An online monitoring method, storage medium, and electronic device for substation switchgear equipment.

Abstract

This invention relates to an online monitoring method, storage medium, and electronic device for substation switchgear equipment, belonging to the field of detection. The method includes acquiring acoustic signals from the target equipment; performing time-domain to frequency-domain conversion of the acoustic signals using high-resolution short-time fractional Fourier transform; mapping the obtained data from the spectral domain to the pitch domain, normalizing the mapping result to obtain a two-dimensional chromaticity image under the corresponding state; comparing this image with chromaticity images obtained from historical normal data to establish an abnormal state early warning map set for the switchgear equipment; and achieving reliable early warning of the switchgear equipment's operating status based on a computationally efficient and robust residual neural network. This technical solution is computationally efficient and accurate, and the obtained comparative chromaticity map contains rich information on the switchgear equipment's status, providing reliable data and technical support for switchgear equipment status identification and defect early warning. It can be widely used in the field of switchgear equipment operation monitoring.

Description

Technical Field

[0001] This invention belongs to the field of detection, and particularly relates to an online monitoring method, storage medium and electronic device based on substation switchgear equipment. Background Technology

[0002] Switchgear is an important component of the power system's substation and distribution links, and its operational reliability plays a vital role in the safety and stability of the entire power system.

[0003] Existing online monitoring methods for switchgear mostly focus on monitoring partial discharge and temperature, with relatively little research on monitoring the mechanical condition of the equipment. Furthermore, according to relevant statistics, mechanical defects such as poor contact of internal components in switchgear equipment are also a significant cause of equipment failure and maintenance. Therefore, seeking an online monitoring method for the operating status of switchgear equipment warrants further research.

[0004] The invention patent CN 104374423 B, authorized on November 2, 2018, discloses "An Online Monitoring Device and Method for Clustered High-Voltage Switchgear." The online monitoring device includes a central control console, Ethernet, a communication converter, a data bus, and clustered switchgear. Its key features are: each switchgear includes a fiber optic temperature sensor, a vacuum circuit breaker fiber optic displacement sensor, a current sensor, a switchgear power monitoring module, and a data processor. After collecting data signals, the sensors and monitoring modules transmit them to the data processor for analysis and processing. The final processing results are then transmitted via the data bus to the communication converter and uploaded to the central control console via the bus or Ethernet. By employing the above monitoring device and method, the operating status of the clustered switchgear can be dynamically monitored in real time, thereby improving the reliability of the power system and facilitating maintenance and inspection by maintenance personnel. However, this technical solution focuses on conventional temperature, current, and voltage detection methods, which cannot avoid the problems of conventional online monitoring methods failing to identify and alarm for defects due to the large differences in the types of defects in switchgear equipment, nor can it intuitively display the current status of the equipment.

[0005] In addition, conventional online monitoring methods for switchgear include ultrasonic testing and ultra-high frequency (UHF) testing, which have achieved certain results. However, ultrasonic testing has low testing efficiency and accuracy, and its signal analysis error is large for electrical equipment with mechanical vibration. UHF technology has high sensitivity and can accurately locate abnormal noise defects in switchgear, but UHF testing is more complex to operate, has high requirements for testing equipment and defect maps, and is susceptible to electromagnetic interference from the field environment. It requires advanced instruments and experienced experts, resulting in high operating costs.

[0006] In view of this, it is necessary to provide an online monitoring method, storage medium and electronic device for substation switchgear equipment. Based on non-contact monitoring methods to improve monitoring security, a highly recognizable one-dimensional signal-two-dimensional map encoding method is used to intuitively display the current status of the equipment. Furthermore, an abnormal status early warning map set for switchgear equipment is established based on historical normal operation data, thereby improving the operational stability of the power system. Summary of the Invention

[0007] The technical problem this invention aims to solve is to provide an online monitoring method, storage medium, and electronic device for substation switchgear equipment. It acquires acoustic signals from the target equipment, utilizes high-resolution short-time fractional Fourier transform (SFT) technology to achieve time-domain to frequency-domain conversion of the acoustic signals, maps the obtained data from the spectral domain to the pitch domain, normalizes the mapping results to obtain a two-dimensional chromaticity image under the corresponding state, and compares it with chromaticity images obtained from historical normal data to establish an abnormal state early warning map set for the switchgear equipment. Finally, it uses a computationally efficient and robust residual neural network to achieve reliable early warning of the switchgear equipment's operating status. In summary, this invention is computationally efficient and accurate, and the obtained comparative chromaticity map contains rich information on the switchgear equipment's status, providing reliable data and technical support for switchgear equipment status identification and defect early warning.

[0008] The technical solution of this invention is: to provide an online monitoring method for substation switchgear equipment, including acquiring the original time-domain acoustic signal of the switchgear equipment using acoustic sensors, characterized in that:

[0009] 1) Time-frequency domain conversion of the time-domain acoustic signal of the switchgear is achieved by using short-time fractional Fourier transform;

[0010] 2) Map the obtained data from the frequency spectrum domain to the pitch domain;

[0011] 3) Normalize the mapping results to obtain the two-dimensional chromaticity image under the corresponding state;

[0012] 4) By collecting time-domain acoustic signals of substation switchgear equipment under normal operating conditions, a chromaticity image set of substation switchgear equipment under normal operating conditions is constructed;

[0013] 5) Compare the two-dimensional chromaticity images under different states with the chromaticity images obtained from historical normal data, and then establish an abnormal state early warning map set for switchgear equipment;

[0014] 6) A reliable early warning of the operating status of switchgear equipment is realized based on residual neural networks.

[0015] Specifically, the time-frequency domain conversion in step 1) includes: using a short-time fractional Fourier transform to convert the acoustic signal y = {y1, y2, ... y} of the switchgear equipment into a frequency domain. n Perform time-to-frequency domain conversion:

[0016]

[0017] In the formula, F(t,u) represents the signal after short-time fractional Fourier transform, t represents time, and u represents the unknown quantity. Let K(τ,u) represent the window function, K(τ,u) represent the kernel function, and τ represent the time interval.

[0018] Specifically, in step 2), the obtained data is mapped from the frequency spectrum domain to the pitch domain, and the mapping calculation formula is as follows:

[0019]

[0020] In the formula, p represents pitch, and its relationship with frequency f is p = 69 + 12log2(f / 440), k∈[0,N-1] represents the number of indexes of frequency components, and N represents the length of the acoustic signal.

[0021] Specifically, in step 3), the chromaticity vector X can be obtained. ve :

[0022]

[0023] In the formula, c represents the variable, X ve (c) represents the energy of the frequency band to which each pitch component belongs, mod represents the remainder calculation, {p∈[0,127]; p mod 12=c} means that the pitch p is an integer between 0 and 127, c is an integer between 0 and 11, and the remainder of p divided by 12 is equal to c;

[0024] In step 3), the chromaticity vector X ve After normalization, we get:

[0025] X norm (c) = X ve (c) / max(X ve (c)).

[0026] Furthermore, in step 5), the two-dimensional chromaticity images under different states are compared with the chromaticity images obtained from historical normal data. The specific formula for this comparison is as follows:

[0027] X ew =||X a -X n ||2 / ||X n||2

[0028] In the formula, X ew This represents a two-dimensional matrix for early warning comparison, X. n X represents the chromaticity matrix of historical normal data. a This represents the chromaticity matrix obtained by converting real-time acquired signals.

[0029] Furthermore, the residual network mapping layer F is defined as follows:

[0030] F = W1σ(W2x)

[0031] In the formula, x represents the input feature information, σ represents the correction function, and W1 and W2 represent the weight information of the preceding and following feature layers, respectively.

[0032] Furthermore, the different states of the switchgear equipment include at least whether the mechanical state of the substation switchgear equipment is normal, and whether the substation switchgear equipment is in a discharge state.

[0033] The technical solution of the present invention also provides an online status monitoring system device based on substation switchgear equipment using the method described in the above claims, characterized in that it specifically includes:

[0034] The system comprises a data sensing module, a data acquisition module, a data processing module, and a results display module.

[0035] Among them, the data sensing module is used to collect the acoustic signals of the target device;

[0036] The data acquisition module is used to process raw acoustic signal data, specifically including collecting the acoustic signals obtained by the data sensing module and amplifying the signals, ADC acquisition and signal storage, and transmitting the data to the data processing module;

[0037] The data processing module is used to process and calculate acoustic data. First, the time-domain to frequency-domain conversion of the acoustic signal is achieved using short-time fractional Fourier transform. Then, the obtained data is mapped from the frequency domain to the pitch domain, and the mapping result is normalized to obtain a two-dimensional chromaticity image under the corresponding state. This image is then compared with the chromaticity image obtained from historical normal data to establish an abnormal state early warning map set for the switchgear equipment. Finally, a reliable early warning of the operating status of the switchgear equipment is achieved based on a computationally efficient and robust residual neural network.

[0038] The results display module is used to display the chromaticity matrix, chromaticity image, and output of chromaticity comparison warning results of the current status of the switchgear equipment.

[0039] The present invention further provides a computer device that operates using any of the methods described in the preceding claims, comprising a storage module and a processing module, wherein the storage module stores a computer program, characterized in that:

[0040] When the processing module executes the computer program, it uses short-time fractional Fourier transform to realize the time-domain to frequency-domain conversion of acoustic signals; realizes the spectrum-pitch domain mapping and normalization of data; compares the output of the two-dimensional chromaticity image with the historical normal chromaticity image to establish an abnormal state early warning map set for switchgear equipment; and uses residual neural network to realize the identification and early warning of the operating status of switchgear equipment.

[0041] Finally, the present invention provides a computer-readable storage medium for storing the method described in any of the preceding claims, characterized in that:

[0042] The computer-readable storage medium stores a computer program, which, when executed by the processing module, is used to implement the status monitoring method for the switchgear equipment as described in any of the preceding claims.

[0043] Compared with the prior art, the advantages of the present invention are:

[0044] 1. The technical solution of the present invention utilizes high-resolution short-time fractional Fourier transform technology to realize the time-domain to frequency-domain conversion of acoustic signals, and maps the obtained data from the spectrum domain to the pitch domain. The mapping result is normalized to obtain a two-dimensional chromaticity image under the corresponding state, and it is compared with the chromaticity image obtained from historical normal data, thereby establishing an abnormal state early warning map set for switchgear equipment.

[0045] 2. The technical solution of the present invention overcomes the problem that conventional online monitoring methods cannot identify and alarm defects due to the large differences in defect types of switchgear equipment. It can effectively sense the degree of change in the working status of switchgear equipment. Based on a computationally efficient and robust residual neural network, it realizes reliable early warning of the operating status of switchgear equipment.

[0046] 3. The technical solution of the present invention is computationally efficient and yields accurate results. The obtained comparative colorimetric spectrum contains rich information on the status of switchgear equipment, providing reliable data and technical support for status identification and defect early warning of switchgear equipment. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0048] Figure 2 This is a schematic diagram of the modular architecture of the system device of the present invention;

[0049] Figure 3 This is a schematic diagram of the computer device and computer-readable storage medium described in this invention;

[0050] Figure 4a This is a schematic diagram of the colorimetric images of the switchgear under normal and mechanical abnormal conditions.

[0051] Figure 4b This is a schematic diagram of the colorimetric images of the switchgear under normal and abnormal discharge conditions.

[0052] Figure 5a This is a schematic diagram of the performance curve of the loss function during the training process of this invention;

[0053] Figure 5b This is a schematic diagram of the classification accuracy performance curve during the training process of this invention;

[0054] Figure 6 This is a schematic diagram of the confusion matrix of the present invention. Detailed Implementation

[0055] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0056] The technical solution of the present invention provides an online monitoring method, storage medium and electronic device based on substation switchgear equipment.

[0057] In a first aspect, the technical solution of the present invention provides an online status monitoring system device based on substation switchgear equipment, specifically including: a data sensing module, a data acquisition module, a data processing module, and a result display module. (System device framework diagram shown) Figure 2 .

[0058] The data sensing module is used to collect acoustic signals from the target device, and the data acquisition module is used to process the raw acoustic signal data. Specifically, this includes collecting the acoustic signals obtained by the data sensing module and amplifying them, performing ADC acquisition and signal storage, and transmitting the data to the data analysis module.

[0059] The data processing module is used to process and calculate acoustic data. First, a short-time fractional Fourier transform is used to convert the acoustic signal from the time domain to the frequency domain. Then, the obtained data is mapped from the frequency domain to the pitch domain, and the mapping result is normalized to obtain a two-dimensional chromaticity image under the corresponding state. This image is then compared with the chromaticity image obtained from historical normal data to establish an abnormal state early warning map set for the switchgear equipment. Finally, a reliable early warning of the operating status of the switchgear equipment is achieved based on a computationally efficient and robust residual neural network.

[0060] The storage medium unit is used to store a computer program, which is used by the data processing unit to execute the relevant steps of the online monitoring method proposed in this paper.

[0061] The results display module is used to display the chromaticity matrix, chromaticity image, and output of chromaticity comparison warning results of the current status of the switch cabinet equipment.

[0062] Secondly, the technical solution of this invention provides a method for online status monitoring of substation switchgear equipment. This method includes: using short-time fractional Fourier transform to achieve time-domain to frequency-domain conversion of acoustic signals; implementing spectrum-pitch domain mapping and normalization of data; comparing the output of a two-dimensional chromaticity image with historical normal chromaticity images to establish an abnormal status early warning map set for the switchgear equipment; and using a residual neural network to identify and warn of the operating status of the switchgear equipment. The specific steps of the online status monitoring method for substation switchgear equipment are described below. Figure 1 As shown in the image.

[0063] Specifically, the implementation steps of this invention are divided into three steps to achieve the identification and early warning of the operating status of switchgear equipment, as follows:

[0064] 1) First, the acoustic signal y = {y1, y2, ... y} of the switchgear equipment is transformed using a short-time fractional Fourier transform. n Perform time-to-frequency domain conversion:

[0065]

[0066] In the formula, F(t,u) represents the signal after short-time fractional Fourier transform, t represents time, and u represents the unknown quantity. Let K(τ,u) represent the window function, K(τ,u) represent the kernel function, and τ represent the time interval.

[0067] 2) Then, the obtained data is mapped from the frequency spectrum domain to the pitch domain, and the calculation formula is as follows:

[0068]

[0069] In the formula, p represents pitch, and its relationship with frequency f is p = 69 + 12log2(f / 440), k∈[0,N-1] represents the number of indexes of frequency components, and N represents the length of the acoustic signal.

[0070] Furthermore, the chromaticity vector X can be obtained. ve :

[0071]

[0072] In the formula, c represents the variable, X ve(c) represents the energy of the frequency band to which each pitch component belongs, mod represents the remainder calculation, {p∈[0,127]; p mod 12=c} means that the pitch p is an integer between 0 and 127, c is an integer between 0 and 11, and the remainder of p divided by 12 is equal to c.

[0073] Furthermore, for the chromaticity vector X ve After normalization, we get:

[0074] X norm (c) = X ve (c) / max(X ve (c)) (4)

[0075] Furthermore, it is compared with the chromaticity image obtained from historical normal data, using the following formula:

[0076] X ew =||X a -X n ||2 / ||X n ||2 (5)

[0077] In the formula, X ew This represents a two-dimensional matrix for early warning comparison, X. n X represents the chromaticity matrix of historical normal data. a This represents the chromaticity matrix obtained by converting real-time acquired signals.

[0078] 3) Finally, a residual neural network with excellent learning performance is used to achieve effective identification of the compared chroma images. The residual network mapping layer F can be defined as:

[0079] F=W1σ(W2x) (6)

[0080] In the formula, x represents the input feature information, σ represents the correction function, and W1 and W2 represent the weight information of the preceding and following feature layers, respectively.

[0081] Thirdly, the present invention provides a computer device, including a storage module and a processing module. The storage module stores a computer program, and the processing module executes the computer program to implement the status monitoring method for switchgear equipment described in any of the second aspects above.

[0082] Fourthly, the present invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processing module, implements the status monitoring method for switchgear equipment described in any of the second aspects above.

[0083] The simplified diagrams for the third and fourth aspects are shown below. Figure 3 As shown in the image.

[0084] Example:

[0085] To highlight the superiority of the method of this invention, a comparison was made using a set of measured data. This set of data included chromaticity images comparing the normal and mechanical abnormal states of the switchgear equipment, and comparing the normal and discharge abnormal states. The results are as follows: Figure 4a and Figure 4b As shown.

[0086] Depend on Figure 4a As shown, the energy of the chromaticity comparison images of the normal-mechanical-abnormal states of the switchgear equipment is mainly concentrated in the 4-12 order, exhibiting a wide-area distribution; from Figure 4b As shown, the energy of the chromaticity comparison image of the normal-discharge abnormal state of the switchgear equipment is mainly concentrated in the 5-6, 7-9 and 10-11 orders, showing a discontinuous distribution.

[0087] The performance curve of the residual neural network training process using the technical solution of this invention is shown below. Figure 5a and Figure 5b As shown in the image.

[0088] in, Figure 5a The middle curve represents the change in classification accuracy. Figure 5b The curve in the middle represents the corresponding loss function value.

[0089] The results show that the loss function quickly reaches a stable value after 130 iterations. The corresponding classification accuracy also stabilizes after 130 iterations, reaching almost 100%.

[0090] Furthermore, to verify the effectiveness and reliability of the technical solution of this invention, a confusion matrix is ​​introduced for quantitative evaluation of the classification effect, such as... Figure 6 Table 1 shows the results of the confusion matrix calculation obtained using the method described in this invention.

[0091] Table 1 Evaluation Parameters

[0092]

[0093] Figure 6 The diagonals of the confusion matrix represent the number of correctly identified samples for each state using the proposed method.

[0094] Combination Figure 6As shown in Table 1, only 3 out of the 400 test cases were misclassified, resulting in a classification accuracy of 99.3% for the test sample set. Table 1 also shows that the average state recognition precision, recall, and F1 score of the method described in this invention are 0.99, 0.99, and 0.99, respectively. This result demonstrates that the online monitoring method described in this invention possesses excellent classification performance, proving the effectiveness and reliability of the method.

[0095] The technical solution of this invention first utilizes acoustic sensors to acquire the original time-domain acoustic signals of the switchgear equipment. Then, it employs a high-resolution short-time fractional Fourier transform to achieve time-frequency domain conversion of the switchgear's time-domain acoustic signals. Next, the obtained data is mapped from the spectral domain to the pitch domain, and the mapping result is normalized to obtain a two-dimensional chromaticity image under the corresponding state. This image is then compared with chromaticity images obtained from historical normal data to establish an abnormal state early warning map set for the switchgear equipment. Finally, based on a computationally efficient and robust residual neural network, reliable early warning of the switchgear equipment's operating status is achieved. Its computational efficiency and accuracy, along with the rich state information of the switchgear equipment in the obtained comparative chromaticity map, provide reliable data and technical support for the state identification and defect early warning of switchgear equipment.

[0096] This invention can be widely used in the field of operation monitoring of switchgear equipment.

Claims

1. An online monitoring method for substation switchgear equipment, comprising acquiring the original time-domain acoustic signal of the switchgear equipment using acoustic sensors, characterized in that: 1) Time-frequency domain conversion of the time-domain acoustic signal of the switchgear is achieved by using short-time fractional Fourier transform; 2) Map the obtained data from the frequency spectrum domain to the pitch domain; 3) Normalize the mapping results to obtain the two-dimensional chromaticity image under the corresponding state; 4) By collecting time-domain acoustic signals of substation switchgear equipment under normal operating conditions, a chromaticity image set of substation switchgear equipment under normal operating conditions is constructed; 5) Compare the two-dimensional chromaticity images under different states with the chromaticity images obtained from historical normal data, and then establish an abnormal state early warning map set for switchgear equipment; 6) A reliable early warning of the operating status of switchgear equipment is achieved based on residual neural networks.

2. The online monitoring method based on substation switchgear equipment according to claim 1, characterized in that: The time-frequency domain conversion in step 1) includes: using a short-time fractional Fourier transform to convert the acoustic signal of the switchgear equipment... Perform time-to-frequency domain conversion: In the formula, Let represent the signal after the short-time fractional Fourier transform, where t represents time and u represents the unknown quantity. Represents the window function. Represents the kernel function. Indicates a time interval.

3. The online monitoring method based on substation switchgear equipment according to claim 1, characterized in that... In step 2), the obtained data is mapped from the frequency spectrum domain to the pitch domain, and the mapping calculation formula is as follows: In the formula, p represents pitch, and its relationship with frequency f is as follows: , The index represents the number of frequency components, and N represents the length of the acoustic signal.

4. The online monitoring method based on substation switchgear equipment according to claim 1, characterized in that... In step 3), the chromaticity vector can be obtained. : In the formula, c represents a variable. This represents the energy level of the frequency band to which each pitch component belongs. mod indicates the remainder calculation. {p∈[0,127]; p mod 12 = c} means that the pitch p is an integer between 0 and 127, c is an integer between 0 and 11, and the remainder of p divided by 12 is equal to c. In step 3), the chromaticity vector After normalization, we get: X norm (c) = X ve (c) / max(X ve (c)) 。 5. The online monitoring method based on substation switchgear equipment according to claim 1, characterized in that: In step 5), the two-dimensional chromaticity images under different states are compared with the chromaticity images obtained from historical normal data. The specific formula for the comparison is as follows: In the formula, This represents a two-dimensional matrix for early warning comparison. The chromaticity matrix representing historical normal data. This represents the chromaticity matrix obtained by converting real-time acquired signals.

6. The online monitoring method based on substation switchgear equipment according to claim 1, characterized in that: The residual network mapping layer F is defined as follows: In the formula, x represents the input feature information. The function is called the correction function, and W1 and W2 are the weight information of the previous and next feature layers, respectively.

7. The online monitoring method based on substation switchgear equipment according to claim 1, characterized in that: The different states of the switchgear equipment include at least whether the mechanical state of the substation switchgear equipment is normal, and whether the substation switchgear equipment is in a discharge state.

8. A condition online monitoring system device for substation switchgear equipment using the method described in claim 1, characterized in that: Specifically, it includes: The system comprises a data sensing module, a data acquisition module, a data processing module, and a results display module. Among them, the data sensing module is used to collect the acoustic signals of the target device; The data acquisition module is used to process raw acoustic signal data, specifically including collecting the acoustic signals obtained by the data sensing module and amplifying the signals, ADC acquisition and signal storage, and transmitting the data to the data processing module; The data processing module is used to process and calculate acoustic data. First, the time-domain to frequency-domain conversion of the acoustic signal is achieved using short-time fractional Fourier transform. Then, the obtained data is mapped from the frequency domain to the pitch domain, and the mapping result is normalized to obtain a two-dimensional chromaticity image under the corresponding state. This image is then compared with the chromaticity image obtained from historical normal data to establish an abnormal state early warning map set for the switchgear equipment. Finally, a reliable early warning of the operating status of the switchgear equipment is achieved based on a computationally efficient and robust residual neural network. The results display module is used to display the chromaticity matrix, chromaticity image, and output of chromaticity comparison warning results of the current status of the switchgear equipment.

9. A computer device operating using the method of any one of claims 1 to 7, comprising a storage module and a processing module, wherein the storage module stores a computer program, characterized in that: When the processing module executes the computer program, it uses the short-time fractional Fourier transform to realize the time-domain to frequency-domain conversion of acoustic signals; and realizes the spectral domain to pitch domain mapping and normalization processing of data. By comparing the output of the two-dimensional chromaticity image with historical normal chromaticity images, an abnormal state early warning map set for switchgear equipment is established; and the operating status identification and early warning of switchgear equipment is realized by using a residual neural network.

10. A computer-readable storage medium for storing the method according to any one of claims 1 to 7, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by the processing module, is used to implement the status monitoring method for the switchgear equipment according to any one of claims 1 to 7.

Images