.1℃), a miniature strain sensor (measurement range 0-5000μm / m, accuracy ±1μm / m), a voltage sensor (measurement range 0-U×1.2, accuracy ±0.5%), and an environmental sensor (measures temperature, humidity, and dust concentration), enabling synchronous acquisition of multi-dimensional parameters. Data processing unit: It adopts an ARM Cortex-M7 processor with a main frequency of 216MHz, supports multi-channel parallel data processing, and completes data filtering, amplification, analog-to-digital conversion and feature extraction. Digital twin virtual modeling unit: Based on ANSYS finite element analysis software, a digital model of the fuse is constructed. The model includes the coupling relationship of temperature field, electric field and stress field. Data is collected by sensors in real time to dynamically simulate the state evolution of the fuse under different working conditions and predict potential faults. Self-calibration unit: Based on the backpropagation neural network algorithm, it dynamically corrects the calibration factor by comparing standard reference data with sensor-acquired data, and compensates for the impact of environmental factors and sensor aging on monitoring accuracy. Wireless communication unit: Adopting a 5G industrial module, it supports low latency and high reliability data transmission, and uploads diagnostic results and monitoring data to the power system monitoring platform; Self-repair trigger unit: Receives instructions from the data processing unit and triggers the melting of the repair medium by heating with a heating wire to achieve self-repair function.

[0024] Auxiliary execution module: includes a miniature heat dissipation module 31 (using a semiconductor cooling chip, symmetrically installed on both sides of the fuse body 1, attached to the outer shell of the fuse body 1, with a cooling power of 5-10W), an arc suppression device 32 (based on a metal oxide varistor, limiting the arc voltage, integrated inside the end cap of the fuse body 1, and directly connected to the electrodes of the fuse body 1), and a status indication unit 33 (embedded in the outer shell surface of the fuse body 1, making it easy for maintenance personnel to observe, and distinguishing the operating status by LED light color: green - normal, yellow - potential fault, red - critical / failure), which is electrically connected to the self-diagnostic integration module 2 and executes corresponding control actions.

[0025] like Figure 3 As shown, the online monitoring method for the operating status of the self-diagnostic high-voltage fuse includes the following steps: S1. Multi-dimensional data acquisition and synchronous calibration: The physical field sensor unit acquires the operating parameters and environmental parameters of the fuse body 1 at the sampling frequency, and the self-calibration unit corrects the acquired data in real time based on the calibration factor K.

[0026] An adaptive calibration algorithm is used to correct the acquired data in real time. The formula for calculating the calibration factor K is as follows: Wherein, K0 is the standard calibration coefficient, with a value ranging from 0.98 to 1.02; The temperature influence coefficient is 3.2 × 10⁻⁶. -4 / ℃), The humidity influence coefficient is 1.8 × 10⁻⁶. -5 / RH%), The dust impact coefficient is 2.5 × 10⁻⁶. -6 / mg·m -3T represents the fuse body temperature (°C), T0 represents the standard calibration temperature (25°C), H represents the ambient relative humidity (RH%), and D represents the ambient dust concentration (mg / m³). 3 ); Calibrated data X cal =X raw ×K, X raw For the raw data collected by the sensor, X cal This is the calibrated data.

[0027] S2. Data Preprocessing and Multi-Factor Feature Extraction: The calibrated data is filtered and denoised to extract core feature parameters such as current change rate, temperature change rate, voltage harmonic components, equivalent resistance, and strain accumulation. This includes the following steps: S21. Data Preprocessing: The preprocessing process adheres to the principles of preserving valid signals, suppressing interference noise, and unifying data dimensions. It employs a three-level processing mechanism, with all operations based on an ARM Cortex-M7 processor (216MHz) for multi-channel parallel processing. Processing latency is ≤10ms, ensuring real-time performance. S211. Data Cleaning (Outlier Removal): The system receives the calibrated data sequence output by S1 (sampling frequency 100Hz-1kHz, data format 16-bit binary). First, outliers are removed based on the 3σ criterion (normal distribution anomaly determination): the mean μ and standard deviation σ of each parameter (I, U, T, ε, H, D, etc.) are calculated, and data exceeding the range of [μ-3σ, μ+3σ] are determined to be outliers. For abnormal fluctuations of 3 or more consecutive sampling points (such as sudden rises and falls in current, sudden temperature changes exceeding 5℃ / 10ms), linear interpolation (based on 2 valid sampling points before and after the abnormal segment) is used to complete the data, so as to avoid local anomalies affecting the overall feature accuracy.

[0028] S212, Filtering and denoising (interference suppression): To address electromagnetic interference (50Hz power frequency interference and high-frequency harmonic interference) and inherent sensor noise in high-voltage power distribution scenarios, an adaptive Kalman filter combined with wavelet threshold denoising algorithm is adopted. The signal-to-noise ratio (SNR) of the filtered data is ≥35dB, ensuring the integrity of the effective signal.

[0029] S213. Data normalization (dimensional unification): Since the physical dimensions and numerical ranges of each characteristic parameter differ greatly (e.g., the unit of current I is A, and the unit of strain ε is μm / m), in order to avoid the difference in numerical magnitude affecting the subsequent weight optimization and diagnostic accuracy, the min-max normalization algorithm is used to map all preprocessed data to the interval [0, 1].

[0030] S22, Multi-factor Feature Extraction: Based on the preprocessed normalized data and combined with the physical characteristics of fuse operation (current thermal effect, mechanical strain effect, electrical harmonic effect, etc.), five types of core feature parameters are extracted. All features are output in the form of a "10-minute time series" (corresponding to the sampling frequency, containing 6000-60000 feature points). The extraction methods for each feature are as follows: Current change rate (di / dt, unit: A / ms): Characterizes the drastic change in current within the fuse body, and is a core feature for identifying initial overload / short circuit faults. Extraction logic: Based on calibrated current data (I) at two consecutive sampling times. k I k-1 The ratio of the current difference between adjacent moments to the time interval is calculated using the sampling time interval Δt (Δt = 10ms when sampling frequency f = 100Hz, Δt = 1ms when f = 1kHz). The formula is: To avoid the impact of instantaneous fluctuations, the calculation results are smoothed by a 3-point moving average, and the size of the smoothing window is adaptively adjusted (the window size is 5 when the fluctuations are severe and 3 when the fluctuations are stable).

[0031] Temperature change rate (dT / dt, unit: ℃ / s): Characterizes the rate at which the temperature of the fuse body rises / falls, reflecting the balance between the current heating effect and the heat dissipation effect. Extraction logic: Based on the calibrated body temperature data (T) collected by the fiber Bragg grating temperature sensor. k T k-1 Using a time interval Δt consistent with di / dt, the ratio of the temperature difference to the time interval (converted to change per second) is calculated using the following formula: Since Δt is in milliseconds (ms), multiplying it by 1000 converts it to seconds (s), while also eliminating sudden changes in ambient temperature (T). env The interference caused by fluctuations (>2℃ / s) is corrected using the ambient temperature compensation coefficient.

[0032] Voltage harmonic components (U h (Unit: V): Characterizes the degree of harmonic pollution of the voltage across the fuse body. Excessive harmonic content will accelerate the aging of the fuse. Extraction logic: Extract the calibrated voltage data across the two ends (U k Perform a Fast Fourier Transform (FFT) to decompose the harmonics into 1-50th harmonics. Based on the GB / T 14549-1993 standard, calculate the square root of the sum of the squares of the effective values ​​of the 2nd-19th harmonics (ignoring the 20th and higher harmonics, whose amplitudes are less than 0.1% of the fundamental frequency). The formula is as follows: in, The effective value of the nth harmonic is given. The number of sampling points for the FFT transformation is 1024, and the frequency resolution is ≤0.5Hz to ensure the accuracy of harmonic component calculation.

[0033] Equivalent resistance (R) eq (Unit: Ω): Characterizes the degree of degradation of the conductivity of the fuse body. Increased resistance is the core characteristic of aging and poor contact. Extraction logic: Based on Ohm's law, combined with the calibrated current (I) at the same sampling time. k ) and the voltage across the terminals (U) k ), minus voltage harmonic components U h The effect of the equivalent resistance is calculated using the following formula: To avoid resistance anomalies caused by instantaneous fluctuations in current and voltage, the average of 10 sampling points is used as the current resistance value (R). eq The value is compared with the initial resistance R0 at the factory, and the resistance decay rate (R0) is calculated. eq -R0) / R0) is used as an auxiliary feature.

[0034] Cumulative strain (Δε, unit: μm / m): Characterizes the cumulative deformation of the fuse body under mechanical stress (thermal expansion, installation stress). Excessive deformation can easily lead to fuse fracture. Extraction logic: Based on the calibrated surface strain data (ε... k The integral is performed with a statistical period of 10 minutes. The cumulative difference between the strain data and the initial strain (ε0, factory calibration value) within the period is calculated using the following formula: Where N is the number of sampling points within 10 minutes (N=30000 when f=500Hz), and the strain accumulation threshold (ε) is set simultaneously. lim When the threshold of 80% is reached, an early warning is triggered, indicating an excessive risk of overreaction.

[0035] S23. Feature Output and Verification: The five core feature parameters (di / dt, dT / dt, U) have been extracted. h R eq Both Δε and Δε are output in the form of "time series + statistical feature values" (mean, maximum value, variance) and stored in the full life cycle database. At the same time, redundant features are removed by feature correlation analysis (Pearson correlation coefficient) (one feature with correlation > 0.95 is retained) to ensure that the feature set input to the self-diagnosis integrated module 2 is independent and representative. The overall accuracy of feature extraction is ≥ 98%, which provides accurate data support for subsequent dynamic threshold self-optimization and fault mode matching.

[0036] S3. Adaptive Operation Status Diagnosis: Based on the multi-parameter coupled status assessment formula, the status assessment index S is calculated. Combined with the simulation results of the digital twin model, the full life cycle data of the fuse body is introduced to construct a dynamic threshold self-optimization model, which adjusts the status judgment threshold in real time and adaptively matches typical fault modes to perform accurate status diagnosis and fault type identification.

[0037] The multi-parameter coupled state evaluation formula is as follows: Where S is the state assessment index. All are dynamic weighting coefficients, satisfying It is obtained through optimization using a genetic algorithm, and responds to parameter changes in real time; I is the real-time operating current, I N T is the rated current of the fuse, and T is the temperature of the fuse body. env For ambient temperature, T max R is the maximum allowable operating temperature of the fuse. eq R0 is the equivalent resistance of the fuse body, and R0 is the initial resistance of the fuse body. For cumulative strain, U represents the maximum allowable strain of the fuse body. h U is the total effective value of voltage harmonics. N This is the rated voltage of the fuse.

[0038] Meanwhile, the self-diagnosis integration module 2 calls the full life cycle database of the fuse body, which includes factory parameters (initial resistance, rated parameters, material properties, etc.), real-time historical data (current / temperature / strain time series data), historical fault records (fault type, occurrence duration, repair status) and self-repair records (number of repairs, repair effect, etc.).

[0039] A dynamic threshold self-optimization model is constructed based on full life cycle data: First, the lifespan decay influence coefficient k is calculated. life Percentage of cumulative runtime The first measure is the impact of quantified lifespan degradation on the threshold (the longer the operation, the lower the threshold adaptability to avoid missing aging-related faults); the second is the calculation of the operating condition severity coefficient k using environmental parameters (humidity, dust, temperature fluctuations) and load parameters (current fluctuations, harmonic content). work Severity of working conditions rating The threshold is adjusted according to the severity of the working conditions (the more severe the working conditions, the higher the threshold is adjusted to avoid misjudgment).

[0040] The optimization formula for the dynamic threshold self-optimization model is: Among them, T std As the standard threshold, k lifeThis is the lifespan degradation impact coefficient. k represents the percentage of cumulative runtime. work The severity coefficient of the working conditions. The severity of the working conditions is rated.

[0041] The optimized normal threshold T is finally calculated based on the dynamic threshold self-optimization model. opt1 Optimized potential fault threshold T opt2 And the optimized failure threshold T opt3 .

[0042] The dynamic threshold determination rule is: S≤T opt1 In normal condition, T opt1 <S≤T opt2 For potential fault conditions, T opt2 <S≤T opt3 The critical melting state is reached when S > T. opt3 It is in a failed state; at the same time, it adaptively matches the fault type based on the fault mode library.

[0043] The system synchronously initiates an adaptive fault mode matching function, matching the core feature parameter sequence with a built-in fault mode library (containing eight typical faults such as overload aging, poor contact, dielectric loss, and arc erosion). A cosine similarity algorithm is used to calculate the matching degree, and the mode with the highest matching degree (≥85%) is identified as the current potential fault type. Combining the state results from dynamic threshold determination with the fault type matching results, the system achieves dual accurate determination of "state diagnosis + fault identification," providing a targeted basis for subsequent lifespan prediction and graded response.

[0044] S4. Remaining Life Dynamic Prediction: The remaining life of the fuse body is calculated using a dynamic aging prediction formula to quantify the failure risk level.

[0045] The remaining life prediction is based on the principle of "irreversible aging of fuse elements + multi-factor coupling influence," constructing a dynamic prediction model of "cumulative damage model + exponential decay correction." The core logic is as follows: fuse aging damage mainly stems from three core factors: current overload heat accumulation, temperature rise thermal stress accumulation, and secondary damage from self-repairing actions. By quantifying the cumulative damage degree of each factor and combining it with an exponential decay model to characterize the change in aging rate with the lifespan, the remaining lifespan is ultimately output. The model is adaptable to all stages of the fuse's life cycle (initial, middle, aging, and critical periods) and can dynamically adjust the weighting coefficients based on the fault types diagnosed in S3, improving prediction accuracy under different operating conditions.

[0046] The input parameters of the prediction model all come from previous steps. After screening and standardization, they are fed into the model to ensure the validity and consistency of the parameters. The core input parameters and processing methods are as follows: Cumulative damage-related parameters (from S2 feature extraction and time series analysis): Cumulative integral value of current overload: Where I is the real-time operating current after calibration, I N The rated current is t, and the cumulative operating time is h. The integral term is calculated using the trapezoidal integration method, with the time step synchronized with the sampling frequency (100Hz-1kHz), which characterizes the cumulative damage of current overload to the conductive substrate of the fuse (the square term of the current is adapted to Joule's law to strengthen the weight of the overload effect).

[0047] Cumulative integral value of temperature increase: Where T is the temperature of the fuse body, T env For ambient temperature, T max To allow the highest operating temperature, the cubic term strengthens the damage weight of the high-temperature region (T≥80℃, close to the upper limit of the rated operating temperature), and the integration range is the cumulative running time t, which is also calculated using the trapezoidal integration method.

[0048] Lifespan impact correction parameters (from S3 diagnostics and full lifecycle database): Self-repair count n: The cumulative number of self-repair triggers is retrieved from the full lifecycle database. Although each self-repair can temporarily restore the conductivity, it will cause damage to the microstructure of the fuse substrate (stress cracks are generated during the melting and solidification process of the repair medium). Therefore, it is used as a negative correction factor in the model.

[0049] Fault type correction coefficient δ: Based on the S3 fault mode adaptive matching results, the aging coefficient weight is dynamically adjusted, with a value range of 0.8-1.2, where: overload aging δ=1.2 (accelerated aging), poor contact δ=1.1 (moderate acceleration), dielectric loss δ=1.0 (normal aging), environmental interference δ=0.9 (slight acceleration), arc erosion δ=1.25 (strong acceleration), and no clear fault δ=0.85 (slow aging).

[0050] Basic parameters (from factory specifications and design standards): Fuse rated service life L0: The rated service life specified by the manufacturer (in this embodiment, it is taken as 10,000h, which can be adjusted according to the voltage level. For 10kV-35kV level, L0=8,000-12,000h; for 110kV-500kV level, L0=10,000-15,000h).

[0051] Inherent aging coefficient: Current aging coefficient k1 = 0.002h -1 Temperature aging coefficient k2 = 0.005h -1 The self-repair cycle influence coefficient k3=0.1, and the coefficient value is based on the aging test data of copper-silver alloy substrate (verified by fitting 1000 cycles of aging test).

[0052] The remaining lifetime calculation adopts a "step-by-step quantization + dynamic correction" process, and is implemented in real time based on the ARM Cortex-M7 processor (calculation latency ≤20ms). The specific steps are as follows: Step 1: Cumulative Damage Quantification: Calculate the cumulative damage D1 from current overload, the cumulative damage D2 from temperature rise, and the self-healing secondary damage D3, respectively, using the following formulas: Wherein, δ is the fault type correction coefficient, which incorporates the impact of fault type on aging acceleration / deceleration into each damage item.

[0053] Step 2: Calculation of total aging damage: The total aging damage D is the sum of the three major damage items (without coupling interference; experiments have verified that each damage item independently affects the aging process), and the formula is: D = D1 + D2 + D3 The total damage D ranges from 0 to ∞. When D ≥ 3.5, it indicates that the fuse is close to failure (remaining life ≤ 500h), and when D ≤ 1.0, it indicates that it is in the mild aging stage (remaining life ≥ 5000h).

[0054] Step 3: Remaining Lifetime Core Calculation: Based on the exponential decay model, and combining the rated service life L0 of the fuse with the total aging damage D, the remaining service life L is calculated. r The core formula is: The exponential function characterizes the accelerating aging rate as damage accumulates (the greater the damage, the faster the remaining life decays, which is consistent with the physical laws of fuse aging).

[0055] Step 4: Dynamic Error Correction To reduce the impact of environmental fluctuations and sensor errors on prediction results, the S3 state assessment index S and the simulation lifetime L of the digital twin model are introduced. sim Error correction is performed using the following formula: Where L is the corrected final remaining lifetime, L sim The remaining lifetime of the digital twin model is simulated based on the current state (error ≤ 3%); 0.2 and 0.1 are correction weights (determined through fitting multiple sets of experiments), and S represents the degree to which the current state deviates from the median value, weakening the prediction error under extreme conditions. The corrected remaining lifetime prediction error is ≤ 5%, meeting the accuracy requirements for engineering applications.

[0056] Step 5: Output and Update Results Store the corrected remaining life L (rounded to the nearest integer, unit: h) in the full life cycle database, and simultaneously output the fault risk level synchronously (L > 2000h: low risk; 1000h < L ≤ 2000h: medium risk; 500h < L ≤ 1000h: high risk; L ≤ 500h: extremely high risk), providing a quantitative basis for hierarchical response. The update cycle of the remaining life is synchronized with the S3 diagnosis cycle (updated once every 30 minutes in the normal state, once every 5 minutes in the potential fault state, and once every 1 minute in the critical / failure state).

[0057] Verification and adaptation of prediction results: To ensure the reliability of the remaining life prediction, a dual verification mechanism is set up: one is real-time verification, comparing the prediction result with the life attenuation curve simulated by the digital twin model. If the deviation > 8%, automatically adjust the fault type correction coefficient δ (±0.05) and recalculate; the other is regular calibration. Every 1000h, fine-tune the inherent aging coefficients (k1, k2, k3) based on the actual operating state (such as the resistance attenuation rate and strain accumulation amount) to adapt to the aging characteristic changes after the fuse body operates for a long time. In addition, the model can receive the actual detection data (such as the off-line resistance measurement value) input by the operation and maintenance personnel through the wireless communication unit, further optimize the prediction accuracy, and adapt to different voltage levels (10kV - 500kV) and complex working conditions (high temperature, high humidity, and multi-dust environment). [

[0058] S5, Hierarchical response and self-diagnosis feedback: According to the state diagnosis result and the remaining life prediction value, trigger the corresponding early warning, self-repair preprocessing or fault isolation actions, and feedback the diagnosis result to the power system monitoring platform.

[0059] The determination rules for the fault risk level are as follows: When S ≤ 0.3 or L > 2000h, it is determined as the normal state: The self-diagnosis module enters the low-power monitoring mode, reduces the sampling frequency to 100Hz, and uploads the status data to the monitoring platform once every 30 minutes through the wireless communication unit. The auxiliary execution module is in the standby state; When 0.3 < S ≤ 0.6 or 1000h < L ≤ 2000h, it is determined as the potential fault state: Trigger a first-level early warning, the 33LED light of the status indicator unit flashes yellow, start the micro cooling module 31 to reduce the temperature of the fuse body, increase the sampling frequency to 500Hz, and upload the status data once every 5 minutes to track the parameter change trend in real time; When 0.6 < S ≤ 0.9 or 500h < L ≤ 1000h, it is determined as the critical fusing state: Trigger a second-level early warning, the 33LED light of the status indicator unit flashes red, start the arc suppression device 32 to prevent the arc from spreading, simultaneously send a fault early warning message to the monitoring platform, increase the sampling frequency to 1kHz, and upload the status data once every 1 minute to remind the operation and maintenance personnel to handle it in time; If S>0.9 or L≤500h, it is determined to be in a failure state: triggering a level three warning, the status indicator unit 33 LED light turns red and stays on, the self-repair trigger unit starts self-repair, and the micro-nano level conductive repair medium is melted by heating with an electric heating wire to fill the fused gap to achieve temporary conductive repair. At the same time, an emergency trip request is sent to the monitoring platform to cooperate with the power system to achieve fault isolation.

[0060] Therefore, the present invention adopts the above-mentioned self-diagnostic high-voltage fuse and its online monitoring method for operating status, which realizes accurate diagnosis, graded response and self-repair linkage of high-voltage fuse operating status, effectively solves the defects of traditional fuse monitoring parameters being single, diagnostic accuracy being low and protection action being lagging, and significantly improves the safety and reliability of high-voltage power distribution system.

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A self-diagnostic high-voltage fuse, characterized in that: It includes a fuse body: made of a new type of self-healing alloy material, with micro-nano conductive repair media embedded inside; Self-diagnosis integration module: integrated in the end caps at both ends of the fuse body, including a physical field sensor unit, a data processing unit, a digital twin virtual modeling unit, a self-calibration unit, a wireless communication unit, and a self-repair trigger unit; Auxiliary execution module: electrically connected to the self-diagnosis integration module, including a micro heat dissipation module, an arc suppression device, and a status indication unit.

2. The self-diagnostic high-voltage fuse according to claim 1, characterized in that: The new type of self-healing alloy material used for the fuse body is based on a copper-silver alloy and added with a nano-aluminum oxide reinforcing phase; the micro-nano conductive repair media is a micro-nano bismuth-tin alloy media, with a melting point of 120-150°C and in a solid-state insulation state below 80°C.

3. The self-diagnostic high-voltage fuse according to claim 1, characterized in that: The physical field sensor unit includes a Rogowski coil current sensor, a fiber Bragg grating temperature sensor, a micro strain sensor, a voltage sensor, and an environmental sensor.

4. The online monitoring method for the operating status of a self-diagnostic high-voltage fuse as described in any one of claims 1-3, characterized in that, It includes the following steps: S1. Multi-dimensional data acquisition and synchronous calibration: The operating parameters and environmental parameters of the fuse body are collected by the physical field sensor unit according to the sampling frequency, and the self-calibration unit performs real-time correction on the collected data based on the calibration factor K; S2. Data preprocessing and multi-factor feature extraction: The calibrated data is filtered and denoised, and the current change rate, temperature change rate, voltage harmonic component, equivalent resistance, and strain accumulation are extracted; S3. Adaptive operating state diagnosis: Based on the multi-parameter coupling state evaluation formula, the state evaluation index S is calculated. Combining with the simulation results of the digital twin model, a dynamic threshold self-optimization model is constructed by introducing the full life cycle data of the fuse body, the state determination threshold is adjusted in real time, and the typical fault modes are adaptively matched to perform accurate state diagnosis and fault type identification; S4. Remaining life dynamic prediction: The remaining life of the fuse body is calculated by the dynamic aging prediction formula to quantify the fault risk level; S5. Hierarchical response and self-diagnosis feedback: According to the state diagnosis result and the remaining life prediction value, the corresponding warning, self-repair preprocessing, or fault isolation actions are triggered, and the diagnosis result is fed back to the power system monitoring platform.

5. The online monitoring method for the operating status of a self-diagnostic high-voltage fuse according to claim 4, characterized in that: In S1, an adaptive calibration algorithm is used to perform real-time correction on the collected data, and the calculation formula of the calibration factor K is: ； Wherein, K0 is the standard calibration coefficient, with a value ranging from 0.98 to 1.02; This is the temperature influence coefficient. Humidity influence coefficient The dust influence coefficient is given by T, where T is the fuse body temperature, T0 is the standard calibration temperature, H is the ambient relative humidity, and D is the ambient dust concentration; the calibrated data X... cal =X raw ×K, X raw For the raw data collected by the sensor, X cal This is the calibrated data.

6. The method for online monitoring of the operating status of a self-diagnostic high-voltage fuse according to claim 5, characterized in that: In S3, the multi-parameter coupling state evaluation formula is: ； Where S is the state assessment index. All are dynamic weighting coefficients, satisfying It is obtained through optimization using a genetic algorithm, and responds to parameter changes in real time; I is the real-time operating current, I N T is the rated current of the fuse, and T is the temperature of the fuse body. env For ambient temperature, T max R is the maximum allowable operating temperature of the fuse. eq R0 is the equivalent resistance of the fuse body, and R0 is the initial resistance of the fuse body. For cumulative strain, U represents the maximum allowable strain of the fuse body. h U is the total effective value of voltage harmonics. N This is the rated voltage of the fuse.

7. The online monitoring method for the operating status of a self-diagnostic high-voltage fuse according to claim 6, characterized in that: The optimization formula of the dynamic threshold self-optimization model is: ； Among them, T std As the standard threshold, k life This is the lifespan degradation impact coefficient. k represents the percentage of cumulative runtime. work The severity coefficient of the working conditions. Rate the severity of the working conditions; The optimized normal threshold T is finally calculated based on the dynamic threshold self-optimization model. opt1 Optimized potential fault threshold T opt2 And the optimized failure threshold T opt3 ; The dynamic threshold determination rule is: S≤T opt1 In normal condition, T opt1 <S≤T opt2 For potential fault conditions, T opt2 <S≤T opt3 The critical melting state is reached when S > T. opt3 It is in a failed state; at the same time, it adaptively matches the fault type based on the fault mode library.

8. The method for online monitoring of the operating status of a self-diagnostic high-voltage fuse according to claim 7, characterized in that: In S4, the dynamic aging prediction formula is: ； Among them, L r L0 represents the remaining lifespan, and L0 represents the rated lifespan of the fuse. Here, k1 is the fault type correction factor, k2 is the current aging factor, k3 is the temperature aging factor, t is the self-repair count influence factor, t is the cumulative running time, and n is the number of self-repair triggers. This is due to the cumulative effect of current overload. This is due to the cumulative effect of increasing temperature.

9. The online monitoring method for the operating status of a self-diagnostic high-voltage fuse according to claim 8, characterized in that: Introducing the state assessment index S in S3 and the simulated lifetime L of the digital twin model sim Error correction is performed using the following formula: ； Among them, L is the final remaining life after correction.

10. The method for online monitoring of the operating status of a self-diagnostic high-voltage fuse according to claim 9, characterized in that, The determination rule for the fault risk level in S5 is: S≤0.3 or L>2000h is determined as the normal state: The self-diagnosis module enters the low-power monitoring mode, reduces the sampling frequency to 100Hz, and uploads the status data to the monitoring platform through the wireless communication unit every 30 minutes. The auxiliary execution module is in the standby state; 0.3<S≤0.6 or 1000h<L≤2000h is determined as the potential fault state: Trigger a first-level warning, the LED light of the status indication unit flashes yellow, start the micro heat dissipation module to reduce the temperature of the fuse body, increase the sampling frequency to 500Hz, and upload the status data every 5 minutes to track the parameter change trend in real time; When 0.6 < S ≤ 0.9 or 500h < L ≤ 1000h, it is determined as the critical fusing state: trigger a secondary warning, the LED light of the status indication unit flashes red, start the arc suppression device to prevent arc diffusion, synchronously send a fault warning message to the monitoring platform, increase the sampling frequency to 1kHz, upload the status data every 1 minute, and remind the operation and maintenance personnel to handle it in time; When S > 0.9 or L ≤ 500h, it is determined as the failure state: trigger a tertiary warning, the LED light of the status indication unit is constantly red, the self-repair trigger unit starts self-repair, melts the micro-nano conductive repair medium by heating the heating wire, fills the fusing gap to achieve temporary conductive repair, and at the same time sends an emergency tripping request to the monitoring platform to cooperate with the power system to achieve fault isolation.

Images