Switch cabinet vibration signal positioning method and device, storage medium and computer equipment

Abstract

According to the switch cabinet vibration signal positioning method and device, the storage medium and the computer equipment provided by the invention, in the operation process of the switch cabinet, the vibration signals generated by the switch cabinet are collected in multiple directions, amplitude detection is carried out after multi-channel time sequence data are formed, and signal segments with potential defects are extracted from the multi-channel time sequence data; further generating a vibration event data packet corresponding to each vibration signal so as to carry out targeted analysis; as the convolution iteration filtering algorithm has better anti-noise capability and higher time resolution, the method can be adopted to extract the signal first arrival time of each vibration event data packet more accurately; and directly performing time difference analysis on the first arrival time of each signal by using the mathematical relationship between the time difference and the vibration source point position established in the finite difference time domain simulation model to obtain the defect vibration source point position of the switch cabinet, so that large-scale disassembly or complex measurement of the switch cabinet is not needed, and the time difference analysis efficiency is improved. Therefore, the position of the defect vibration source point can be accurately determined.

Description

Technical Field

[0001] This application relates to the field of power system technology, and in particular to a method, device, storage medium and computer equipment for locating vibration signals of switchgear. Background Technology

[0002] High-voltage switchgear, as the core control and protection equipment in power transmission and transformation systems, directly impacts the safety of the entire power system through its operational stability. Due to its long-term operation under high voltage, high load, and complex electromagnetic environments, the equipment is highly susceptible to latent defects such as partial discharge, poor contact, and arc breakdown. These defects are often accompanied by sudden energy changes, which in turn cause weak and transient mechanical vibrations within the equipment casing or supporting structure. Therefore, by capturing and analyzing vibration signals, it is hoped to achieve rapid detection and intelligent location of defects within the switchgear.

[0003] In recent years, with the development of sensor technology and signal processing algorithms, defect detection based on vibration signals has become an important auxiliary method after current, voltage, and gas analysis. Especially in the early stages of partial discharge, the mechanical response can be detected even before the electrical signal is leaked, making vibration signals an important information source for early warning. However, vibration signals caused by partial discharge usually have non-stationary characteristics such as small amplitude, short duration, and complex spectrum, and are easily submerged in background interference such as environmental noise, structural resonance, or operational disturbances, making it difficult to accurately extract the signal time difference, which seriously restricts the spatial inversion accuracy based on the three-point positioning method.

[0004] Traditional vibration source localization often employs the TDOA (Time Difference of Arrival) method, which calculates the source coordinates by measuring the time difference of arrival of signals from multiple points and combining this with known sensor locations. However, the complex internal structure of switchgear and the dense arrangement of metal components mean that signal propagation paths often involve multiple refractions and reflections, leading to uncertain propagation speeds and rendering the straight-line propagation model inapplicable, thus causing localization errors. Furthermore, under low signal-to-noise ratio conditions, conventional thresholding, envelope, or cumulative energy methods lack stability in determining the signal origin, further impacting the reliability of the localization system. Summary of the Invention

[0005] The purpose of this application is to at least solve one of the above-mentioned technical defects, in particular the technical defect that the vibration source positioning method in the prior art has poor positioning reliability under the influence of factors such as the complex internal structure of the switch cabinet, the dense metal components, and the low signal-to-noise ratio.

[0006] This application provides a method for locating vibration signals in a switchgear, the method comprising:

[0007] The signals generated during the operation of the switchgear are collected from multiple directions to form multi-channel timing data; the multi-channel timing data consists of multiple vibration signals.

[0008] Amplitude detection is performed on the multi-channel time-series data to obtain detection results. Based on the detection results, signal segments with potential defects are extracted from the multi-channel time-series data to generate a vibration event data packet corresponding to each vibration signal.

[0009] The first arrival time of the signal in each vibration event data packet is extracted using a convolutional iterative filtering algorithm.

[0010] A finite-difference time-domain simulation model is determined, and time difference analysis is performed on the first arrival time of each signal using the finite-difference time-domain simulation model to obtain the location of the defect vibration source point of the switchgear.

[0011] Optionally, the step of acquiring signals generated during the operation of the switchgear from multiple directions to form multi-channel time-series data includes:

[0012] Multiple high-frequency vibration sensors are used to synchronously collect signals generated during the operation of the switchgear, resulting in multiple channel signals;

[0013] An analog-to-digital converter is used to convert the signals of each channel into digital signals to form multiple vibration signals, and multi-channel timing data is generated based on each vibration signal.

[0014] Optionally, the step of performing amplitude detection on the multi-channel time-series data to obtain the detection result includes:

[0015] The sampling time sequence of the multi-channel time-series data is determined, and multiple sampling points are collected from each vibration signal based on the sampling time sequence; the sampling time sequence contains multiple sampling time points.

[0016] For each sampling time point, extract the amplitude value of the sampling point corresponding to that sampling time point in each vibration signal, and determine whether there is a sampling point whose amplitude value is greater than a preset threshold;

[0017] If so, it is confirmed that the multi-channel time-series data has a potential defect at that sampling time point;

[0018] If not, then it is confirmed that the multi-channel time-series data does not have any potential defects at that sampling time point.

[0019] Optionally, the step of extracting signal segments with potential defects from the multi-channel time-series data based on the detection results to generate a vibration event data packet corresponding to each vibration signal includes:

[0020] Based on the detection results, the sampling time point with potential defects is determined, and a signal segment of each vibration signal in the multi-channel time series data at the sampling time point is extracted based on a preset extraction range.

[0021] Generate a vibration event data packet corresponding to each vibration signal based on each signal segment.

[0022] Optionally, the step of extracting the signal first arrival time of each vibration event data packet using a convolutional iterative filtering algorithm includes:

[0023] For each vibration event data packet, Savitzky-Golay smoothing convolution filtering is performed on the vibration event data packet, and the signal start window in the vibration event data packet is identified based on the filtering result;

[0024] The signal's initial window is processed using a moving average method to construct an energy growth curve; the energy growth curve is used to show the change of signal energy over time.

[0025] The time point corresponding to the first signal energy exceeding the preset noise threshold in the energy growth curve is marked as the signal arrival time of the vibration event data packet.

[0026] Optionally, determining the finite-difference time-domain simulation model includes:

[0027] Obtain the structural diagram of the switch cabinet, divide the switch cabinet into spatial grids according to the structural diagram, and generate an initial simulation model based on the multiple spatial grids obtained from the division;

[0028] The virtual defect source points in each spatial network are identified, and unit excitations are injected into each virtual defect source point to simulate the propagation time from each virtual defect source point to each vibration sensor.

[0029] The virtual time difference of each virtual defect source point is constructed based on the propagation time of each virtual defect source point, and each virtual time difference is labeled in the corresponding spatial grid in the initial simulation model to form a finite difference time domain simulation model.

[0030] Optionally, the step of performing time difference analysis on the first arrival time of each signal using the finite difference time-domain simulation model to obtain the location of the defect vibration source point of the switchgear includes:

[0031] The true time difference is constructed based on the first arrival time of each signal, and the Euclidean distance between the true time difference and each virtual time difference in the finite difference time domain simulation model is determined.

[0032] The virtual defect source point corresponding to the virtual time difference with the smallest Euclidean distance is marked as the location of the defect vibration source point of the switchgear.

[0033] This application also provides a switchgear vibration signal positioning device, comprising:

[0034] The signal acquisition module is used to acquire vibration signals generated during the operation of the switchgear from multiple directions to form multi-channel time-series data; the multi-channel time-series data consists of multiple vibration signals.

[0035] The signal detection module is used to perform amplitude detection on the multi-channel time-series data, obtain detection results, and extract signal segments with potential defects from the multi-channel time-series data based on the detection results, so as to generate a vibration event data packet corresponding to each vibration signal.

[0036] The time extraction module is used to extract the first arrival time of the signal in each vibration event data packet using a convolutional iterative filtering algorithm;

[0037] The source point analysis module is used to determine the finite difference time domain simulation model, and to perform time difference analysis on the first arrival time of each signal through the finite difference time domain simulation model to obtain the location of the defect vibration source point of the switchgear.

[0038] This application also provides a storage medium storing computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to perform the steps of the switch cabinet vibration signal positioning method as described in any of the above embodiments.

[0039] This application also provides a computer device, including: one or more processors, and memory;

[0040] The memory stores computer-readable instructions, which, when executed by the one or more processors, perform the steps of the switch cabinet vibration signal positioning method as described in any of the above embodiments.

[0041] As can be seen from the above technical solutions, the embodiments of this application have the following advantages:

[0042] The method, apparatus, storage medium, and computer equipment for locating vibration signals of switchgear provided in this application can synchronously collect vibration signals generated by the switchgear from multiple directions during operation, forming multi-channel time-series data, which can more comprehensively reflect the vibration state of the switchgear. Then, amplitude detection can be performed on the multi-channel time-series data to obtain the detection results, and based on the detection results, signal segments with potential defects can be extracted from the multi-channel time-series data, thereby generating vibration event data packets corresponding to each vibration signal, which can greatly reduce the amount of data processing and facilitate subsequent targeted analysis. Since the convolutional iterative filtering algorithm has good noise resistance and high time resolution, this application can use the convolutional iterative filtering algorithm to extract the signal first arrival time of each vibration event data packet more accurately. At the same time, the mathematical relationship between the time difference and the vibration source point location established in the finite difference time-domain simulation model can be used to directly perform time difference analysis on the first arrival time of each signal to obtain the location of the defect vibration source point of the switchgear. In this way, this application can accurately determine the location of the defect vibration source point without large-scale disassembly or complex measurement of the switchgear. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 A flowchart illustrating a method for locating vibration signals in a switchgear, as provided in an embodiment of this application;

[0045] Figure 2 A flowchart illustrating a multi-channel time-series data amplitude detection process provided in an embodiment of this application;

[0046] Figure 3 A flowchart illustrating a signal first arrival time extraction process provided in an embodiment of this application;

[0047] Figure 4 This is a schematic diagram of the structure of a switchgear vibration signal positioning device provided in an embodiment of this application;

[0048] Figure 5 This is a schematic diagram of the internal structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0049] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0050] Traditional vibration source localization often employs the TDOA (Time Difference of Arrival) method, which calculates the source coordinates by measuring the time difference of arrival of signals from multiple points and combining this with known sensor locations. However, the complex internal structure of switchgear and the dense arrangement of metal components mean that signal propagation paths often involve multiple refractions and reflections, leading to uncertain propagation speeds and rendering the straight-line propagation model inapplicable, thus causing localization errors. Furthermore, under low signal-to-noise ratio conditions, conventional thresholding, envelope, or cumulative energy methods lack stability in determining the signal origin, further impacting the reliability of the localization system.

[0051] Based on this, this application proposes the following technical solution, as detailed below:

[0052] In one embodiment, such as Figure 1 As shown, Figure 1 This is a flowchart illustrating a method for locating vibration signals in a switchgear, as provided in an embodiment of this application. The method specifically includes the following:

[0053] S110: Collects signals generated during the operation of the switchgear from multiple directions to form multi-channel timing data; the multi-channel timing data consists of multiple vibration signals.

[0054] In this step, during the operation of the switchgear, the computer equipment can collect the vibration signals generated by it from multiple directions simultaneously to form multi-channel time-series data. This can more comprehensively reflect the vibration status of the switchgear and avoid subsequent missed or misjudged defects due to missing information.

[0055] Specifically, during the operation of the switchgear, computer equipment deploys multiple vibration sensors within the switchgear. This allows for the simultaneous, multi-directional acquisition of vibration signals generated at different locations, such as the inner wall or key parts of the switchgear casing, generating multi-channel time-series data with a time-corresponding relationship. This acquisition method not only covers the vibration characteristics of key components and structural nodes of the switchgear but also improves the spatial resolution and temporal accuracy of signal acquisition, enabling the computer equipment to comprehensively perceive the dynamic vibration state of the switchgear during operation.

[0056] It is understandable that multi-channel time-series data consists of multiple vibration signals, each of which is acquired by a vibration sensor at a specific location. Therefore, through the fusion analysis of multi-channel time-series data, this application can promptly detect weak mechanical vibration signals caused by defects such as partial discharge, discharge breakdown, or loose contact, thereby enhancing the comprehensiveness and accuracy of defect detection and avoiding the risk of misjudgment and missed detection due to information loss caused by single-channel or asynchronous acquisition.

[0057] S120: Perform amplitude detection on multi-channel time-series data, obtain detection results, and extract signal segments with potential defects from the multi-channel time-series data based on the detection results to generate vibration event data packets corresponding to each vibration signal.

[0058] In this step, after acquiring multi-channel time-series data through step S110, the computer device can perform amplitude detection on the multi-channel time-series data, obtain the detection results, and extract signal segments with potential defects from the multi-channel time-series data based on the detection results, thereby generating a vibration event data packet corresponding to each vibration signal. This can greatly reduce the amount of data processing, so as to facilitate subsequent targeted analysis.

[0059] Specifically, when performing amplitude detection on multi-channel time-series data, computer equipment can quickly identify channel signals that may exhibit abnormal vibrations and their locations by analyzing the amplitude changes of each channel signal on the time axis, thus obtaining preliminary detection results. Since the vibration signals in the multi-channel time-series data are acquired synchronously, when extracting signal segments containing potential defects, the computer equipment can accurately extract signal segments from each vibration signal based on the identified locations. These segments typically exhibit abnormal amplitudes, abrupt changes, or dynamic variations similar to typical fault modes. Therefore, the computer equipment can generate a vibration event data packet corresponding to each vibration signal based on the extracted signal segments.

[0060] Understandably, vibration event data packets not only possess highly structured feature representations but also effectively summarize the core characteristics of vibration events, greatly reducing the burden of analyzing redundant data during subsequent processing and improving data processing efficiency and storage utilization. Through vibration event data packets, the originally massive and complex continuous vibration data can be transformed into several event units with clear diagnostic significance, providing more targeted data support for subsequent in-depth fault location, thereby achieving high efficiency and accuracy in switchgear condition monitoring.

[0061] S130: The first arrival time of the signal in each vibration event data packet is extracted using a convolutional iterative filtering algorithm.

[0062] In this step, after generating the vibration event data packet in step S120, the computer device can use a convolutional iterative filtering algorithm to extract the signal arrival time of each vibration event data packet. The good noise resistance and high time resolution of the convolutional iterative filtering algorithm can improve the accuracy of the time extraction results.

[0063] Understandably, the signal first arrival time refers to the earliest time point in each channel where the vibration event occurs. When extracting the signal first arrival time from vibration event data packets, the convolutional iterative filtering algorithm constructs multi-scale convolution kernels and performs multiple rounds of iterative convolution processing with the vibration signal. This effectively enhances the abrupt change edge features of the signal while suppressing interference components such as environmental noise and mechanical resonance, thereby achieving high-precision localization of the starting point of minute signals. Compared to traditional threshold detection or sliding window methods, convolutional iterative filtering has stronger noise resistance and higher temporal resolution in complex backgrounds, maintaining good recognition accuracy even when the signal waveform is blurred and vibration characteristics are not obvious.

[0064] S140: Determine the finite difference time-domain simulation model, and perform time difference analysis on the first arrival time of each signal using the finite difference time-domain simulation model to obtain the location of the defect vibration source point of the switchgear.

[0065] In this step, after the signal arrival time of each vibration signal is extracted in step S130, the computer equipment can also use the mathematical relationship between the time difference and the vibration source point location established in the finite difference time domain simulation model to directly perform time difference analysis on the signal arrival time and obtain the location of the defect vibration source point of the switchgear. In this way, this application can determine the location of the defect vibration source point more accurately without large-scale disassembly or complex measurement of the switchgear.

[0066] Understandably, through a finite-difference time-domain simulation model, computer equipment can perform time difference analysis on the arrival times of signals collected by various vibration sensors, thereby calculating the specific spatial location of the source point causing the vibration event within the switchgear. The finite-difference time-domain simulation model refers to a time-domain analysis method used for numerical simulation of wave propagation. It discretizes the continuous wave equation in time and space, and iteratively calculates the wave propagation process in the medium using a difference approximation. Therefore, this application can simulate the propagation path and velocity of vibration signals within the internal structure of the switchgear using a finite-difference time-domain simulation model, constructing a high-precision function model between time difference and physical location. This allows computer equipment to quickly calculate the coordinate information of the vibration source point in practical applications by only processing the arrival times of multi-channel signals. Compared to traditional fault location methods, this application eliminates the need for large-scale disassembly of the switchgear or complex physical measurements, thus avoiding location errors and maintenance costs caused by equipment downtime or structural interference, significantly improving the efficiency and accuracy of fault location.

[0067] In the above embodiments, during the operation of the switchgear, vibration signals generated by it can be collected synchronously from multiple directions to form multi-channel time-series data, which can more comprehensively reflect the vibration state of the switchgear. Then, amplitude detection can be performed on the multi-channel time-series data to obtain the detection results. Based on the detection results, signal segments with potential defects can be extracted from the multi-channel time-series data, thereby generating vibration event data packets corresponding to each vibration signal. This can greatly reduce the amount of data processing, so as to facilitate subsequent targeted analysis. Since the convolutional iterative filtering algorithm has good noise resistance and high time resolution, this application can use the convolutional iterative filtering algorithm to extract the signal first arrival time of each vibration event data packet more accurately. At the same time, the mathematical relationship between the time difference and the vibration source point location established in the finite difference time-domain simulation model can be used to directly perform time difference analysis on the first arrival time of each signal to obtain the location of the defect vibration source point of the switchgear. In this way, this application can determine the location of the defect vibration source point more accurately without large-scale disassembly or complex measurement of the switchgear.

[0068] In one embodiment, the process of acquiring signals generated during the operation of the switchgear from multiple directions to form multi-channel time-series data in step S110 may include:

[0069] S111: Multiple high-frequency vibration sensors are used to synchronously collect signals generated during the operation of the switchgear, resulting in multiple channel signals.

[0070] S112: An analog-to-digital converter is used to convert the signals of each channel into digital signals to form multiple vibration signals, and multi-channel timing data is generated based on each vibration signal.

[0071] In this embodiment, during signal acquisition, the computer equipment can synchronously acquire signals generated during the operation of the switch cabinet by using high-frequency vibration sensors deployed in different locations within the switch cabinet, obtaining multiple channel signals. Then, an analog-to-digital converter can be used to convert each channel signal into an analog signal, forming multiple digital vibration signals. Subsequently, multi-channel timing data can be generated based on each vibration signal.

[0072] Specifically, through various high-frequency vibration sensors, the computer equipment can synchronously acquire minute mechanical vibration signals generated by the switchgear during operation, thereby ensuring that vibration responses from different spatial locations are obtained at the same time. These high-frequency vibration sensors have high sensitivity and wide bandwidth response capabilities, and can accurately capture high-frequency vibration characteristics caused by defects such as partial discharge, discharge breakdown, or loose contact during equipment operation, including environmental background noise such as electromagnetic interference, structural resonance, and operational vibration.

[0073] Subsequently, the analog signals collected by each vibration sensor can be converted into digital signals in real time by an analog-to-digital converter, transforming the continuous analog vibration waveforms into digital signals and forming multiple channels of digital vibration signals. The vibration signals of each channel are precisely synchronized on the time axis, thus forming structured multi-channel time-series data. This multi-channel time-series data can accurately reflect the dynamic operating status inside the switchgear and has good comparability and analyzability.

[0074] In one embodiment, such as Figure 2 As shown, Figure 2 A flowchart illustrating a multi-channel time-series data amplitude detection process provided in an embodiment of this application; Figure 2 In step S120, the process of performing amplitude detection on multi-channel time-series data to obtain the detection result may include:

[0075] S121: Determine the sampling time sequence of the multi-channel time series data, and collect multiple sampling points from each vibration signal based on the sampling time sequence; the sampling time sequence contains multiple sampling time points.

[0076] S122: For each sampling time point, extract the amplitude value of the sampling point corresponding to that sampling time point in each vibration signal, and determine whether there is a sampling point whose amplitude value is greater than a preset threshold.

[0077] S123: If so, then it is confirmed that there is a potential defect in the multi-channel time series data at this sampling time point.

[0078] S124: If not, then confirm that the multi-channel time-series data does not have any potential defects at this sampling time point.

[0079] In this embodiment, when performing amplitude detection on multi-channel time-series data, the computer device can first determine the sampling time sequence of the multi-channel time-series data, and then collect multiple sampling points from each vibration signal based on the sampling time sequence containing multiple sampling time points. For each sampling time point, the computer device can extract the amplitude value of the sampling point corresponding to that sampling time point in each vibration signal, and determine whether there is a sampling point whose amplitude value is greater than a preset threshold. Based on the determination result, it can then determine whether there are potential defects in the multi-channel time-series data at that sampling time point.

[0080] It should be noted that before amplitude detection, the computer equipment can first construct a unified sampling time series for time-domain alignment of the data from all channels. This sampling time series contains multiple sampling time points with fixed time intervals, ensuring that the vibration signals from each channel are processed synchronously under the same time reference. Therefore, based on this sampling time series, the computer equipment can extract the sampling points corresponding to each sampling time point from each vibration signal, thereby forming a set of vibration data samples at the same moment.

[0081] Specifically, for each sampling time point, the computer equipment can extract the amplitude values ​​of all channels at that time, construct a cross-channel amplitude distribution vector, and determine one by one whether the amplitude values ​​of one or more channels exceed a preset vibration anomaly threshold. This threshold can be set according to the vibration baseline noise level during normal operation of the switchgear, and is used to effectively distinguish between normal fluctuations and abnormal disturbances. If there is a sampling point with an amplitude value exceeding the preset threshold, the computer equipment can consider that there is a vibration anomaly at that sampling time point, i.e., there is a potential defect; conversely, if there is no sampling point with an amplitude exceeding the threshold, the computer equipment can consider that there is no vibration anomaly at that sampling time point.

[0082] In one embodiment, step S120, which involves extracting signal segments with potential defects from multi-channel time-series data based on the detection results to generate a vibration event data packet corresponding to each vibration signal, may include:

[0083] S125: Based on the test results, determine the sampling time point where there is a potential defect, and extract the signal segment of each vibration signal in the multi-channel time series data at the sampling time point based on the preset interception range.

[0084] S126: Generate a vibration event data packet corresponding to each vibration signal based on each signal segment.

[0085] In this embodiment, after the computer device determines the sampling time point with potential defects based on the detection results, it can extract the signal segment of each vibration signal in the multi-channel time series data at the sampling time point based on the preset interception range, and then generate a vibration event data packet corresponding to each vibration signal based on each signal segment.

[0086] Specifically, the computer equipment can use the sampling time point as the center and, in conjunction with a pre-set interception range, perform interception operations on each vibration signal in multi-channel time-series data, thereby extracting the signal segment of each channel near that time point. This pre-set interception range is typically centered on the sampling time point, extending forward and backward by a fixed time window to ensure it includes the complete onset, development, and decay process of the vibration anomaly, making the extracted signal segments more representative and diagnostically valuable. Here, this application can retain signal segments 200 microseconds before and after the sampling time point. After intercepting the signal segment of each channel, the computer equipment can organize the signal segments into structured data units, generating a vibration event data packet for each channel containing the entire vibration response process, thereby reducing interference from invalid data and lowering data storage and processing costs.

[0087] In one embodiment, such as Figure 3 As shown, Figure 3 A flowchart illustrating a signal first arrival time extraction process provided in an embodiment of this application; Figure 3 In step S130, the process of extracting the signal first arrival time of each vibration event data packet using a convolutional iterative filtering algorithm may include:

[0088] S131: For each vibration event data packet, perform Savitzky-Golay smoothing convolution filtering on the vibration event data packet, and identify the signal start window in the vibration event data packet based on the filtering result.

[0089] S132: The moving average method is used to perform a moving average process on the signal starting window to construct an energy growth curve; the energy growth curve is used to show the change of signal energy over time.

[0090] S133: Mark the time point corresponding to the first signal energy exceeding the preset noise threshold in the energy growth curve as the signal arrival time of the vibration event data packet.

[0091] In this embodiment, for each vibration event data packet, the computer device can perform Savitzky-Golay smoothing convolution filtering on the vibration event data packet, and identify the signal start window in the vibration event data packet based on the filtering result. Then, the moving average method can be used to perform moving average processing on the signal start window to construct an energy growth curve, and the time point corresponding to the first signal energy exceeding the preset noise threshold in the energy growth curve is marked as the signal first arrival time of the vibration event data packet.

[0092] Savitzky-Golay smoothing convolutional filtering refers to a smoothing filtering method based on local polynomial fitting. Its core idea is to perform low-order polynomial regression fitting on the signal within a sliding window, and then use the value of the fitted curve to replace the original data to achieve signal smoothing. The moving average method is a basic signal smoothing technique that reduces short-term fluctuations and highlights long-term trends by taking the arithmetic mean of data within a certain time window.

[0093] Specifically, the computer device fits a polynomial and performs convolution calculations within a local window of the signal using Savitzky-Golay smoothing convolution filtering. This effectively preserves the signal's trend characteristics and waveform details while suppressing high-frequency noise and transient disturbances, resulting in a smoother and distortion-free filtering result. After obtaining the smoothed signal curve, the computer device can identify the time region that may contain the signal's initial waveform, i.e., the signal's initial window, based on the characteristic rate of change or predefined rules. Subsequently, the computer device can perform a moving average process on this initial window to progressively calculate the steady growth trend of the signal energy, thereby constructing a continuous energy growth curve. This curve can intuitively reflect the dynamic change process of energy accumulation in the initial stage of the signal.

[0094] More specifically, the expression for the energy growth curve can be shown below:

[0095]

[0096] In the formula, This represents the signal energy at time t. Indicates the length of the signal start window; This represents the vibration event data packet for the i-th channel.

[0097] Furthermore, the computer equipment can perform threshold comparison on the energy growth curve according to a pre-set noise threshold, and mark the time point when the first signal energy value exceeds the threshold as the signal arrival time of the vibration event data packet. This enables highly sensitive extraction of the arrival time of weak signals, effectively avoiding misjudgments caused by noise disturbances or local anomalies, thereby improving the reliability of vibration source timing localization.

[0098] For example, when selecting a preset noise threshold, the average background energy of the switchgear can be determined first. with standard deviation At this time, the preset noise threshold It can be set to ,in, In the energy growth curve, the first moment that is satisfied. It can be defined as the first arrival time of the signal. .

[0099] In one embodiment, the process of determining the finite-difference time-domain simulation model in step S140 may include:

[0100] S141: Obtain the switch cabinet structure diagram, divide the switch cabinet into spatial meshes according to the switch cabinet structure diagram, and generate an initial simulation model based on the multiple spatial meshes obtained from the division.

[0101] S142: Determine the virtual defect source point in each spatial network and inject a unit excitation into each virtual defect source point to simulate the propagation time from each virtual defect source point to each vibration sensor.

[0102] S143: Based on the propagation time of each virtual defect source point, the virtual time difference of each virtual defect source point is constructed, and each virtual time difference is labeled in the corresponding spatial grid in the initial simulation model to form a finite difference time domain simulation model.

[0103] In this embodiment, when constructing the finite-difference time-domain simulation model of the switchgear, the computer equipment can first obtain the structural diagram of the switchgear, then divide the switchgear into spatial meshes according to the structural diagram, and generate an initial simulation model based on the multiple spatial meshes obtained from the division. Next, virtual defect source points in each spatial network can be determined, and unit excitation can be injected into each virtual defect source point to simulate the propagation time from each virtual defect source point to each vibration sensor. Then, based on the propagation time of each virtual defect source point, a virtual time difference of each virtual defect source point can be constructed, and each virtual time difference can be labeled in the corresponding spatial mesh in the initial simulation model to form a finite-difference time-domain simulation model.

[0104] Understandably, the switchgear structural diagram describes the geometry, spatial distribution, and material properties of the various components inside the switchgear. Based on this diagram, computer equipment can spatially discretize the entire switchgear, using a finite difference mesh generation method to divide the three-dimensional space of the switchgear into a large number of small spatial units with regular shapes, i.e., multiple spatial meshes. An initial simulation model can then be generated based on these spatial meshes, serving as the foundational computational domain for subsequent wave propagation simulations.

[0105] In this application, when using the finite difference mesh generation method, the simulation mesh granularity is 20–50 mm, and the time step is less than 0.01 μs, thus ensuring the accuracy of the propagation time difference. In this application, the step size can be set to 50 mm.

[0106] Specifically, before simulating wave propagation in the initial simulation model, the computer equipment can set a virtual defect source point in each spatial grid and sequentially inject a unit excitation signal, such as a unit pulse or excitation wave, into each virtual defect source point to simulate the propagation behavior of the signal in the internal structure of the switchgear when a small vibration event occurs at that location. Here, the computer equipment can calculate the propagation time required for each excitation signal to travel from the virtual source point to each deployed vibration sensor using the finite difference time-domain method, thereby obtaining the virtual time difference corresponding to each source point.

[0107] When constructing the virtual time difference for each source point, the computer equipment can use a specific vibration sensor as a reference to convert each set of propagation times into the virtual time difference corresponding to that source point, i.e., the propagation time difference pattern generated by that source point location among all vibration sensors. These virtual time differences have obvious locational characteristics, so the computer equipment can annotate the time difference results of each virtual source point onto the corresponding spatial grid in the simulation model, gradually constructing a finite difference time-domain simulation model that covers the entire switchgear space and has high-precision time response characteristics.

[0108] For example, suppose a switch cabinet is equipped with three vibration sensors, L1, L2, and L3, and the propagation time from a virtual defect source to these three vibration sensors is respectively... , , At this point, the computer equipment can use sensor L1 as a reference to construct the virtual time difference of the virtual defect source point: .

[0109] In one embodiment, step S140, which involves performing time difference analysis on the first arrival time of each signal using a finite difference time-domain simulation model to obtain the location of the defect vibration source point in the switchgear, may include:

[0110] S144: Construct the true time difference based on the first arrival time of each signal, and determine the Euclidean distance between the true time difference and each virtual time difference in the finite difference time domain simulation model.

[0111] S145: Mark the virtual defect source point corresponding to the virtual time difference with the smallest Euclidean distance as the location of the defect vibration source point of the switchgear.

[0112] In this embodiment, the computer device can construct the real time difference based on the first arrival time of each signal, determine the Euclidean distance between the real time difference and each virtual time difference in the finite difference time domain simulation model, and then mark the virtual defect source point corresponding to the virtual time difference with the smallest Euclidean distance as the location of the defect vibration source point of the switch cabinet.

[0113] Specifically, after constructing the true time difference of the first arrival time of each signal, the computer equipment can match this true time difference with each virtual time difference in the finite difference time-domain simulation model. Here, the computer equipment can calculate the Euclidean distance between each virtual time difference and the true time difference to measure the similarity between the two. The smaller the Euclidean distance, the closer the virtual time difference is to the actual vibration propagation, meaning the more consistent the spatial location of the virtual source point is with the actual vibration source. Finally, the computer equipment can select the virtual defect source point corresponding to the virtual time difference with the smallest Euclidean distance and mark it as the actual defect source point location where vibration anomalies occur, serving as the spatial location result of potential structural faults or component anomalies in the switchgear.

[0114] For example, when the actual time difference is At that time, it is compared with each virtual time difference in the finite difference time domain simulation model. The formula for comparing Euclidean distances can be shown below:

[0115]

[0116] Among them, computer equipment can select the point with the smallest error. As the location of the defect source, i.e. .

[0117] The following describes the switchgear vibration signal positioning device provided in the embodiments of this application. The switchgear vibration signal positioning device described below can be referred to in correspondence with the switchgear vibration signal positioning method described above.

[0118] In one embodiment, as shown in 4 Figure 4 This application provides a schematic diagram of the structure of a switchgear vibration signal positioning device according to an embodiment of the present application. The present application also provides a switchgear vibration signal positioning device, including a signal acquisition module 210, a signal detection module 220, a time extraction module 230, and a source point analysis module 240, specifically comprising the following:

[0119] The signal acquisition module 210 is used to acquire vibration signals generated during the operation of the switchgear from multiple directions to form multi-channel time-series data; the multi-channel time-series data consists of multiple vibration signals.

[0120] The signal detection module 220 is used to perform amplitude detection on multi-channel time-series data, obtain detection results, and extract signal segments with potential defects from the multi-channel time-series data based on the detection results to generate a vibration event data packet corresponding to each vibration signal.

[0121] The time extraction module 230 is used to extract the signal arrival time of each vibration event data packet using a convolutional iterative filtering algorithm.

[0122] The source point analysis module 240 is used to determine the finite difference time domain simulation model and to perform time difference analysis on the first arrival time of each signal through the finite difference time domain simulation model to obtain the location of the defect vibration source point of the switchgear.

[0123] In the above embodiments, during the operation of the switchgear, vibration signals generated by it can be collected synchronously from multiple directions to form multi-channel time-series data, which can more comprehensively reflect the vibration state of the switchgear. Then, amplitude detection can be performed on the multi-channel time-series data to obtain the detection results. Based on the detection results, signal segments with potential defects can be extracted from the multi-channel time-series data, thereby generating vibration event data packets corresponding to each vibration signal. This can greatly reduce the amount of data processing, so as to facilitate subsequent targeted analysis. Since the convolutional iterative filtering algorithm has good noise resistance and high time resolution, this application can use the convolutional iterative filtering algorithm to extract the signal first arrival time of each vibration event data packet more accurately. At the same time, the mathematical relationship between the time difference and the vibration source point location established in the finite difference time-domain simulation model can be used to directly perform time difference analysis on the first arrival time of each signal to obtain the location of the defect vibration source point of the switchgear. In this way, this application can determine the location of the defect vibration source point more accurately without large-scale disassembly or complex measurement of the switchgear.

[0124] In one embodiment, the signal acquisition module 210 may include:

[0125] The signal acquisition submodule is used to synchronously acquire signals generated during the operation of the switchgear using multiple high-frequency vibration sensors, thereby obtaining signals from multiple channels.

[0126] The analog-to-digital conversion submodule is used to perform analog-to-digital conversion on each channel signal using an analog-to-digital converter to form multiple vibration signals, and generate multi-channel timing data based on each vibration signal.

[0127] In one embodiment, the signal detection module 220 may include:

[0128] The sampling point acquisition submodule is used to determine the sampling time sequence of multi-channel time series data, and to acquire multiple sampling points from each vibration signal based on the sampling time sequence; the sampling time sequence contains multiple sampling time points.

[0129] The amplitude determination submodule is used to extract the amplitude value of the sampling point corresponding to the sampling time point in each vibration signal for each sampling time point, and to determine whether there is a sampling point whose amplitude value is greater than a preset threshold.

[0130] The first result confirmation submodule is used to confirm, when present, that there is a potential defect in the multi-channel time series data at that sampling time point.

[0131] The second result confirmation submodule is used to confirm that, if the result is not present, the multi-channel time-series data does not have any potential defects at the sampling time point.

[0132] In one embodiment, the signal detection module 220 may further include:

[0133] The signal interception submodule is used to determine the sampling time point where there is a potential defect based on the detection results, and to intercept the signal segment of each vibration signal in the multi-channel time series data at the sampling time point based on the preset interception range.

[0134] The data packet generation submodule is used to generate vibration event data packets corresponding to each vibration signal based on each signal segment.

[0135] In one embodiment, the time extraction module 230 may include:

[0136] The window recognition submodule is used to perform Savitzky-Golay smoothing convolution filtering on each vibration event data packet and identify the signal start window in the vibration event data packet based on the filtering result.

[0137] The curve construction submodule is used to perform a moving average process on the signal starting window to construct an energy growth curve; the energy growth curve is used to show the change of signal energy over time.

[0138] The time stamping submodule is used to mark the time point corresponding to the first signal energy exceeding the preset noise threshold in the energy growth curve as the signal arrival time of the vibration event data packet.

[0139] In one embodiment, the process of determining the finite-difference time-domain simulation model in step S140 may include:

[0140] The model building submodule is used to obtain the switch cabinet structure diagram, divide the switch cabinet into spatial meshes according to the switch cabinet structure diagram, and generate an initial simulation model based on the multiple spatial meshes obtained from the division.

[0141] The defect simulation submodule is used to determine the virtual defect source points in each spatial network and inject unit excitation into each virtual defect source point to simulate the propagation time from each virtual defect source point to each vibration sensor.

[0142] The model determination submodule is used to construct the virtual time difference of each virtual defect source point based on the propagation time of each virtual defect source point, and to label each virtual time difference into the corresponding spatial grid in the initial simulation model to form a finite difference time domain simulation model.

[0143] In one embodiment, the source analysis module 240 may include:

[0144] The distance calculation submodule is used to construct the true time difference based on the first arrival time of each signal, and to determine the Euclidean distance between the true time difference and each virtual time difference in the finite difference time domain simulation model.

[0145] The location marking submodule is used to mark the virtual defect source point corresponding to the virtual time difference with the smallest Euclidean distance as the location of the defect vibration source point of the switchgear.

[0146] In one embodiment, this application also provides a storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the switch cabinet vibration signal positioning method as described in any of the above embodiments.

[0147] In one embodiment, this application also provides a computer device storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the switch cabinet vibration signal positioning method as described in any of the above embodiments.

[0148] Indicatively, such as Figure 5 As shown, Figure 5 This is a schematic diagram of the internal structure of a computer device 300 provided in an embodiment of this application. The computer device 300 can be provided as a server. (Refer to...) Figure 5 The computer device 300 includes a processing component 302, which further includes one or more processors, and memory resources represented by memory 301 for storing instructions, such as application programs, that can be executed by the processing component 302. The application programs stored in memory 301 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 302 is configured to execute instructions to perform the switchgear vibration signal positioning method of any of the above embodiments.

[0149] The computer device 300 may also include a power supply component 303 configured to perform power management of the computer device 300, a wired or wireless network interface 304 configured to connect the computer device 300 to a network, and an input / output (I / O) interface 305. The computer device 300 may operate on an operating system stored in memory 301, such as Windows Server™, Mac OS X™, Unix™, Linux™, Free BSD™, or similar.

[0150] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0151] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0152] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0153] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A switchgear vibration signal positioning method, characterized in that, The method comprises: Collecting signals generated during operation of the switch cabinet from multiple directions to form multi-channel time sequence data; the multi-channel time sequence data is composed of multiple vibration signals; Detecting amplitudes of the multi-channel time sequence data to obtain a detection result, and extracting a signal segment with potential defects from the multi-channel time sequence data according to the detection result to generate a vibration event data packet corresponding to each vibration signal; Extracting a signal first arrival time of each vibration event data packet by using a convolution iterative filtering algorithm; Determining a finite difference time domain simulation model, and performing time difference analysis on the signal first arrival times by using the finite difference time domain simulation model to obtain a defect vibration source point position of the switch cabinet.

2. The switchgear vibration signal positioning method according to claim 1, characterized in that, The collecting signals generated during operation of the switch cabinet from multiple directions to form multi-channel time sequence data comprises: Synchronously collecting signals generated during operation of the switch cabinet by using multiple high-frequency vibration sensors to obtain multiple channel signals; Converting the channel signals into analog signals by using an analog-to-digital converter to form multiple vibration signals, and generating multi-channel time sequence data according to the vibration signals.

3. The switchgear vibration signal positioning method of claim 1, wherein The detecting amplitudes of the multi-channel time sequence data to obtain a detection result comprises: Determining a sampling time sequence of the multi-channel time sequence data, and collecting multiple sampling points from each vibration signal based on the sampling time sequence; the sampling time sequence comprises multiple sampling time points; For each sampling time point, extracting an amplitude value of a sampling point corresponding to the sampling time point in each vibration signal, and determining whether there is a sampling point with an amplitude value greater than a preset threshold; If yes, it is determined that the multi-channel time sequence data has potential defects at the sampling time point; If no, it is determined that the multi-channel time sequence data has no potential defects at the sampling time point.

4. The switchgear vibration signal positioning method according to claim 3, characterized in that, The extracting a signal segment with potential defects from the multi-channel time sequence data according to the detection result to generate a vibration event data packet corresponding to each vibration signal comprises: Determining a sampling time point with potential defects according to the detection result, and intercepting a signal segment of each vibration signal at the sampling time point in the multi-channel time sequence data based on a preset interception range; Generating a vibration event data packet corresponding to each vibration signal according to each signal segment.

5. The switchgear vibration signal positioning method of claim 1, wherein, The extracting a signal first arrival time of each vibration event data packet by using a convolution iterative filtering algorithm comprises: For each vibration event data packet, performing Savitzky-Golay smoothing convolution filtering on the vibration event data packet, and identifying a signal starting window in the vibration event data packet according to a filtering result; Performing sliding average processing on the signal starting window by using a moving average method to construct an energy growth curve; the energy growth curve is used to show changes of signal energy over time; Marking a time point corresponding to a first signal energy exceeding a preset noise threshold in the energy growth curve as a signal first arrival time of the vibration event data packet.

6. The switchgear vibration signal positioning method of claim 1, wherein The determining a finite difference time domain simulation model comprises: An opening switch cabinet structure diagram is acquired, a spatial grid is divided according to the switch cabinet structure diagram, and an initial simulation model is generated according to the plurality of spatial grids obtained by the division; A virtual defect source point in each spatial network is determined, and a unit excitation is injected into each virtual defect source point to simulate the propagation time from each virtual defect source point to each vibration sensor; A virtual time difference of each virtual defect source point is constructed based on the propagation time of each virtual defect source point, and each virtual time difference is labeled in the corresponding spatial grid in the initial simulation model to form a finite difference time domain simulation model.

7. The switchgear vibration signal positioning method of claim 6, wherein, The time difference analysis of each signal first arrival time based on the finite difference time domain simulation model obtains the defect vibration source point position of the switch cabinet, including: A real time difference is constructed according to each signal first arrival time, and the Euclidean distance between the real time difference and each virtual time difference in the finite difference time domain simulation model is determined; The virtual defect source point corresponding to the virtual time difference with the minimum Euclidean distance is marked as the defect vibration source point position of the switch cabinet.

8. A switchgear vibration signal positioning device, characterized in that, Including: A signal acquisition module is configured to acquire vibration signals generated during the operation of the switch cabinet from multiple directions to form multi-channel time sequence data; The multi-channel time sequence data is composed of a plurality of vibration signals; A signal detection module is configured to detect the amplitude of the multi-channel time sequence data to obtain a detection result, and extract a signal segment with a potential defect from the multi-channel time sequence data according to the detection result to generate vibration event data packets corresponding to each vibration signal; A time extraction module is configured to extract the signal first arrival time of each vibration event data packet using a convolution iterative filtering algorithm; A source point analysis module is configured to determine a finite difference time domain simulation model, and perform time difference analysis on each signal first arrival time based on the finite difference time domain simulation model to obtain the defect vibration source point position of the switch cabinet.

9. A storage medium characterized by: The storage medium stores computer readable instructions, and the computer readable instructions are executed by one or more processors to cause the one or more processors to perform the steps of the switch cabinet vibration signal positioning method according to any one of claims 1 to 7.

10. A computer device, comprising: Including: One or more processors and a memory; The memory stores computer readable instructions, and the computer readable instructions are executed by the one or more processors to perform the steps of the switch cabinet vibration signal positioning method according to any one of claims 1 to 7.

Images