Large transformer bushing vibration monitoring method based on millimeter wave radar

Abstract

The present invention relates to the field of vibration monitoring technology, and in particular to a large transformer bushing vibration monitoring method based on millimeter wave radar, comprising the following steps: non-contact real-time monitoring of the vibration of the transformer bushing by a millimeter wave radar device to generate an original echo signal; pre-processing the original echo signal to eliminate abnormal mutation frequency bands and generate a stable vibration signal; separating the multipath effect of the vibration signal to construct a multi-dimensional vibration signal; generating a vibration characteristic curve of the frequency, phase and amplitude of the corresponding vibration state; obtaining the vibration change pattern of the vibration characteristic curve by time series analysis, and analyzing whether there is an abnormality. The present invention generates a characteristic curve reflecting the vibration state, performs time series analysis on the characteristic curve, and identifies the dynamic change pattern by decomposing the curve into trend components, periodic components and random components.

Description

Technical Field

[0001] The invention relates to the technical field of vibration monitoring, and in particular to a large transformer bushing vibration monitoring method based on millimeter wave radar. Background Art

[0002] As the core equipment in the power system, the operating status of the transformer is crucial to the safety and stability of the entire power grid, especially the vibration status of the bushing of large transformers, which can reflect the health of the internal structure of the equipment and the potential risk of failure. Traditional vibration monitoring mostly relies on the contact monitoring method of mechanical sensors, but this method is complicated to install, inconvenient to maintain, and difficult to avoid electromagnetic interference during equipment operation, resulting in unstable measurement data. In addition, the long-term use of mechanical sensors is susceptible to wear, which reduces the monitoring accuracy and service life.

[0003] As a non-contact, high-precision detection technology with strong anti-interference ability, millimeter-wave radar has shown significant advantages in vibration monitoring. The use of millimeter-wave radar for non-contact vibration detection of transformer bushings can penetrate electromagnetic interference and adapt to complex power environments. However, the original echo signal detected by millimeter-wave radar contains a lot of environmental noise and multipath effects, resulting in unstable signal quality. In addition, the current vibration monitoring technology based on millimeter-wave radar usually lacks the ability to deeply extract multi-dimensional features (such as frequency, phase and amplitude) in vibration signals, and it is difficult to provide accurate vibration state change patterns. Summary of the invention

[0004] The invention provides a large transformer bushing vibration monitoring method based on millimeter wave radar.

[0005] The large transformer bushing vibration monitoring method based on millimeter wave radar includes the following steps:

[0006] S1, non-contact real-time monitoring of transformer bushing vibration is performed using a millimeter-wave radar device to generate raw echo (vibration) signals;

[0007] S2, pre-processing the original echo signal, removing abnormal mutation frequency bands, and generating a stable vibration signal;

[0008] S3, separates the multipath effect of the vibration signal, generates a vibration mode decoupling matrix using the phase difference between the echo characteristics of the millimeter-wave radar and the reflected signal on the surface of the transformer bushing, separates the independent signals of the three vibration dimensions of basic vibration, resonant frequency and high-frequency vibration, and constructs a multi-dimensional vibration signal;

[0009] S4, based on the decoupled multidimensional vibration signal, frequency domain analysis, wavelet packet decomposition and Hilbert-Huang transform are used to extract vibration characteristics and generate vibration characteristic curves of frequency, phase and amplitude corresponding to the vibration state;

[0010] S5, based on the dynamic change of the characteristic curve on the time axis, obtain the vibration change pattern of the vibration characteristic curve through time series analysis, set the threshold range of vibration frequency, phase shift, and amplitude fluctuation based on the historical data of the characteristic curve under normal working conditions, classify different vibration change patterns into normal and abnormal types, form a classification standard for abnormal patterns, compare the vibration change pattern with the classification standard for abnormal patterns, and analyze whether there is an abnormality.

[0011] Optionally, the acquiring of the original millimeter-wave radar signal in S1 specifically includes:

[0012] The millimeter-wave radar device is positioned at the alignment position of the transformer bushing, and the radar transmission and reception angles are preset so that the radar signal can effectively cover the bushing surface. The millimeter-wave radar device continuously transmits high-frequency millimeter-wave signals to the transformer bushing. By adjusting the transmission frequency to adapt to the optimal detection frequency band under the transformer operating environment, it is ensured that the signal penetrates electromagnetic interference and is stably reflected to the radar receiver. The receiver of the millimeter-wave radar device synchronously captures the original millimeter-wave signal reflected back from the bushing surface to form an original echo signal including vibration frequency, vibration amplitude, and phase change information.

[0013] Optionally, the preprocessing in S2 identifies abnormal frequency bands outside the target vibration frequency range by performing frequency domain analysis on the original echo signal, uses a bandpass filter to retain only signals within the target vibration frequency range, and eliminates the identified abnormal frequency bands.

[0014] Optionally, the S3 specifically includes:

[0015] S31, extracting and calculating the phase of the echo signal received by the millimeter wave radar, and identifying the echo signal components in different paths based on the phase difference between the millimeter wave radar and the reflected signal on the surface of the transformer bushing;

[0016] S32, using the extracted multipath phase difference data, establish a vibration mode decoupling matrix , decompose the superimposed signal caused by the multipath effect in the original vibration signal, separate the three-dimensional signals of basic vibration, resonant frequency and high-frequency vibration, and form a multi-dimensional vibration signal , and obtain independent signals reflecting different vibration characteristics.

[0017] Optionally, in the echo signal of the millimeter wave radar, signals of different paths will produce different phase changes, resulting in a multipath effect. The original echo signal emitted by the millimeter wave radar is ,in, is the amplitude of the signal; is the frequency of the millimeter wave signal, is the initial phase. The echo signal reflected from the transformer bushing has a phase shift due to the delay of different paths. Suppose the received echo data is: ,in, It is The amplitude of the signal on each path, It is The phase delay of each path depends on the path length and the location of the radar device. Expressed as and Phase difference of the paths: , by measuring the phase difference between different paths ,determine the relative delay of each path.

[0018] Optional, construction of vibration mode decoupling matrix: To separate fundamental vibration, resonant frequency and high frequency vibration, a vibration mode decoupling matrix is ​​constructed using phase difference , the vibration mode decoupling matrix It is composed of phase difference data, which is used to represent the distribution of different vibration components, including the fundamental vibration , resonant frequency and high frequency vibration , then the vibration mode decoupling matrix Expressed as:

[0019] ;

[0020] in, are the weight coefficients corresponding to the basic vibration components, are the weight coefficients corresponding to the resonant frequency components, are the weight coefficients corresponding to the high-frequency vibration components, and the matrix Determined based on multipath phase difference data, each type of vibration component is expressed independently in the signal.

[0021] Optionally, the multidimensional vibration signal is decoupled by a matrix The inverse matrix right By transforming, we can get the separated fundamental vibration, resonant frequency and high-frequency vibration signals:

[0022] ,in, represents the basic vibration signal, represents the resonant frequency signal, Represents a high frequency vibration signal, is the vector expression of the original echo signal, expressed as: .

[0023] Optionally, the S4 specifically includes:

[0024] S41, frequency domain analysis: decoupled multidimensional vibration signals Perform fast Fourier transform to convert the signal from time domain to frequency domain and extract the main frequency component of the signal to obtain the basic frequency characteristics of the vibration signal. The fast Fourier transform is expressed as:

[0025] ,in, The vibration signal representation in the frequency domain describes the energy distribution of the vibration signal at different frequencies by calculating , identify the main frequency components of the signal, is the frequency, As a Fourier transform operator, the main frequency component provides a benchmark for the subsequent vibration characteristic curve, and can clearly mark the main frequency in the frequency, phase and amplitude characteristic curves, thereby helping to more clearly observe the main vibration mode of the vibration signal and its dynamic changes;

[0026] S42, wavelet packet decomposition: After frequency domain analysis, multidimensional vibration signals Perform wavelet packet decomposition, decompose the signal energy of different frequency bands layer by layer, extract vibration components with high resolution in both time and frequency, capture subtle changes in vibration signals, generate layered frequency and amplitude feature data, and decompose the levels The following components are expressed as:

[0027] ,in, Represents the components of the vibration signal after wavelet packet decomposition, including the decomposition level and decomposition nodes The frequency band information, is the number of levels of decomposition, represents a decomposition node, It is a wavelet basis function, which decomposes the vibration signals of different frequency bands. In each level of decomposition, the wavelet packet coefficients are calculated. , obtain the amplitude characteristics of each frequency band, It is a discrete representation of the original time domain signal, which can describe the vibration state of different frequency bands in detail;

[0028] S43, Hilbert-Huang transform: Apply Hilbert-Huang transform to the decomposed vibration signal to extract the instantaneous frequency and instantaneous amplitude characteristics of the signal, construct the frequency, phase and amplitude characteristic curves of the vibration signal, and reflect the change of the vibration state on the time axis;

[0029] S44, characteristic curve generation: the extracted frequency, phase and amplitude characteristics are plotted on the time axis to generate a vibration characteristic curve to reflect the dynamic characteristics of the current vibration state.

[0030] Optionally, in the Hilbert-Huang transform of S43, the signal after wavelet decomposition is Perform empirical mode decomposition and decompose it into a series of intrinsic mode functions IMFs, denoted as : ,in, It is IMF components, for each IMF component Perform Hilbert transform to obtain instantaneous frequency and instantaneous amplitude: ,in, is the instantaneous amplitude, reflecting the vibration signal at a point in time The amplitude characteristics of is the instantaneous frequency, which represents the signal time point The frequency characteristics of , is the instantaneous phase, which indicates the phase angle of the signal at a certain moment. is the number of IMFs, which determines the number of decomposition levels of the vibration signal.

[0031] Optionally, the vibration characteristic curve in S5 is expressed as , which indicates the frequency, phase or amplitude characteristics in time The value of , Cycle Components and random components Combination of:

[0032] ,in, represents the trend component of the signal, represents the periodic component of the signal, Represents the random component in the signal;

[0033] The trend component Use the sliding average method to extract and smooth the long-term trend of the characteristic curve. For the sliding average method, define a window size , calculate the trend component: , Represents the characteristic value at time point i, trend component It reflects the long-term changes of the characteristic curve, such as the continuous upward or downward trend of the vibration signal;

[0034] The periodic component The extraction of periodic components is done by periodic smoothing or based on known periodic information. , periodic component Indicates repetitive fluctuations in a signal;

[0035] The random component By In , Remove the obtained.

[0036] Beneficial effects of the present invention:

[0037] The present invention realizes high-precision extraction of transformer bushing vibration signals through contactless monitoring of millimeter-wave radar combined with noise suppression. Environmental noise and abnormal frequency bands in the original echo signal are eliminated through frequency domain analysis and bandpass filter to obtain a stable vibration signal. Thereafter, a vibration mode decoupling matrix is ​​generated by utilizing multipath effect separation and phase difference calculation to accurately separate the three dimensions of basic vibration, resonant frequency and high-frequency vibration, thereby ensuring clear extraction of independent components of the signal and laying a high-quality signal foundation for subsequent feature analysis.

[0038] The present invention realizes the all-round dynamic feature extraction of frequency, phase and amplitude through multi-level processing, extracts the main frequency features by frequency domain analysis, captures the subtle vibration changes layer by layer by wavelet packet decomposition, and finally uses Hilbert-Huang transform to extract the instantaneous frequency and instantaneous amplitude, generates a characteristic curve reflecting the vibration state, performs time series analysis on the characteristic curve, identifies the dynamic change pattern by decomposing the curve into trend components, periodic components and random components, sets the abnormal threshold based on the dynamic change of vibration characteristics and historical operating data, forms a classification standard for abnormal patterns, and realizes real-time intelligent early warning and classification feedback. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 A schematic diagram of a monitoring method flow chart of an embodiment of the present invention;

[0041] Figure 2 Schematic diagram of generating a vibration characteristic curve according to an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0043] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiments", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0044] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0045] like Figure 1-Figure 2 As shown, the large transformer bushing vibration monitoring method based on millimeter wave radar includes the following steps:

[0046] S1, non-contact real-time monitoring of transformer bushing vibration is performed using a millimeter-wave radar device to generate raw echo (vibration) signals;

[0047] S2, the original echo signal contains vibration information, but also includes environmental noise and interference components. Therefore, the original echo signal is preprocessed to remove abnormal mutation frequency bands and generate a stable vibration signal;

[0048] S3, separates the multipath effect of the vibration signal, generates a vibration mode decoupling matrix using the phase difference between the echo characteristics of the millimeter-wave radar and the reflected signal on the surface of the transformer bushing, separates the independent signals of the three vibration dimensions of basic vibration, resonant frequency and high-frequency vibration, and constructs a multi-dimensional vibration signal;

[0049] S4, based on the decoupled multidimensional vibration signal, frequency domain analysis, wavelet packet decomposition and Hilbert-Huang transform are used to extract vibration characteristics and generate vibration characteristic curves of frequency, phase and amplitude corresponding to the vibration state;

[0050] S5, based on the dynamic change of the characteristic curve on the time axis, obtain the vibration change pattern of the vibration characteristic curve through time series analysis, set the threshold range of vibration frequency, phase shift, and amplitude fluctuation based on the historical data of the characteristic curve under normal working conditions, classify different vibration change patterns into normal and abnormal types, form a classification standard for abnormal patterns, compare the vibration change pattern with the classification standard for abnormal patterns, and analyze whether there is an abnormality.

[0051] The acquisition of the original millimeter-wave radar signal in S1 specifically includes:

[0052] The millimeter-wave radar device is positioned at the alignment position of the transformer bushing, and the radar transmission and reception angles are preset so that the radar signal can effectively cover the bushing surface. The millimeter-wave radar device continuously transmits high-frequency millimeter-wave signals to the transformer bushing. By adjusting the transmission frequency to adapt to the optimal detection frequency band under the transformer operating environment, it is ensured that the signal penetrates electromagnetic interference and is stably reflected to the radar receiver. The receiver of the millimeter-wave radar device synchronously captures the original millimeter-wave signal reflected back from the bushing surface to form an original echo signal including vibration frequency, vibration amplitude, and phase change information.

[0053] The preprocessing in S2 performs frequency domain analysis on the original echo signal to identify abnormal frequency bands outside the target vibration frequency range, and uses a bandpass filter to retain only the signals within the target vibration frequency range and eliminate the identified abnormal frequency bands.

[0054] First, the original echo signal Perform a Fourier transform to convert the signal from the time domain to the frequency domain: ,in, is the echo signal representation in the frequency domain (i.e., vibration representation), is the frequency, is the original echo signal (vibration signal) in the time domain; by calculating , identify the spectral components of the signal, and thus distinguish abnormal frequency bands outside the target vibration frequency range.

[0055] Identify the target frequency range and abnormal frequency band: Assume the target vibration frequency range is , defines a bandpass filter transfer function To filter out abnormal frequency bands:

[0056] , here, the transfer function Will only remain in arrive The frequency components between , and the components of other frequency bands are completely filtered out (set to 0);

[0057] Apply a bandpass filter to a frequency domain signal After that, the filtered frequency domain signal is obtained. : ,so, Only the signal components within the target vibration frequency range are retained;

[0058] Reconstruct the filtered signal and reconstruct the filtered frequency domain signal Perform inverse Fourier transform and restore to the time domain to obtain a stable vibration signal :

[0059] ,in, It is a stable vibration signal in the time domain after removing the abnormal frequency band interference.

[0060] S3 specifically includes:

[0061] S31, extracting and calculating the phase of the echo signal received by the millimeter wave radar, and identifying the echo signal components in different paths based on the phase difference between the millimeter wave radar and the reflected signal on the surface of the transformer bushing;

[0062] S32, using the extracted multipath phase difference data, establish a vibration mode decoupling matrix , decompose the superimposed signal caused by the multipath effect in the original vibration signal, separate the three-dimensional signals of basic vibration, resonant frequency and high-frequency vibration, and form a multi-dimensional vibration signal , and obtain independent signals reflecting different vibration characteristics.

[0063] In the echo signal of the millimeter-wave radar, signals of different paths will produce different phase changes, resulting in a multipath effect. The original echo signal emitted by the millimeter-wave radar is ,in, is the amplitude of the signal; is the frequency of the millimeter wave signal, is the initial phase. The echo signal reflected from the transformer bushing has a phase shift due to the delay of different paths. Suppose the received echo data is: ,in, It is The amplitude of the signal on each path, It is The phase delay of each path depends on the path length and the location of the radar device. Expressed as and Phase difference of the paths: , by measuring the phase difference between different paths ,determine the relative delay of each path.

[0064] Construction of vibration mode decoupling matrix: To separate fundamental vibration, resonant frequency and high-frequency vibration, the vibration mode decoupling matrix is ​​constructed using phase difference , vibration mode decoupling matrix It is composed of phase difference data, which is used to represent the distribution of different vibration components, including the fundamental vibration , resonant frequency and high frequency vibration , then the vibration mode decoupling matrix Expressed as:

[0065] ;

[0066] in, are the weight coefficients corresponding to the basic vibration components, are the weight coefficients corresponding to the resonant frequency components, are the weight coefficients corresponding to the high-frequency vibration components, and the matrix Determined based on multipath phase difference data, each type of vibration component is expressed independently in the signal.

[0067] Multidimensional vibration signal through decoupling matrix The inverse matrix right By transforming, we can get the separated fundamental vibration, resonant frequency and high-frequency vibration signals:

[0068] ,in, represents the basic vibration signal, represents the resonant frequency signal, Represents a high frequency vibration signal, is the vector expression of the original echo signal, expressed as: .

[0069] S4 specifically includes:

[0070] S41, frequency domain analysis: decoupled multidimensional vibration signals Perform fast Fourier transform to convert the signal from time domain to frequency domain, extract the main frequency component of the signal, and obtain the basic frequency characteristics of the vibration signal. The fast Fourier transform here is the same as the fast Fourier transform frequency domain analysis method in the preprocessing of S2. The difference is that the fast Fourier transform in S2 focuses on the identification and elimination of abnormal frequency bands. The main purpose is to detect and filter out noise or abnormal frequency bands outside the target vibration frequency range, so that the signal is purer, more stable, and suitable for subsequent processing. The fast Fourier transform here focuses on the extraction of basic frequency characteristics, that is, accurately extracting the main frequency characteristics from the decoupled multi-dimensional vibration signal, providing key frequency information for in-depth analysis of the vibration state and generation of characteristic curves. The fast Fourier transform is expressed as:

[0071] ,in, The vibration signal representation in the frequency domain describes the energy distribution of the vibration signal at different frequencies by calculating , identify the main frequency components of the signal, is the frequency, As a Fourier transform operator, the main frequency component provides a benchmark for the subsequent vibration characteristic curve, and can clearly mark the main frequency in the frequency, phase and amplitude characteristic curves, thereby helping to more clearly observe the main vibration mode of the vibration signal and its dynamic changes;

[0072] S42, wavelet packet decomposition: After frequency domain analysis, the main frequency components of the vibration signal are provided, which can quickly identify the main frequency range of the signal and help set the frequency band decomposition layer and frequency band of the wavelet packet decomposition, so as to more accurately perform wavelet packet decomposition near the main frequency, focus on more important frequency components, determine the range of the target frequency interval, and perform wavelet packet decomposition in these frequency ranges in a targeted manner, reducing the decomposition calculation of irrelevant frequency bands. Furthermore, for multidimensional vibration signals, Perform wavelet packet decomposition, decompose the signal energy of different frequency bands layer by layer, extract vibration components with high resolution in both time and frequency, capture subtle changes in vibration signals, generate layered frequency and amplitude feature data, and decompose the levels The following components are expressed as:

[0073] ,in, Represents the components of the vibration signal after wavelet packet decomposition, including the decomposition level and decomposition nodes The frequency band information, is the number of levels of decomposition, represents a decomposition node, It is a wavelet basis function, which decomposes the vibration signals of different frequency bands. In each level of decomposition, the wavelet packet coefficients are calculated. , obtain the amplitude characteristics of each frequency band, It is a discrete representation of the original time domain signal, which can describe the vibration state of different frequency bands in detail;

[0074] S43, Hilbert-Huang transform: Apply Hilbert-Huang transform to the decomposed vibration signal to extract the instantaneous frequency and instantaneous amplitude characteristics of the signal, construct the frequency, phase and amplitude characteristic curves of the vibration signal, and reflect the change of the vibration state on the time axis;

[0075] S44, characteristic curve generation: The extracted frequency, phase and amplitude characteristics are plotted on the time axis to generate a vibration characteristic curve to reflect the dynamic characteristics of the current vibration state, provide real-time characteristic curves under different vibration dimensions, and provide basic data for vibration anomaly detection and analysis.

[0076] In the Hilbert-Huang transform of S43, the signal after wavelet decomposition is Perform empirical mode decomposition and decompose it into a series of intrinsic mode functions IMFs, denoted as : ,in, It is IMF components, for each IMF component Perform Hilbert transform to obtain instantaneous frequency and instantaneous amplitude: ,in, is the instantaneous amplitude, reflecting the vibration signal at a point in time The amplitude characteristics of is the instantaneous frequency, which represents the signal time point The frequency characteristics of , is the instantaneous phase, which indicates the phase angle of the signal at a certain moment. is the number of IMFs, which determines the number of decomposition levels of the vibration signal.

[0077] The vibration characteristic curve in S5 is expressed as , which indicates the frequency, phase or amplitude characteristics in time The value of , Cycle Components and random components Combination of:

[0078] ,in, represents the trend component of the signal, represents the periodic component of the signal, Represents the random component in the signal;

[0079] Trending ingredients Use the sliding average method to extract and smooth the long-term trend of the characteristic curve. For the sliding average method, define a window size , calculate the trend component: , Represents the characteristic value at time point i, which can be the frequency, phase or amplitude of the characteristic curve, trend component Reflects the long-term changes in the characteristic curve, such as the continuous upward or downward trend of the vibration signal; periodic component The extraction of periodic components is done by periodic smoothing or based on known periodic information. , periodic component Represents repetitive fluctuations in a signal, such as daily or weekly cycles; random component By In , Remove the obtained.

[0080] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0081] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A large transformer bushing vibration monitoring method based on millimeter wave radar, characterized in that: The following steps are involved: S1, non-contact real-time monitoring of transformer bushing vibration is performed through a millimeter-wave radar device to generate an original echo signal; S2, pre-processing the original echo signal, removing abnormal mutation frequency bands, and generating a stable vibration signal; S3, separates the multipath effect of the vibration signal, generates a vibration mode decoupling matrix using the phase difference between the echo characteristics of the millimeter-wave radar and the reflected signal on the surface of the transformer bushing, separates the independent signals of the three vibration dimensions of basic vibration, resonant frequency and high-frequency vibration, and constructs a multi-dimensional vibration signal; S4, based on the decoupled multidimensional vibration signal, frequency domain analysis, wavelet packet decomposition and Hilbert-Huang transform are used to extract vibration characteristics and generate vibration characteristic curves of frequency, phase and amplitude corresponding to the vibration state; S5, based on the dynamic change of the characteristic curve on the time axis, obtain the vibration change pattern of the vibration characteristic curve through time series analysis, set the threshold range of vibration frequency, phase shift, and amplitude fluctuation based on the historical data of the characteristic curve under normal working conditions, classify different vibration change patterns into normal and abnormal types, form a classification standard for abnormal patterns, compare the vibration change pattern with the classification standard for abnormal patterns, and analyze whether there is an abnormality.

2. The large transformer bushing vibration monitoring method based on millimeter wave radar according to claim 1 is characterized in that: The acquisition of the original millimeter wave radar signal in S1 specifically includes: The millimeter-wave radar device is positioned at the alignment position of the transformer bushing, and the radar transmission and reception angles are preset. The millimeter-wave radar device continuously transmits high-frequency millimeter-wave signals to the transformer bushing. The receiver of the millimeter-wave radar device synchronously captures the original millimeter-wave signal reflected from the bushing surface to form an original echo signal including vibration frequency, vibration amplitude, and phase change information.

3. The large transformer bushing vibration monitoring method based on millimeter wave radar according to claim 1 is characterized in that: The preprocessing in S2 performs frequency domain analysis on the original echo signal to identify abnormal frequency bands outside the target vibration frequency range, and uses a bandpass filter to retain only the signals within the target vibration frequency range and eliminate the identified abnormal frequency bands.

4. The large transformer bushing vibration monitoring method based on millimeter wave radar according to claim 1 is characterized in that: The S3 specifically includes: S31, extracting and calculating the phase of the echo signal received by the millimeter wave radar, and identifying the echo signal components in different paths based on the phase difference between the millimeter wave radar and the reflected signal on the surface of the transformer bushing; S32, using the extracted multipath phase difference data, establish a vibration mode decoupling matrix , decompose the superimposed signal caused by the multipath effect in the original vibration signal, separate the three-dimensional signals of basic vibration, resonant frequency and high-frequency vibration, and form a multi-dimensional vibration signal .

5. The large transformer bushing vibration monitoring method based on millimeter wave radar according to claim 4 is characterized in that: The original echo signal emitted by the millimeter wave radar is ,in, is the amplitude of the signal; is the frequency of the millimeter wave signal, is the initial phase. The echo signal reflected from the transformer bushing has a phase shift due to the delay of different paths. Suppose the received echo data is: ,in, It is The amplitude of the signal on each path, It is Phase delay of each path, phase difference Expressed as and Phase difference of the paths: , by measuring the phase difference between different paths ,determine the relative delay of each path.

6. The large transformer bushing vibration monitoring method based on millimeter wave radar according to claim 5 is characterized in that: The vibration mode decoupling matrix It is composed of phase difference data, which is used to represent the distribution of different vibration components, including the fundamental vibration , resonant frequency and high frequency vibration .

7. The large transformer bushing vibration monitoring method based on millimeter wave radar according to claim 6 is characterized in that: The multidimensional vibration signal is decoupled by a matrix The inverse matrix right By transforming, we can get the separated fundamental vibration, resonant frequency and high-frequency vibration signals: ,in, represents the basic vibration signal, represents the resonant frequency signal, Represents a high frequency vibration signal, is the vector expression of the original echo signal, expressed as: .

8. The large transformer bushing vibration monitoring method based on millimeter wave radar according to claim 1 is characterized in that: The S4 specifically includes: S41, frequency domain analysis: decoupled multidimensional vibration signals Perform fast Fourier transform to convert the signal from time domain to frequency domain and extract the main frequency component of the signal to obtain the basic frequency characteristics of the vibration signal. The fast Fourier transform is expressed as: ,in, The vibration signal representation in the frequency domain describes the energy distribution of the vibration signal at different frequencies by calculating , identify the main frequency components of the signal, is the frequency, is the Fourier transform operator; S42, wavelet packet decomposition: After frequency domain analysis, multidimensional vibration signals Perform wavelet packet decomposition, decompose the signal energy of different frequency bands layer by layer, extract the vibration components with high resolution in both time and frequency, capture the changes in the vibration signal, generate layered frequency and amplitude feature data, and decompose the level The following components are expressed as: ,in, Represents the components of the vibration signal after wavelet packet decomposition, including the decomposition level and decomposition nodes The frequency band information, is the number of levels of decomposition, represents a decomposition node, It is a wavelet basis function, which decomposes the vibration signals of different frequency bands. In each level of decomposition, the wavelet packet coefficients are calculated. , obtain the amplitude characteristics of each frequency band, is the discrete representation of the original time domain signal; S43, Hilbert-Huang transform: Apply Hilbert-Huang transform to the decomposed vibration signal to extract the instantaneous frequency and instantaneous amplitude characteristics of the signal, construct the frequency, phase and amplitude characteristic curves of the vibration signal, and reflect the change of the vibration state on the time axis; S44, characteristic curve generation: the extracted frequency, phase and amplitude characteristics are plotted on the time axis to generate a vibration characteristic curve to reflect the dynamic characteristics of the current vibration state.

9. The large transformer bushing vibration monitoring method based on millimeter wave radar according to claim 8 is characterized in that: In the Hilbert-Huang transform of S43, the signal after wavelet decomposition is Perform empirical mode decomposition and decompose it into a series of intrinsic mode functions IMFs, denoted as : ,in, It is IMF components, for each IMF component Perform Hilbert transform to obtain instantaneous frequency and instantaneous amplitude: ,in, is the instantaneous amplitude, reflecting the vibration signal at a point in time The amplitude characteristics of is the instantaneous frequency, which represents the signal time point The frequency characteristics of , is the instantaneous phase, which indicates the phase angle of the signal at a certain moment. is the number of IMFs, which determines the number of decomposition levels of the vibration signal.

10. The large transformer bushing vibration monitoring method based on millimeter wave radar according to claim 1, characterized in that: The vibration characteristic curve in S5 is expressed as , which indicates the frequency, phase or amplitude characteristics in time The value of , Cycle Components and random components Combination of: ,in, represents the trend component of the signal, represents the periodic component of the signal, Represents the random component in the signal; The trend component Use the sliding average method to extract and smooth the long-term trend of the characteristic curve. For the sliding average method, define a window size , calculate the trend component: , Represents the characteristic value at time point i, trend component It reflects the long-term changes in the characteristic curve; The periodic component The extraction of periodic components is done by periodic smoothing or based on known periodic information. , periodic component Indicates repetitive fluctuations in a signal; The random component By In , Remove the obtained.

Images