Composite insulator aging evaluation method based on adaptive nuclear magnetic resonance signal analysis

Abstract

The invention relates to the technical field of power equipment operation and inspection, in particular to a composite insulator aging evaluation method based on self-adaptive nuclear magnetic resonance signal analysis, which acquires signals through a multi-array nuclear magnetic resonance probe and combines self-adaptive CPMG pulse optimization, transverse relaxation time spectrum decomposition, environmental factor dynamic compensation and multi-dimensional mapping modeling. The method comprises the following steps: extracting characteristic parameters such as a main peak, a spectral width, total signal strength and a peak area ratio of a transverse relaxation time distribution T2 spectral line, constructing a mapping relation between the characteristic parameters and the mechanical strength of the bonding interface, and establishing a prediction model by utilizing a vector machine algorithm, so as to evaluate the aging state and predict the residual life. The internal structure of the bonding interface of the composite insulator is deeply detected by using the nuclear magnetic resonance principle, the device can adapt to outdoor working conditions, dynamically compensate environmental factors such as temperature and humidity and electromagnetic interference, evaluate the aging state in real time and predict the residual life, and a theoretical basis is provided for state maintenance of power transmission equipment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric power equipment operation and inspection. BACKGROUND

[0002] With the continuous development of ultra-high voltage and extra-high voltage power transmission technology of power system, composite insulators are widely used in high-voltage transmission lines due to their light weight, high mechanical strength, good pollution flashover resistance and other advantages. However, the adhesive interface between the shed and the sheath of the composite insulator is prone to structural degradation with the extension of service life, and phenomena such as interface debonding, water tree aging, micro-cracks may occur, which may lead to flashover, breakdown and other safety accidents, seriously affecting the stability of the transmission line operation.

[0003] The existing composite insulator detection methods mainly include visual image recognition, infrared thermal imaging, ultrasonic detection and laboratory sampling means. However, these methods often have high false detection rate, rely on manual experience, and cannot deeply detect the internal changes of the interface material. In particular, in terms of online detection, the existing methods cannot achieve high-precision, non-destructive and real-time monitoring of the aging state of the adhesive interface of the composite insulator.

[0004] As a mature non-destructive testing technology, nuclear magnetic resonance technology has been widely used in medical imaging, material detection and other fields, and has the characteristics of high resolution and the ability to detect microstructure changes. However, the traditional nuclear magnetic resonance equipment is bulky and the detection parameter setting is complex, which makes it difficult to adapt to the field use in complex outdoor environments, especially in strong electromagnetic interference environments. Therefore, it is urgent to develop a nuclear magnetic resonance detection and analysis system suitable for field application of power transmission lines, which has self-adaptive parameter adjustment and environmental compensation capability, for accurate evaluation of the aging state of composite insulators. SUMMARY

[0005] In order to overcome the problem that the existing technology cannot achieve high-precision, non-destructive and real-time monitoring of the aging state of the adhesive interface of the composite insulator, the present application provides a composite insulator aging evaluation method based on adaptive nuclear magnetic resonance signal analysis.

[0006] The technical solution adopted by the present application to achieve the above-mentioned purpose is: a composite insulator aging evaluation method based on adaptive nuclear magnetic resonance signal analysis, comprising the following steps:

[0007] S1, using an existing multi-array nuclear magnetic resonance probe to collect a composite insulator, obtaining an original nuclear magnetic resonance echo signal;

[0008] S2, inputting the original nuclear magnetic resonance echo signal into an adaptive CPMG pulse sequence optimization algorithm to obtain an optimized nuclear magnetic resonance echo signal;

[0009] S3, performing echo signal fitting and transverse relaxation time distribution T2 spectral line decomposition on the optimized nuclear magnetic resonance echo signal to obtain decomposed multiple transverse relaxation time components in the composite insulator;

[0010] S4, constructing a prediction model, inputting the transverse relaxation time components into the prediction model, evaluating the aging state of the composite insulator and predicting the remaining life.

[0011] Preferably, in step S2, the adaptive CPMG pulse sequence optimization algorithm formula is: ;

[0012] wherein, is the optimal echo time interval, is the signal response under the current parameters, is the standard reference signal, is the sampling complexity penalty term, is the weight factor.

[0013] Preferably, in step S3, the non-negative least squares method is used to decompose the transverse relaxation time distribution T2 spectral line, and the decomposition formula is as follows: ;

[0014] wherein, is the measured signal value at the observation time point , is the th transverse relaxation time component, is the number of transverse relaxation time components, is the component intensity corresponding to the transverse relaxation time, is the fitting error.

[0015] Preferably, in step S3, feature parameters are extracted from the decomposition results of the transverse relaxation time distribution T2 spectral line, including: main peak T2 time, spectral width, total signal intensity and peak area ratio;

[0016] The main peak T2 time is the transverse relaxation time corresponding to the maximum ;

[0017] The spectral width is: ;

[0018] wherein, is the maximum value of all transverse relaxation time components in the transverse relaxation time distribution T2 spectral line, is the minimum value of all transverse relaxation time components in the transverse relaxation time distribution T2 spectral line;

[0019] ​​Total signal intensity is: ;

[0020] Peak area ratio is; ;

[0021] wherein, is the signal intensity of each transverse relaxation time component.

[0022] Preferably, step S3 further comprises establishing an environmental factor real-time compensation mechanism: ;

[0023] wherein, is the result after environmental compensation correction, is the preliminary measurement result without environmental compensation, respectively are temperature, humidity and electromagnetic interference offset, respectively are temperature, humidity and electromagnetic interference corresponding sensitivity coefficients.

[0024] Preferably, in step S4, a prediction model is constructed by using a vector machine regression algorithm: ;

[0025] wherein, is the interface bonding strength, is the weight of the support vector machine, is the kernel function, is the transverse relaxation time component characteristic parameter vector, is the bias term, and the aging state of the composite insulator is judged according to the interface bonding strength.

[0026] Preferably, in step S4, based on the existing long-term service sample data and experimental fatigue loading data, the remaining service life of the composite insulator is predicted : ;

[0027] wherein, is the failure criterion threshold, is the current detection characteristic value, is the current degradation rate.

[0028] The beneficial effects of the present application are:

[0029] The application utilizes the principle of nuclear magnetic resonance to deeply detect the internal structure of the bonding interface of the composite insulator, and high-precision detection can be realized without disassembling components; the multi-array probe used in the application and the automatic parameter adjustment capability support rapid deployment under complex outdoor working conditions, breaking through the field application limitation of traditional nuclear magnetic equipment; the application comprehensively considers environmental factors such as dynamic compensation temperature and humidity, electromagnetic interference, and the like, and enhances the consistency and reliability of detection data; the application realizes real-time evaluation of the aging state and prediction of the remaining life, and provides a theoretical basis for state maintenance of power transmission equipment. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 is a flowchart of an embodiment of the application. DETAILED DESCRIPTION

[0031] The embodiment of the application provides a composite insulator aging evaluation method based on adaptive nuclear magnetic resonance signal analysis, comprising the following steps:

[0032] S1, using an existing multi-array nuclear magnetic resonance probe, an arc-shaped probe array is designed for the bonding interface area between the shed and the sheath of the composite insulator, the probe array simultaneously collects nuclear magnetic resonance signals of multiple areas of the composite insulator in a parallel mode, obtains original nuclear magnetic resonance echo signals, realizes spatial distribution detection of different parts of the composite insulator, and acquires information such as the internal structure, material distribution, moisture, bonding quality and microstructure of the composite insulator.

[0033] S2, the nuclear magnetic resonance signal collection adopts a Carr-Purcell-Meiboom-Gill (CPMG) sequence to obtain transverse relaxation time T2 spectral line information, the original nuclear magnetic resonance echo signals are input into an adaptive CPMG pulse sequence optimization algorithm, key parameters of the CPMG sequence are adjusted in real time, and optimized nuclear magnetic resonance echo signals are obtained.

[0034] The formula of the adaptive CPMG pulse sequence optimization algorithm is:

[0035] Among them, is an optimal echo time interval, is a signal response under a current parameter, is a standard reference signal, is a sampling complexity penalty term, is a weight factor, the algorithm can automatically adapt to the magnetic resonance characteristics of materials at different aging stages, adjust the CPMG sequence parameters, and improve the detection sensitivity and robustness of the echo signals.

[0036] S3, echo signal fitting and transverse relaxation time distribution T2 spectral line decomposition are performed on the optimized nuclear magnetic resonance echo signals, and a plurality of transverse relaxation time components of the decomposed composite insulator are obtained.​

[0037] Because the echo signal of the insulator has relatively high noise, in order to decompose multiple transverse relaxation time distribution T2 spectral line components within the material and reflect the attenuation characteristics corresponding to different micro-regions (moisture-bound state, interface defect region, and filler aggregation region) of the composite insulator, the non-negative least squares method is used to decompose the transverse relaxation time distribution T2 spectral lines. The decomposition formula is as follows: ;

[0038] in, At the observation time point The measured signal value at time [time]. For the first A horizontal relaxation time component This represents the number of transverse relaxation time components. The component intensities corresponding to the transverse relaxation time distribution T2 spectral lines are... To determine the fitting error, we analyze different... Weight distribution and corresponding The changing trend can be used to extract parameters of several key features to quantify the degree of material aging;

[0039] Feature parameters were extracted from the decomposition results of the transverse relaxation time distribution T2 spectral lines. The parameters included: main peak T2 time, spectral width, total signal intensity, and peak area ratio.

[0040] The peak T2 time has the maximum corresponding The most prominent transverse relaxation time component in the signal can reflect the main attenuation characteristics of the signal. The main peak T2 time reflects the dominant microenvironment in the material. For example, debonding at the interface of composite insulators or moisture intrusion can cause the main peak T2 time to shift.

[0041] Spectral width represents the range of relaxation time distributions of various components in a signal. It is commonly used to measure the diversity of signal attenuation and can reflect the internal structural characteristics of composite insulators. For example, as a composite insulator ages and its structure becomes looser, the spectral width will increase. for: ;

[0042] in, This represents the maximum value of all transverse relaxation time components in the T2 spectral line of the transverse relaxation time distribution. It represents the minimum value of all transverse relaxation time components in the T2 spectral line of the transverse relaxation time distribution;

[0043] The total signal intensity is the sum of all transverse relaxation time components of the signal, which reflects the total content of all substances in the material, and is usually related to the density of moisture, fillers or other substances, and can directly reflect the aging damage of the composite insulator, the total signal intensity is: ;

[0044] The peak area ratio is used to describe the relative contribution of different relaxation time components, which reflects the proportion of different regions (such as moisture, defect area, filler, etc.) in the material, and the peak area ratio is; ;

[0045] wherein, is the signal intensity of each transverse relaxation time component;

[0046] In order to exclude the influence of environmental temperature, humidity, electric field interference and other factors on the detection results, the temperature and humidity sensor and the electromagnetic interference monitoring module are embedded in the embodiment to collect real-time environmental parameters, and a physical-statistical hybrid compensation model is constructed, and the real-time compensation mechanism of environmental factors is established according to the following formula to correct the characteristic value: ;

[0047] wherein, is the result after environmental compensation correction, is the preliminary measurement result without environmental compensation, respectively, the temperature, humidity and electromagnetic interference offset, respectively, the temperature, humidity and electromagnetic interference corresponding sensitivity coefficient.

[0048] S4, constructing a prediction model, inputting the transverse relaxation time component into the prediction model, evaluating the aging state of the composite insulator and predicting the remaining life;

[0049] A mapping model between the nuclear magnetic resonance characteristic parameters (main peak T2 time, spectral width, total signal intensity, peak area ratio) and the mechanical strength of the bonding interface is established, and a prediction model is constructed by using vector machine regression algorithm: ;

[0050] wherein, is the interface bonding strength, is the weight of support vector machine, is the kernel function, transverse relaxation time component characteristic parameter vector, As a bias term, the aging state of the composite insulator is judged according to the bonding strength of the interface, strong robustness is obtained through training, the mapping relationship between different feature vectors and the aging state of the composite insulator is analyzed, and the support vector machine regression model can effectively predict the aging degree of the bonding interface of the composite insulator, consistent evaluation between different batches of insulators can be realized, a plurality of aging modes can be adapted, and a theoretical basis is provided for residual life evaluation;

[0051] Based on the existing long-term service sample data and experimental fatigue loading data, an aging grade evaluation standard and a deterioration rate model are constructed, the distribution range of the nuclear magnetic resonance characteristic parameters under different aging states is counted, a historical degradation curve is established, the current detection characteristic value is combined with the historical degradation curve, and the residual service life of the composite insulator is predicted : ;

[0052] Wherein, is a failure criterion threshold, is a current detection characteristic value, is a current degradation rate.

[0053] The present application is described by embodiments, and those skilled in the art know that various changes or equivalent replacements can be made to these features and embodiments without departing from the spirit and scope of the present application. In addition, under the guidance of the present application, these features and embodiments can be modified to adapt to specific conditions and materials without departing from the spirit and scope of the present application. Therefore, the present application is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of the present application belong to the protection scope of the present application.

Claims

1. A method for aging assessment of composite insulators based on adaptive analysis of nuclear magnetic resonance signals, characterized by, The method comprises the following steps: S1, acquiring nuclear magnetic resonance signals of the composite insulator using a multi-array nuclear magnetic resonance probe to obtain original nuclear magnetic resonance echo signals; S2, inputting the original nuclear magnetic resonance echo signals into an adaptive CPMG pulse sequence optimization algorithm to obtain optimized nuclear magnetic resonance echo signals; S3, performing echo signal fitting and transverse relaxation time distribution T2 spectral line decomposition on the optimized nuclear magnetic resonance echo signals to obtain decomposed multiple transverse relaxation time components in the composite insulator; S4, constructing a prediction model, inputting the transverse relaxation time components into the prediction model, and evaluating the aging state of the composite insulator and predicting the remaining life.

2. The composite insulator aging assessment method based on adaptive nuclear magnetic resonance signal analysis according to claim 1, characterized in that, In step S2, the adaptive CPMG pulse sequence optimization algorithm formula is: ； wherein, is the optimal echo time interval, is the signal response under the current parameters, is the standard reference signal, is the sampling complexity penalty term, is the weight factor.

3. The composite insulator aging assessment method based on adaptive nuclear magnetic resonance signal analysis according to claim 1, characterized in that, In step S3, the non-negative least square method is used to decompose the transverse relaxation time distribution T2 spectral line, and the decomposition formula is as follows: ； wherein is the measured signal value at the observation time point is the measured signal value at the observation time point is the Tl component, is the T2 component, is the number of transverse relaxation time components, is the component intensity corresponding to the transverse relaxation time, is the fitting error.

4. The composite insulator aging assessment method based on adaptive nuclear magnetic resonance signal analysis according to claim 3, characterized in that, In step S3, the decomposition results of the transverse relaxation time distribution T2 spectral line are extracted for characteristic parameters, including: main peak T2 time, spectral width, total signal intensity and peak area ratio; The main peak T2 time is the time with the maximum corresponding ; Spectrum width Is: ； wherein T2maxis the maximum of all transverse relaxation time components in the transverse relaxation time distribution T2spectrum, T2minis the minimum of all transverse relaxation time components in the transverse relaxation time distribution T2spectrum. Total signal intensity Is: ； Peak area ratio Is; ； wherein, is the signal intensity for each transverse relaxation time component.

5. The composite insulator aging assessment method based on adaptive nuclear magnetic resonance signal analysis according to claim 1, characterized in that, In step S3, an environmental factor real-time compensation mechanism is established: ； wherein, is the result after environmental compensation correction, is the preliminary measurement result without environmental compensation, are temperature, humidity and electromagnetic interference offset, respectively, are temperature, humidity and electromagnetic interference corresponding sensitivity coefficients, respectively.

6. The composite insulator aging assessment method based on adaptive nuclear magnetic resonance signal analysis according to claim 1, characterized in that, In step S4, a prediction model is constructed using a vector machine regression algorithm: ； Wherein, is the interface bonding strength, is the weight of the support vector machine, is the kernel function, is the transverse relaxation time component characteristic parameter vector, is the bias term, and the aging state of the composite insulator is judged according to the interface bonding strength.

7. The composite insulator aging assessment method based on adaptive nuclear magnetic resonance signal analysis according to claim 1, characterized in that, In step S4, the remaining service life of the composite insulator is predicted based on the existing long-term service sample data and the experimental fatigue loading data : ； wherein, is a failure criterion threshold, is a current detected characteristic value, is a current degradation rate.

Images