A porcelain insulator zero value rapid detection method based on impulse current characteristics

Abstract

The application discloses a porcelain insulator zero value rapid detection method and device based on impact current characteristics, and relates to the technical field of power equipment detection. The method comprises the following steps: constructing a nano-amorphous double-magnetic-ring differential sensing array; applying a high-voltage impact voltage; processing a signal through a fractional-order charge amplification integration-lifting wavelet adaptive filtering-Hilbert phase compensation differential link; extracting equivalent distributed resistance and fractal box dimension characteristics; and determining the insulator deterioration state based on a two-dimensional phase plane. The device comprises a sensing array, a high-voltage generator, a signal processing unit, a main control unit and a temperature compensation unit. The application solves the problems that the prior art cannot distinguish between internal zero value and surface damp, is weak in anti-interference and signal distortion, realizes accurate identification of the deterioration mode of the porcelain insulator, reduces operation and maintenance costs, adapts to complex operating environments, and has important engineering application value.

Description

Technical Field

[0001] This invention relates to the field of power equipment testing technology, and in particular to a rapid method for detecting zero values ​​in porcelain insulators based on impulse current characteristics. Background Technology

[0002] Disc-type suspension porcelain insulators are critical insulation components for high-voltage transmission lines. During long-term operation, they are prone to internal breakdown (zero value), surface dampness, and contamination, directly threatening transmission safety. According to DL / T 2453-2021, "Specification for Zero-Value High-Voltage Impulse Testing of Disc-Type Suspension Porcelain Insulators," the current mainstream testing technology is the high-voltage impulse voltage method, which determines the zero value by measuring the residual voltage amplitude of the insulator. However, this existing technology has significant drawbacks: (1) Inability to distinguish between degradation modes: Existing voltage detection methods can only identify complete breakdown (zero value), and cannot distinguish between internal zero value (ceramic body breakdown and short circuit) and surface degradation (increased surface conductivity due to moisture and dirt). The two are completely different in terms of operational risks and disposal strategies: zero-value insulators need to be replaced immediately, while surface moisture can be restored by cleaning and drying. The misjudgment of existing methods leads to unnecessary power outages for replacement, resulting in high operation and maintenance costs.

[0003] (2) The contradiction between sensor resolution and anti-interference is prominent: the traditional Rogowski coil or ordinary magnetic ring used to measure impulse current is large in size and has low permeability. When measuring impulse current, it is severely affected by common-mode interference such as spatial corona and switching operation, resulting in low signal-to-noise ratio. Furthermore, it cannot be installed close to the insulator, introducing additional lead inductance and distorting the current rising edge characteristics.

[0004] (3) Lack of physical matching in signal processing links: In terms of data processing, the existing RC integral / differential links are integer-order linear systems, while the impedance response of internal defects (cracks, air gaps, moisture) of insulators has Cole-Cole distributed parameter characteristics. Traditional integer-order algorithms cannot accurately restore the transient process of impact current, resulting in feature extraction distortion.

[0005] To address the aforementioned problems, this invention introduces a nano-amorphous magnetic ring to construct a micro-sized through-hole sensing array, designs a fractional integral-lifting wavelet adaptive filtering-Hilbert phase compensation differential signal processing link, and proposes a two-dimensional nonlinear determination method based on the equivalent distributed resistance and the fractal box dimension of the current waveform, thereby realizing the physical essence identification of insulator degradation modes. Summary of the Invention

[0006] This invention aims to overcome the shortcomings of existing technologies and provide a rapid detection method and device for zero values ​​of porcelain insulators based on impulse current characteristics. It solves the problems of existing technologies being unable to distinguish between internal zero values ​​and surface moisture, having weak anti-interference capabilities, and having signal processing distortion. This enables accurate and rapid identification of porcelain insulator degradation modes, reduces operation and maintenance costs, and ensures the safety of transmission lines.

[0007] A rapid method for detecting zero values ​​in porcelain insulators based on impulse current characteristics includes the following steps: S1: A nano-amorphous magnetic ring is used as the core sensing element and is connected in a through-hole style to the current lead of the insulator to be tested. S2: Apply a high-voltage impulse voltage to the insulator; S3: Extract the equivalent distributed resistance characteristics and fractal box dimension characteristics of the impact current through a fractional charge amplification integral-lifting wavelet adaptive filtering-Hilbert phase compensation differential signal processing link; S4: Based on the two-dimensional combination of the equivalent distributed resistance and the fractal box dimension, determine whether the insulator is in a normal state, a surface damp state, or an internal zero-value state.

[0008] Furthermore, a dual-magnetic-ring differential sensing array is constructed using two nano-amorphous magnetic rings. The main sensing magnetic ring is fitted with the current receiving lead, and the auxiliary compensation magnetic ring is suspended to induce common-mode interference. The common-mode rejection ratio is ≥100dB, which is achieved through differential output of the instrumentation amplifier.

[0009] Furthermore, the transfer function of the fractional-order charge amplification integral is H(s) = 1 / (τ1s). α , where τ1=5ms, fractional order α∈[0.5,1.0], is achieved through a tree-shaped RC fractal network to match the distribution parameter characteristics of internal defects in the insulator.

[0010] Furthermore, the lifting wavelet adaptive filtering employs a second-generation wavelet lifting algorithm to decompose the current signal into five levels, and performs adaptive soft thresholding on the detail coefficients, with a threshold coefficient λ=2.5 and a noise standard deviation σ. n The calculation is based on the real-time estimation of the auxiliary compensation magnetic ring.

[0011] Furthermore, the Hilbert phase compensation differential is used to construct the analytic signal z(t) = i(t) + j of the current signal. (t) extracts the instantaneous amplitude and instantaneous phase, and calculates the instantaneous slope as a characteristic of the current change rate, thus avoiding the phase distortion of traditional RC differentiation.

[0012] Furthermore, the fractal box dimension D f The fractal dimension was obtained by performing multi-scale box counting on the current waveform within a 200µs time window, with the scale ε decreasing in powers of 2, and then obtaining the fractal dimension through linear regression.

[0013] Furthermore, the two-dimensional combination decision employs a support vector machine nonlinear classifier, with the input vector being X=[log 10 R eq D f ] TThe equivalent distributed resistance-fractal box dimension phase plane is divided into normal region, moisture region and zero value region.

[0014] Furthermore, the amplitude of the high-voltage impulse voltage is not less than 50kV, and the rated energy is not less than 0.2J.

[0015] A zero-value detection device for porcelain insulators based on impulse current characteristics, used in the above method, includes: Nanocrystalline amorphous magnetic ring dual-magnetic ring differential sensing array; High-voltage impulse voltage generator with an output voltage of not less than 50kV; Fractional charge amplification and integration circuit; Enhance the wavelet adaptive filtering processing unit; Hilbert phase compensation differential processing unit; And the main control unit.

[0016] Furthermore, it also includes a temperature compensation unit, which monitors the magnetic ring temperature through a built-in NTC thermistor and corrects the mutual inductance coefficient in real time according to a temperature coefficient of -0.08% / ℃, ensuring measurement accuracy in an environment of -40℃ to 60℃.

[0017] The beneficial effects of this invention are as follows: 1. Nano-amorphous magnetic ring micro-sized through-hole differential array: For the first time, a nano-amorphous magnetic ring with an inner diameter of 6.5mm is used for micro-sensing of insulator impulse current. The dual-ring differential structure suppresses common-mode interference to less than 1 / 100 of the traditional single-ring solution, solving the problem of accurate measurement of micro-current in strong electromagnetic environment.

[0018] 2. Fractional-Wavelet-Hilbert Three-Level Signal Link: Breaking through the limitations of traditional integer-order RC links, this link uses fractional-order integrals to match the physical nature of the insulator's distributed parameters, enhances wavelet processing for time-frequency adaptive filtering, and utilizes Hilbert transform for distortion-free differentiation. This link improves signal fidelity from 75% in traditional schemes to over 98%.

[0019] 3. Fractal box dimension-equivalent resistance two-dimensional nonlinear determination: For the first time, fractal geometry theory is introduced into insulator condition diagnosis. The fractal complexity of the current waveform is used to distinguish between internal arc discharge (zero value) and surface uniform conductivity (moisture), which solves the long-standing technical bottleneck of "misjudgment of zero value / moisture" in the power industry. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a structural diagram of the zero-value insulator detection system based on impulse current characteristics according to the present invention; Figure 2 This is a schematic diagram comparing the impulse current waveforms of the zero-value insulator and the normal insulator of the present invention; Figure 3 This is a flowchart of the current waveform wavelet packet decomposition feature extraction process of the present invention; Figure 4 This is a schematic diagram of the zero-value recognition decision boundary based on the SVM classifier of this invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0023] A rapid method for detecting zero values ​​in porcelain insulators based on impulse current characteristics includes the following steps: S1: Constructing a nanocrystalline amorphous dual magnetic ring differential sensing array Two nanocrystalline amorphous magnetic rings (outer diameter 12mm, inner diameter 6.5mm, height 4mm, material 1K107B, initial permeability μ) were used. i ≥100000, saturation magnetic induction intensity B s (≥1.23T), construct a differential sensing array: Main sensing magnetic ring: It is a through-hole sleeve connected to the short-circuit wire between the metal accessory (iron cap or steel foot) of the insulator under test and the output terminal of the high voltage pulse generator. Its inner diameter of 6.5mm is precisely matched with the commonly used insulator connecting wire (about 4mm in diameter) to achieve micro-coupling and reduce the influence of lead inductance on di / dt measurement. Auxiliary compensation magnetic ring: It has the same geometric parameters as the main magnetic ring, is suspended in the same spatial electric field but is not fitted with the test current lead, and only induces common-mode spatial electromagnetic interference; Differential output: The secondary windings of two magnetic rings (each wound with 1-2 turns of reinforced insulated flexible wire) are connected to an instrumentation amplifier chip (such as INA333), outputting V. diff (t)= Vmain (t)− V comp (t), with a common-mode rejection ratio (CMRR) ≥100dB, effectively eliminating common-mode noise such as corona discharge and power frequency electromagnetic coupling.

[0024] The relationship between the magnetic ring output voltage and the inrush current is as follows: M is the mutual inductance coefficient, which is guaranteed by the extremely high permeability of the nano-amorphous magnetic ring. Even if the single impact current is only tens of amperes, the secondary side can still output a signal of hundreds of millivolts.

[0025] S2: Fractional charge amplification integral—lifting wavelet adaptive filtering—Hilbert phase compensation differential signal processing link For the differential voltage signal output by the magnetic ring, a three-level processing link with physical matching of distributed parameters is constructed: 1. Fractional-order charge amplification integral (distributed parameter matching): The traditional integer-order integrator transfer function is H(s) = 1 / (τs), which cannot describe the memory and dispersion characteristics of the defect medium inside the insulator. This section uses a fractional-order integrator with the following transfer function: The fractional order α is determined through offline calibration: for a known zero-value insulator, α is adjusted to minimize the simulation response error between the output current waveform and the standard equivalent circuit (Appendix A of DL / T 2453-2021).

[0026] Implementation method: A tree-shaped RC fractal network (5-level RC branches, with the resistance-capacitance ratio distributed geometrically) is used to simulate 1 / s. 0.8 The characteristics include a pre-amplifier (ADA4530-1 op-amp, input bias current <1fA) to achieve high impedance conversion and integration. This stage also performs signal restoration and high-frequency noise pre-suppression (fractional-order operators naturally have low-pass characteristics).

[0027] 2. Enhanced wavelet adaptive filtering (time-frequency adaptive denoising): Traditional fixed bandpass filters (100Hz~1MHz) cannot adapt to the non-stationary spectrum of inrush current. This section employs a second-generation wavelet lifting algorithm to perform a 5-level lifting wavelet decomposition on the integrated current signal: Splitting: Dividing the signal into even-numbered samples and odd-numbered samples; Prediction: Predict odd samples using even samples to obtain detail coefficients d; Update: Update even-numbered samples with detail coefficients to obtain approximation coefficient a; Perform adaptive soft thresholding on detail coefficients: Where λ is the threshold coefficient (λ=2.5), and is the noise standard deviation (estimated in real time by the auxiliary compensation magnetic ring). This algorithm runs in real time on the MCU processor, taking <1ms. Compared to traditional filters, it improves the signal-to-noise ratio by 45dB while preserving the rising edge of the inrush current.

[0028] 3. Hilbert phase-compensated differential (accurate extraction of instantaneous slope): Traditional RC differentiating circuits amplify high-frequency noise and introduce phase distortion when extracting di / dt. This section uses Hilbert transform to construct an analytic signal: For the filtered current signal i(t), its Hilbert transform is: (t), construct the analytic signal z(t) = i(t) + j (t).

[0029] Given the instantaneous amplitude A(t) = |z(t)| and the instantaneous phase ϕ(t) = arg(z(t)), the instantaneous slope is: The real part is used as the current slope characteristic, and the imaginary part is used as the phase abrupt change characteristic. This method eliminates the group delay distortion of traditional differentiation through phase compensation, and can accurately capture the leading edge of arc current oscillation during zero-value insulator breakdown.

[0030] S3: Dual Feature Parameter Extraction (Nonlinear Geometric Features) Within a 200 μs time window after the application of the impulse voltage, two physically essential nonlinear features are extracted: 1. Equivalent distributed resistance R eq : Steady-state current amplitude I output by fractional integral ss With applied impulse voltage V imp (50kV) Calculation: Where η(α) is the fractional order correction coefficient (when α=0.8, η=0.92). R eq R reflects the overall insulation resistance level of the insulator; its internal value is zero. eq <1MΩ, 1MΩ~500MΩ when damp, >500MΩ under normal conditions.

[0031] 2. Current waveform fractal box dimension D f : Box counting analysis was performed on the current waveform i(t) within a 200μs time window: Divide the time-current plane into boxes with side length ε; The number of boxes N(ε) required to cover the waveform is statistically determined. fractal dimension The actual calculation uses multi-scale linear regression.

[0032] Physical meaning: Normal insulator: current waveform is smooth, approximately a one-dimensional curve, D f =1.15~1.35; Surface-moistened insulators: The surface water film homogenizes conductivity, the current waveform shows a monotonically rising trend followed by exponential decay, and the surface roughness is low. f =1.05~1.20; Internal zero-value insulator: Ceramic body breakdown generates arc discharge, the current waveform is rich in high-frequency glitches and oscillations, and the geometric complexity is significantly increased, D f =1.60~1.85.

[0033] S4: Two-dimensional nonlinear state determination (fractal-resistive phase plane) Establish "equivalent distributed resistance R" eq —Fractal box dimension D f "In a two-dimensional phase plane, the three insulator states form separable clustered regions on the phase plane:" The decision algorithm uses a Support Vector Machine (SVM) non-linear classifier. Kernel function: RBF, trained with 200 sets of calibration samples; Input vector: X=[log 10 (R eq ), D f ] T ; Output: Distance to the boundary of the three decision classes, using the nearest neighbor decision.

[0034] When the sample falls into the "fuzzy boundary region" (such as R), eq = 8~12MΩ and D f When the voltage is between 1.4 and 1.6, a secondary confirmation pulse is automatically triggered (the voltage is reduced to 30kV for retesting) to avoid misjudgment.

[0035] S5: Temperature Drift Adaptive Compensation The temperature coefficient of magnetic permeability of the nanocrystalline amorphous magnetic ring is approximately −0.08% / ℃. Within the temperature range of −40℃ to 60℃ (compliant with Clause 4.4 of DL / T2453-2021), a built-in NTC thermistor monitors the magnetic ring temperature T in real time, and the mutual inductance coefficient is corrected using a lookup table method. Ensure R eq Measurement error <±2%, D fThe calculation is unaffected by temperature (the fractal dimension is a geometric invariant).

[0036] Example 1: Portable Zero Value and Moisture Comprehensive Detector Hardware architecture: Sensing unit: Two nano-amorphous magnetic rings 1204 (1K107B, Symmetrically installed on both sides of the insulator's steel foot, with a spacing of 15mm, differential output; Fractional order integral: ADA4530-1 charge amplifier + tree-shaped RC fractal network (5 levels) , , ),accomplish characteristic; Lifting wavelet filtering: ARM Cortex-M4 MCU runs 5-layer lifting wavelet in real time with adaptive thresholding. ; Hilbert Differentiation: After sampling by the ADC (14-bit, 1MS / s), the instantaneous slope is calculated by the MCU through a digital Hilbert filter (FIR, 128th order); Fractal dimension calculation: MCU for Waveform execution box counting method, scale Pick ( ), linear regression .

[0037] On-site testing case: Testing of XZN-300 insulators on a 500kV line: Sample A: Measured , It falls into the normal zone; Sample B: Measured , It fell into the damp area. Maintenance personnel performed a live cleaning and then retested. Restore to , Confirmed that no replacement is needed; Sample C: Measured , It fell into the zero-value zone. After a power outage and replacement, an internal dissection confirmed the existence of a through-breakdown channel, proving the judgment to be accurate.

[0038] Example 2: Equivalent Circuit Parameter Calibration and Fractional Order Optimization Establish the equivalent circuit model of Appendix A of DL / T2453-2021 in the laboratory: Normal insulator: , ; Moisture-damaged insulators: in parallel ; Zero-value insulator: , (Simulated breakdown).

[0039] Applying a 50kV impulse voltage, the results are analyzed using traditional integer-order integrals ( ) and the fractional integral of this invention ( Restore the current waveform.

[0040] Result comparison: Correlation coefficient between integer-order restored waveform and theoretically simulated waveform: (Zero-value sample) (Damp sample); Fractional order ( The correlation coefficient between the restored waveform and the theoretically simulated waveform: (Zero-value sample) (Damp sample).

[0041] Prove that fractional-order operators and the distributed parameter impedance of insulators have a physically essential matching property.

[0042] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A rapid method for detecting zero value in porcelain insulators based on impulse current characteristics, characterized in that, Includes the following steps: S1: A nano-amorphous magnetic ring is used as the core sensing element and is connected in a through-hole style to the current lead of the insulator to be tested. S2: Apply a high-voltage impulse voltage to the insulator; S3: Extract the equivalent distributed resistance characteristics and fractal box dimension characteristics of the impact current through a fractional charge amplification integral-lifting wavelet adaptive filtering-Hilbert phase compensation differential signal processing link; S4: Based on the two-dimensional combination of the equivalent distributed resistance and the fractal box dimension, determine whether the insulator is in a normal state, a surface damp state, or an internal zero-value state.

2. The method according to claim 1, characterized in that, A dual-magnetic-ring differential sensing array is constructed using two nano-amorphous magnetic rings. The main sensing magnetic ring is fitted with the current receiving lead, and the auxiliary compensation magnetic ring is suspended to induce common-mode interference. The signal is output differentially through an instrumentation amplifier, with a common-mode rejection ratio ≥100dB.

3. The method according to claim 1, characterized in that, The transfer function of the fractional charge amplification integral is H(s) = 1 / (τ1s). α , where τ1=5ms, fractional order α∈[0.5,1.0], is achieved through a tree-shaped RC fractal network to match the distribution parameter characteristics of internal defects in the insulator.

4. The method according to claim 1, characterized in that, The lifting wavelet adaptive filtering employs a second-generation wavelet lifting algorithm to decompose the current signal into five levels, and performs adaptive soft thresholding on the detail coefficients, with a threshold coefficient λ=2.5 and a noise standard deviation σ. n The calculation is based on the real-time estimation of the auxiliary compensation magnetic ring.

5. The method according to claim 1, characterized in that, The Hilbert phase-compensated differential is used to construct the analytic signal z(t) = i(t) + j of the current signal. (t) extracts the instantaneous amplitude and instantaneous phase, and calculates the instantaneous slope as a characteristic of the current change rate, thus avoiding the phase distortion of traditional RC differentiation.

6. The method according to claim 1, characterized in that, The fractal box dimension D f The fractal dimension was obtained by performing multi-scale box counting on the current waveform within a 200µs time window, with the scale ε decreasing in powers of 2, and then obtaining the fractal dimension through linear regression.

7. The method according to claim 1, characterized in that, The two-dimensional combination decision uses a support vector machine nonlinear classifier, with the input vector being X=[log 10 R eq D f ] T The equivalent distributed resistance-fractal box dimension phase plane is divided into normal region, moisture region and zero value region.

8. The method according to claim 1, characterized in that, The amplitude of the high-voltage impulse voltage is not less than 50kV, and the rated energy is not less than 0.2J.

9. A zero-value detection device for porcelain insulators based on impulse current characteristics, characterized in that, A method for implementing any one of claims 1-8, comprising: Nanocrystalline amorphous magnetic ring dual-magnetic ring differential sensing array; High-voltage impulse voltage generator with an output voltage of not less than 50kV; Fractional charge amplification and integration circuit; Enhance the wavelet adaptive filtering processing unit; Hilbert phase compensation differential processing unit; And the main control unit.

10. The apparatus according to claim 9, characterized in that, It also includes a temperature compensation unit, which monitors the magnetic ring temperature through a built-in NTC thermistor and corrects the mutual inductance coefficient in real time according to a temperature coefficient of -0.08% / ℃, ensuring measurement accuracy in environments ranging from -40℃ to 60℃.

Images