Breaker contact ablation prediction method, system, equipment and medium
By collecting spectral data of the electric arc between circuit breaker contacts, extracting wavelength and time domain parameters using a non-orthogonal Gaussian processor, and combining this with a linear regression model, the problem of large prediction errors in circuit breaker contact ablation in existing technologies has been solved, achieving higher-precision prediction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA THREE GORGES RENEWABLES YANGJIANG POWER CO LTD
- Filing Date
- 2025-12-29
- Publication Date
- 2026-05-05
AI Technical Summary
Existing methods ignore the influence of material elements in the arc spectrum when predicting circuit breaker contact erosion, resulting in prediction errors exceeding 30% under high current conditions, which cannot meet the accuracy requirements of condition-based maintenance of equipment.
Spectral data of the electric arc between circuit breaker contacts are collected, processed by a non-orthogonal Gaussian processor, and wavelength and time domain parameters are extracted. A preset linear regression model is then used to predict contact quality loss.
It improves the accuracy of circuit breaker contact erosion prediction, reduces prediction errors under high current conditions, meets the accuracy requirements of equipment condition-based maintenance, and has high versatility.
Smart Images
Figure CN121978471A_ABST
Abstract
Description
Technical Field
[0001] This invention application relates to the field of circuit breaker contact erosion prediction, and more particularly to a method, system, device and medium for predicting circuit breaker contact erosion. Background Technology
[0002] High-voltage SF6 circuit breakers are critical protection devices in power systems. During current interruption, the arc contacts are subjected to the energy of the electric arc, resulting in material vaporization and droplet splashing, leading to mass loss and surface morphology deterioration, ultimately reducing the equipment's breaking capacity. Current research on contact ablation prediction shows a development trend based on multi-source data acquisition and fusion, with advanced algorithm models at its core.
[0003] Currently, existing methods mainly rely on electrical or mechanical parameters to indirectly estimate the amount of ablation. This method of predicting contact ablation ignores the influence of material elements in the arc spectrum, and the prediction error exceeds 30% under high current conditions, failing to meet the accuracy requirements of condition-based maintenance. Therefore, the prediction of contact ablation suffers from low accuracy. Summary of the Invention
[0004] This invention application provides a method, system, device, and medium for predicting circuit breaker contact erosion, in order to solve the technical problem of how to improve the accuracy of predicting circuit breaker contact erosion.
[0005] To address the aforementioned technical problems, this invention provides a method for predicting circuit breaker contact erosion, comprising: Collect spectral data of the electric arc between the circuit breaker contacts; The spectral data is processed by a first non-orthogonal Gaussian processor to obtain a first processing result; and wavelength domain parameters are extracted based on the first processing result; wherein the first non-orthogonal Gaussian processor is a wavelength domain processor; The wavelength domain parameters are processed by a second non-orthogonal Gaussian processor to obtain a second processing result; and the time domain parameters are extracted based on the second processing result; wherein the second non-orthogonal Gaussian processor is a time domain processor; Obtain the mass loss parameters corresponding to the circuit breaker contacts; select solution parameters from the time domain parameters and the wavelength domain parameters, substitute the mass loss parameters and the solution parameters into a preset linear regression model, obtain the contact mass loss of the circuit breaker contacts, and realize the prediction of circuit breaker contact erosion.
[0006] As a preferred embodiment, the preset linear regression model includes: ; Where C is a constant, β n P is the regression coefficient. nTo solve for the parameters, where n is the number of parameters to solve for. This refers to the loss of contact quality.
[0007] As a preferred embodiment, the extraction of wavelength domain parameters based on the first processing result includes: Perform an xyz transformation on the first processing result to obtain the relative intensity; The first processing result is subjected to HLS transformation to obtain the parameters: dominant wavelength, effective signal strength, and equivalent signal width.
[0008] As a preferred embodiment, the step of processing the spectral data using a first non-orthogonal Gaussian processor to obtain a first processing result specifically involves: Obtain the first response curves of the three first non-orthogonal Gaussian processors with respect to the parameter domain of the spectral data; The spectral data is then processed using the first response curve to obtain the outputs of three first non-orthogonal Gaussian processors. The outputs of the three first non-orthogonal Gaussian processors are determined as the first processing result.
[0009] As a preferred embodiment, the three first non-orthogonal Gaussian processors have different center wavelengths and are evenly distributed within a preset wavelength range.
[0010] As a preferred embodiment, the step of processing the wavelength domain parameters using a second non-orthogonal Gaussian processor to obtain a second processing result specifically involves: Obtain the second response curves of the three second non-orthogonal Gaussian processors in the parameter domain where the wavelength domain parameters are located; wherein the peak times of the three second non-orthogonal Gaussian processors are different; The wavelength domain parameters are then processed using the second response curve to obtain the outputs of three second non-orthogonal Gaussian processors. The outputs of the three second non-orthogonal Gaussian processors are determined as the second processing result.
[0011] As a preferred embodiment, the acquisition of spectral data of the electric arc between the circuit breaker contacts includes: Sensor data of electric arc between circuit breaker contacts is collected using fiber optic sensors. The spectral data is obtained by analyzing the sensor data using a spectrometer.
[0012] Accordingly, this invention application also provides a prediction system for circuit breaker contact erosion, comprising a spectral data acquisition module, a wavelength domain parameter extraction module, a time domain parameter extraction module, and a prediction module; wherein, The spectral data acquisition module is used to acquire the spectral data of the electric arc between the circuit breaker contacts; The wavelength domain parameter extraction module is used to process the spectral data through a first non-orthogonal Gaussian processor to obtain a first processing result; and to extract wavelength domain parameters based on the first processing result; wherein the first non-orthogonal Gaussian processor is a wavelength domain processor; The time-domain parameter extraction module is used to process the wavelength-domain parameters through a second non-orthogonal Gaussian processor to obtain a second processing result; and to extract time-domain parameters based on the second processing result; wherein the second non-orthogonal Gaussian processor is a time-domain processor; The prediction module is used to obtain the mass loss parameters corresponding to the circuit breaker contacts; select solution parameters from the time domain parameters and the wavelength domain parameters; substitute the mass loss parameters and the solution parameters into a preset linear regression model to obtain the contact mass loss of the circuit breaker contacts, thereby realizing the prediction of circuit breaker contact erosion.
[0013] As a preferred embodiment, the preset linear regression model includes: ; Where C is a constant, β n P is the regression coefficient. n To solve for the parameters, where n is the number of parameters to solve for. This refers to the loss of contact quality.
[0014] As a preferred embodiment, the wavelength domain parameter extraction module extracts wavelength domain parameters based on the first processing result, including: The wavelength domain parameter extraction module performs an xyz transformation on the first processing result to obtain the relative intensity. The first processing result is subjected to HLS transformation to obtain the parameters: dominant wavelength, effective signal strength, and equivalent signal width.
[0015] As a preferred embodiment, the wavelength domain parameter extraction module processes the spectral data using a first non-orthogonal Gaussian processor to obtain a first processing result, specifically: The wavelength domain parameter extraction module obtains the first response curves of the three first non-orthogonal Gaussian processors with respect to the parameter domain in which the spectral data is located; The spectral data is then processed using the first response curve to obtain the outputs of three first non-orthogonal Gaussian processors. The outputs of the three first non-orthogonal Gaussian processors are determined as the first processing result.
[0016] As a preferred embodiment, the three first non-orthogonal Gaussian processors have different center wavelengths and are evenly distributed within a preset wavelength range.
[0017] As a preferred embodiment, the time-domain parameter extraction module processes the wavelength-domain parameters using a second non-orthogonal Gaussian processor to obtain a second processing result, specifically: The time-domain parameter extraction module obtains the second response curves of the three second non-orthogonal Gaussian processors in the parameter domain where the wavelength-domain parameters are located; wherein, the peak times of the three second non-orthogonal Gaussian processors are different; The wavelength domain parameters are then processed using the second response curve to obtain the outputs of three second non-orthogonal Gaussian processors. The outputs of the three second non-orthogonal Gaussian processors are determined as the second processing result.
[0018] As a preferred embodiment, the spectral data acquisition module acquires the spectral data of the electric arc between the circuit breaker contacts, including: The spectral data acquisition module acquires sensor data of the electric arc between the circuit breaker contacts through an optical fiber sensor. The spectral data is obtained by analyzing the sensor data using a spectrometer.
[0019] Accordingly, this application also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the method for predicting circuit breaker contact erosion.
[0020] Accordingly, this application also provides a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the circuit breaker contact erosion prediction method.
[0021] Compared with the prior art, this invention application has the following beneficial effects: This invention provides a method, system, device, and medium for predicting circuit breaker contact erosion. The prediction method includes: acquiring spectral data of an electric arc between circuit breaker contacts; processing the spectral data using a first non-orthogonal Gaussian processor to obtain a first processing result; and extracting wavelength domain parameters based on the first processing result; wherein the first non-orthogonal Gaussian processor is a wavelength domain processor; processing the wavelength domain parameters using a second non-orthogonal Gaussian processor to obtain a second processing result; and extracting time domain parameters based on the second processing result; wherein the second non-orthogonal Gaussian processor is a time domain processor; obtaining mass loss parameters corresponding to the circuit breaker contacts; selecting solution parameters from the time domain parameters and the wavelength domain parameters; and substituting the mass loss parameters and the solution parameters into a preset linear regression model to obtain the contact mass loss of the circuit breaker contacts, thereby predicting circuit breaker contact erosion. This invention collects spectral data of the arc between circuit breaker contacts, processes it using a non-orthogonal Gaussian processor to extract wavelength domain parameters, and then extracts second-order time domain parameters based on these parameters. Solving parameters are selected from both wavelength and time domain parameters and substituted into a preset linear regression model for predicting mass loss. Compared to existing technologies that rely on electrical or mechanical parameters, this invention effectively mines characteristic spectral line information of material elements in the spectrum. By analyzing the spectral information, the degree of material ablation can be inferred, thereby effectively improving the accuracy of circuit breaker contact ablation prediction and reducing prediction errors under complex operating conditions such as high current. This meets the accuracy requirements of condition-based maintenance. Furthermore, the technical solution of this application does not require remodeling for different models or types of circuit breakers, exhibiting high versatility. Attached Figure Description
[0022] Figure 1 This is a flowchart illustrating an embodiment of the method for predicting circuit breaker contact erosion provided in this invention application.
[0023] Figure 2 This is a schematic diagram illustrating the variation of wavelength domain parameters of the copper contact provided in this invention application during circuit breaker ablation.
[0024] Figure 3 This is a schematic diagram illustrating the variation of wavelength domain parameters of the copper contact provided in this invention application during circuit breaker ablation.
[0025] Figure 4 This is a schematic diagram illustrating the variation of wavelength domain parameters of the copper contact provided in this invention application during circuit breaker ablation.
[0026] Figure 5 This is a schematic diagram illustrating the variation of the second-order time-domain parameters of the copper contact provided in this invention application during circuit breaker ablation.
[0027] Figure 6 This is a schematic diagram illustrating the variation of the second-order time-domain parameters of the copper contact provided in this invention application during circuit breaker ablation.
[0028] Figure 7 This is a schematic diagram illustrating the variation of the second-order time-domain parameters of the copper-tungsten contact provided in this invention application during circuit breaker ablation.
[0029] Figure 8 This is a schematic diagram of an embodiment of the circuit breaker contact erosion prediction system provided in this application. Detailed Implementation
[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0031] Example 1 According to relevant technical records, high-voltage SF6 circuit breakers are critical protection equipment in power systems. During current interruption, the arcing contacts are subjected to arc energy, resulting in material vaporization and droplet splashing, leading to mass loss and surface morphology deterioration, ultimately reducing the equipment's breaking capacity. Current research on contact ablation prediction shows a development trend based on multi-source data acquisition and fusion, with advanced algorithm models at its core. At the multi-source data level, relevant technologies collect multi-dimensional signals such as opening and closing coil current, vibration, stroke, radio frequency, angular displacement, field emission current, arc-extinguishing chamber voltage, dynamic capacitance, insulation resistance, and temperature rise. Through preprocessing methods such as time-scale alignment, sampling rate normalization, adaptive filtering, and wavelet transform, a multi-information fusion feature set is constructed to provide comprehensive data support for prediction.
[0032] At the algorithm model level, the relevant technologies utilize multilayer perceptrons, attention-based long short-term memory neural networks, lightweight convolutional recurrent neural networks (CRNN), backpropagation neural networks, RVM algorithms, and double exponential smoothing combined with trend analysis, along with optimization strategies such as mechanical correction factors, environmental corrosion indices, and fuzzy logic, to improve the accuracy of ablation degree classification and remaining life prediction.
[0033] However, the aforementioned technologies also have significant shortcomings: In terms of model universality, some methods rely on specific equipment structural parameters (such as two-dimensional axisymmetric simulation models and pre-breakdown characteristics of vacuum interrupters), and different models or types of circuit breakers need to be remodeled, which limits their versatility; In terms of computational complexity, models such as multilayer perceptrons and CRNNs contain multiple modules and a large number of learnable parameters, which require cloud computing power support and make it difficult to achieve millisecond-level real-time prediction on edge devices.
[0034] Currently, existing methods mainly rely on electrical or mechanical parameters to indirectly estimate the amount of ablation. This method of predicting contact ablation ignores the influence of material elements in the arc spectrum, and the prediction error exceeds 30% under high current conditions, failing to meet the accuracy requirements of condition-based maintenance. Therefore, the prediction of contact ablation suffers from low accuracy.
[0035] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the method for predicting circuit breaker contact erosion provided in this invention application.
[0036] The method for predicting circuit breaker contact erosion can be applied to computer equipment, including but not limited to smartphones, laptops, tablets, desktop computers, and physical servers and cloud servers connected to display units. The display unit can be used to display the prediction results in real time, allowing relevant personnel to be aware of the erosion status of the circuit breaker contacts.
[0037] Figure 1 The illustrated embodiment includes steps S101 to S104; each step is described in detail below: Step S101: Collect spectral data of the electric arc between the circuit breaker contacts.
[0038] During the current breaking process of a circuit breaker, a high-temperature electric arc is generated between the contacts. The arc plasma contains vaporized components of the contact materials (such as copper, tungsten, etc.), which emit light of specific wavelengths when excited at high temperatures, forming an arc spectrum.
[0039] Spectral data reflects the intensity and distribution of characteristic spectral lines of different elements in the electric arc, and is directly related to the ablation state of the contact material. For example, copper atoms have significant spectral lines at about 550 nanometers, and tungsten also has its own unique spectral lines. The intensity and changes of these spectral lines can be used to infer the degree of material ablation.
[0040] In some embodiments, a quartz optical window may be opened on the side wall of the arc-extinguishing cavity for high-voltage SF6 circuit breakers or vacuum circuit breakers. Sensor data of the electric arc between the circuit breaker contacts is collected using a fiber optic sensor; the sensor data is then analyzed using a spectrometer to obtain the spectral data.
[0041] Three of the aforementioned fiber optic sensors can be deployed at 120-degree angles to each other, ensuring no blind spots in the arc root region covered by the electric arc. The fiber optic sensors can be connected to a high-resolution spectrometer, and the synchronous triggering unit is connected to the circuit breaker's opening and closing coil control circuit. The timing of spectral acquisition and circuit breaker operation is synchronized via a TTL level signal, with a synchronization accuracy ≤100 nanoseconds.
[0042] The time-resolved spectrum of the electric arc is captured using a spectrometer. The arc spectrum S(λ, t) is a function of both wavelength λ and time t. Therefore, by analyzing the time-varying spectrum of the arc, we can obtain information not only about the arc components (wavelength domain) but also about the changes in the arc components over time (time domain).
[0043] Step S102: The spectral data is processed by a first non-orthogonal Gaussian processor to obtain a first processing result; and wavelength domain parameters are extracted based on the first processing result.
[0044] In this step, the first non-orthogonal Gaussian processor is a wavelength domain processor, which can process the spectral data through three first non-orthogonal Gaussian processors with different center wavelengths and uniformly distributed within a preset wavelength range to obtain the first processing result.
[0045] For example, three first non-orthogonal Gaussian processors can be represented as (R w,ev G w,ev B w,ev (R, G, B), where the subscript w represents wavelength and ev represents uniformity. In some examples, the center wavelengths of the three first non-orthogonal Gaussian processors can be located at 475nm, 550nm, and 625nm, respectively, with the overlap of adjacent processors being 50% of the peak value, covering a wavelength range of 200-800nm.
[0046] The step of processing the spectral data using a first non-orthogonal Gaussian processor to obtain a first processing result specifically involves: acquiring the first response curves of three first non-orthogonal Gaussian processors with respect to the parameter domain of the spectral data; processing the spectral data using the first response curves to obtain the outputs of the three first non-orthogonal Gaussian processors; and determining the outputs of the three first non-orthogonal Gaussian processors as the first processing result.
[0047] The output of the three first non-orthogonal Gaussian processors can be expressed as: ; ; ; Where R0, G0, and B0 represent the outputs of the first non-orthogonal Gaussian processors R, G, and B, respectively, p is the parameter domain in which the spectral data is located, S(p) represents the spectral data, and R(p), G(p), and B(p) are the first response curves of the first non-orthogonal Gaussian processors R, G, and B, respectively.
[0048] For the first processing result, its wavelength domain parameters can be extracted. In a preferred embodiment, extracting the wavelength domain parameters based on the first processing result includes: performing an xyz transform on the first processing result to obtain the relative intensity x. w y w z w Perform HLS transformation on the first processing result to obtain the dominant wavelength H. w Effective signal strength L w And equivalent signal width 1-S w .
[0049] When obtaining the above wavelength domain parameters, the chromatic method can be used. The chromatic method originates from people's understanding of the photic field and color science. The perception of light by mammals is a typical example: the three types of cone cells in the human eye are sensitive to light of different wavelengths. Without precise measurement, the brain processes the output of these three types of cone cells in real time and can easily identify different colors (the basic idea of the chromatic method).
[0050] Although the chromatic method was originally derived from the processing of optical signals, its applications extend far beyond optics. To date, the chromatic method has been successfully applied in various fields, including plasma monitoring, transformer oil degradation monitoring, radio frequency signal analysis, image recognition, location and behavior monitoring, and event monitoring.
[0051] In short, the chromatic method uses comparison to uncover mathematical cross-correlation between processed data. Through chromatic processing, unexpected events in the monitored system can be easily identified, the system's state can be quantitatively assessed, and effective information can be conveyed intuitively, rather than simply providing data.
[0052] The wavelength domain parameters in this embodiment can also be understood as chromatic parameters. When the original signal is a function of multiple variables, performing chromatic processing on different variables sequentially is called higher-order chromatic processing. Each chromatic processing compresses the amount of information in the corresponding data domain, achieving dimensionality reduction. Unlike machine learning methods such as neural networks, data processed by chromatic processing has retrospective properties, and the characteristics and trends of lower-order data can be derived by analyzing the higher-order chromatic parameters.
[0053] like Figure 2 As shown, due to the existence of relation x w +y w +z w =1, parameter z w The changing trend is already implied in this graph. Therefore, a single rectangular coordinate graph can simultaneously express x. w y w z w The trends of these three chromatic parameters are shown in the graph, where the color of the scatter points represents the mass loss of the arc contact. As can be seen from the graph, x... w (x in the diagram) w (tp)) and y w (The corresponding y in the figure) w The values of (tp) all decrease as the contact mass loss increases.
[0054] Figure 3 H is shown w (corresponding to H in the diagram) w (t p )) and L w (corresponding to L in the diagram) w (t p The trend of change with contact mass loss. When the contact mass loss is small, the effective signal intensity L of the arc spectrum w The value is very low, close to zero. As the contact mass loss increases, the effective signal intensity L of the arc spectrum decreases. w This increases accordingly. Furthermore, when the contact mass loss is high, the dominant wavelength H of the arc spectrum... w It tends to increase with the increase of contact quality loss (the direction is indicated by the solid arrow in the figure).
[0055] Figure 4 H is shown w (corresponding to H in the diagram) w (t p )) and 1-S w (corresponding to 1-S in the figure) w (t pThe graph shows the trend of change with contact mass loss. It can be seen from the figure that when the contact mass loss is small, the dominant wavelength H... w Located in the range of 60 to 90 degrees. As contact mass loss increases, H... w It shifts to a range of approximately 180 to 210 degrees. The equivalent signal width is 1-S. w Within different ranges of mass loss, the value decreases as the contact mass loss increases.
[0056] Step S103: The wavelength domain parameters are processed by a second non-orthogonal Gaussian processor to obtain a second processing result; and the time domain parameters are extracted based on the second processing result.
[0057] In this step, the second non-orthogonal Gaussian processor is a time-domain processor. The wavelength-domain parameters (the aforementioned relative intensity x) can be processed using a second non-orthogonal Gaussian processor with three peak times representing the initial, middle, and later stages of arcing. w y w z w Parameters: dominant wavelength H w Effective signal strength L w And equivalent signal width 1-S w The process is performed to obtain a second processing result. The peak time settings of the three second non-orthogonal Gaussian processors can cover the entire arc burning cycle, with the overlap area of adjacent processors being 50% of the peak value.
[0058] Preferably, the step of processing the wavelength domain parameters through a second non-orthogonal Gaussian processor to obtain a second processing result specifically involves: obtaining second response curves of the three second non-orthogonal Gaussian processors in the parameter domain where the wavelength domain parameters are located; wherein the peak times of the three second non-orthogonal Gaussian processors are different; and processing the wavelength domain parameters through the second response curves to obtain the outputs of the three second non-orthogonal Gaussian processors; and determining the outputs of the three second non-orthogonal Gaussian processors as the second processing result.
[0059] Furthermore, the second processing result can be processed using the chromatic method and subjected to xyz transform and HLS transform to obtain time-domain parameters (relative intensity x). t y t z t Parameters: dominant wavelength H t Effective signal strength L t And equivalent signal width 1-S t The time-domain analysis in step S103 is an analysis of the variation of the wavelength-domain parameters over time.
[0060] like Figure 5 As shown, with x t (corresponding to x in the figure)t,xw ) and y t (corresponding to y in the figure) t,xw The figure (x) represents the ablation characteristics of copper contacts in nitrogen (N2). t With y t Representing wavelength domain parameters x w The relative intensity in the early and middle stages of the arc. As can be seen from the figure, x increases with increasing contact mass loss. t y keeps increasing t It continues to decrease.
[0061] like Figure 6 The image shows H t (corresponding to H in the diagram) t,hw ) and L t (corresponding to L in the diagram) t,hw The symbol represents the ablation characteristics of copper contacts in nitrogen (N2), where H... t,hw Indicates the dominant wavelength H w The relative timing of the occurrence of the maximum value, L t,hw Indicates the dominant wavelength H w The equivalent signal intensity. As shown in the figure, the chromatic parameter shifts along the direction of the arrow as the contact mass loss increases. For experiments with small contact mass loss, H... t,hw The values are roughly distributed between 240 degrees and 360 degrees, representing H. w Two equivalent peak values occur, one in the early stage and the other in the late stage of the arc. Meanwhile, L... t,hw The relatively low value indicates that the dominant wavelength is located in R for most of the time. w,ew and G w,ev Within the covered area (H) w (The value is relatively low). As the contact mass loss increases, H... t,hw Moving clockwise indicates H w The peak time is occurring earlier.
[0062] For copper-tungsten contacts, such as Figure 7 As shown, through correlation analysis, parameter L t (corresponding to L in the diagram) t,lw ) and 1-S t (corresponding to 1-S in the figure) t,yw ) represent the effective spectral intensity and parameter y, respectively. w The equivalent signal width has a strong linear correlation with the quality loss of the copper-tungsten arc contact. The time-domain parameter L... t,lw and 1-S t,yw Plotted on the same graph, the relationship between these two parameters and the mass loss of the copper-tungsten contact ablation can be clearly seen.
[0063] Step S104: Obtain the mass loss parameters corresponding to the circuit breaker contacts; select solution parameters from the time domain parameters and the wavelength domain parameters, substitute the mass loss parameters and the solution parameters into a preset linear regression model, obtain the contact mass loss of the circuit breaker contacts, and realize the prediction of circuit breaker contact erosion.
[0064] In this step, several solution parameters with high correlation can be selected from the time domain parameters and wavelength domain parameters based on their correlation.
[0065] Preferably, the preset linear regression model includes: ; Where C is a constant, β n P is the regression coefficient. n To solve for the parameters, where n is the number of parameters to solve for. This refers to the loss of contact quality.
[0066] The preset linear regression model can be pre-built by fitting experimental data.
[0067] In this embodiment, when using SPSS analysis to determine the linear regression model to be used, the regression coefficients and constants (the above-mentioned mass loss parameters) corresponding to the circuit breaker can be obtained by fitting. Substituting the contact mass loss parameters and solution parameters (selecting time domain parameters or wavelength domain parameters with high correlation after analysis) can yield the arc contact mass loss.
[0068] For copper-tungsten contacts, comparing the standard error of the estimate (SEE) of the traditional arc energy method and the prediction method of this embodiment, the SEE value obtained by the traditional arc energy method is 66 mg, while the SEE value obtained by the prediction method of this embodiment using wavelength domain parameters is 51.3 mg, and the SEE value obtained by the prediction method of this embodiment using time domain parameters is 40.2 mg. Compared with the prior art, this can effectively improve the prediction accuracy of mass loss.
[0069] Accordingly, such as Figure 8 As shown, this invention application also provides a circuit breaker contact erosion prediction system 800, comprising a spectral data acquisition module 801, a wavelength domain parameter extraction module 802, a time domain parameter prediction module 803, and a prediction module 804; wherein, The spectral data acquisition module 801 is used to acquire the spectral data of the electric arc between the contacts of the circuit breaker; The wavelength domain parameter extraction module 802 is used to process the spectral data through a first non-orthogonal Gaussian processor to obtain a first processing result; and to extract wavelength domain parameters based on the first processing result; wherein the first non-orthogonal Gaussian processor is a wavelength domain processor; The time-domain parameter advance module 803 is used to process the wavelength-domain parameters through a second non-orthogonal Gaussian processor to obtain a second processing result; and to extract time-domain parameters based on the second processing result; wherein the second non-orthogonal Gaussian processor is a time-domain processor; The prediction module 804 is used to obtain the mass loss parameters corresponding to the circuit breaker contacts; select solution parameters from the time domain parameters and the wavelength domain parameters; substitute the mass loss parameters and the solution parameters into a preset linear regression model to obtain the contact mass loss of the circuit breaker contacts, thereby realizing the prediction of circuit breaker contact erosion.
[0070] As a preferred embodiment, the preset linear regression model includes: ; Where C is a constant, β n P is the regression coefficient. n To solve for the parameters, where n is the number of parameters to solve for. This refers to the loss of contact quality.
[0071] As a preferred embodiment, the wavelength domain parameter extraction module 802 extracts wavelength domain parameters based on the first processing result, including: The wavelength domain parameter extraction module 802 performs xyz transformation on the first processing result to obtain the relative intensity; The first processing result is subjected to HLS transformation to obtain the parameters: dominant wavelength, effective signal strength, and equivalent signal width.
[0072] As a preferred embodiment, the wavelength domain parameter extraction module 802 processes the spectral data using a first non-orthogonal Gaussian processor to obtain a first processing result, specifically: The wavelength domain parameter extraction module 802 acquires the first response curves of the three first non-orthogonal Gaussian processors with respect to the parameter domain in which the spectral data is located; The spectral data is then processed using the first response curve to obtain the outputs of three first non-orthogonal Gaussian processors. The outputs of the three first non-orthogonal Gaussian processors are determined as the first processing result.
[0073] As a preferred embodiment, the three first non-orthogonal Gaussian processors have different center wavelengths and are evenly distributed within a preset wavelength range.
[0074] As a preferred embodiment, the time-domain parameter advance module 803 processes the wavelength-domain parameters through a second non-orthogonal Gaussian processor to obtain a second processing result, specifically: The time-domain parameter advance module 803 acquires the second response curves of the parameter domain in which the wavelength-domain parameters of the three second non-orthogonal Gaussian processors are located; wherein, the peak times of the three second non-orthogonal Gaussian processors are different; The wavelength domain parameters are then processed using the second response curve to obtain the outputs of three second non-orthogonal Gaussian processors. The outputs of the three second non-orthogonal Gaussian processors are determined as the second processing result.
[0075] As a preferred embodiment, the spectral data acquisition module 801 acquires the spectral data of the electric arc between the circuit breaker contacts, including: The spectral data acquisition module 801 acquires sensor data of the electric arc between the circuit breaker contacts through an optical fiber sensor. The spectral data is obtained by analyzing the sensor data using a spectrometer.
[0076] Accordingly, this application also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the method for predicting circuit breaker contact erosion.
[0077] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal, connecting various parts of the terminal via various interfaces and lines.
[0078] The memory can be used to store the computer program. The processor implements various functions of the terminal by running or executing the computer program stored in the memory and calling the data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.
[0079] Accordingly, this application also provides a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the circuit breaker contact erosion prediction method.
[0080] Wherein, if the modules of the device / terminal equipment / system integration are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, it can implement the steps of the various method embodiments described above. Wherein, the computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.
[0081] Compared with the prior art, this invention application has the following beneficial effects: This invention provides a method, system, device, and medium for predicting circuit breaker contact erosion. The prediction method includes: acquiring spectral data of an electric arc between circuit breaker contacts; processing the spectral data using a first non-orthogonal Gaussian processor to obtain a first processing result; and extracting wavelength domain parameters based on the first processing result; wherein the first non-orthogonal Gaussian processor is a wavelength domain processor; processing the wavelength domain parameters using a second non-orthogonal Gaussian processor to obtain a second processing result; and extracting time domain parameters based on the second processing result; wherein the second non-orthogonal Gaussian processor is a time domain processor; obtaining mass loss parameters corresponding to the circuit breaker contacts; selecting solution parameters from the time domain parameters and the wavelength domain parameters; and substituting the mass loss parameters and the solution parameters into a preset linear regression model to obtain the contact mass loss of the circuit breaker contacts, thereby predicting circuit breaker contact erosion. This invention collects spectral data of the electric arc between circuit breaker contacts, processes it using a non-orthogonal Gaussian processor to extract wavelength domain parameters, and then extracts second-order time domain parameters based on these wavelength domain parameters. Solving parameters are selected from the wavelength and time domain parameters and substituted into a preset linear regression model for predicting mass loss. Compared to existing technologies that rely on electrical or mechanical parameters, this invention effectively mines the characteristic spectral information of material elements in the arc spectrum. By analyzing the spectral information, the degree of material ablation can be inferred, thereby effectively improving the accuracy of circuit breaker contact ablation prediction and reducing prediction errors under complex operating conditions such as high current. This meets the accuracy requirements of equipment condition-based maintenance. Furthermore, the technical solution of this application does not require remodeling for different models or types of circuit breakers, exhibiting high versatility.
[0082] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.
Claims
1. A method for predicting contact erosion in circuit breakers, characterized in that, include: Collect spectral data of the electric arc between the circuit breaker contacts; The spectral data is processed by a first non-orthogonal Gaussian processor to obtain a first processing result; Based on the first processing result, wavelength domain parameters are extracted; wherein, the first non-orthogonal Gaussian processor is a wavelength domain processor; The wavelength domain parameters are processed by a second non-orthogonal Gaussian processor to obtain a second processing result; and the time domain parameters are extracted based on the second processing result; wherein the second non-orthogonal Gaussian processor is a time domain processor; Obtain the mass loss parameters corresponding to the circuit breaker contacts; select solution parameters from the time domain parameters and the wavelength domain parameters, substitute the mass loss parameters and the solution parameters into a preset linear regression model, obtain the contact mass loss of the circuit breaker contacts, and realize the prediction of circuit breaker contact erosion.
2. The method for predicting contact erosion of a circuit breaker as described in claim 1, characterized in that, The preset linear regression model includes: ； Where C is a constant, β n P is the regression coefficient. n To solve for the parameters, where n is the number of parameters to solve for. This refers to the loss of contact quality.
3. The method for predicting contact erosion of a circuit breaker as described in claim 1, characterized in that, The step of extracting wavelength domain parameters based on the first processing result includes: Perform an xyz transformation on the first processing result to obtain the relative intensity; The first processing result is subjected to HLS transformation to obtain the parameters: dominant wavelength, effective signal strength, and equivalent signal width.
4. The method for predicting contact erosion of a circuit breaker as described in claim 1, characterized in that, The first processing result obtained by processing the spectral data using a first non-orthogonal Gaussian processor is as follows: Obtain the first response curves of the three first non-orthogonal Gaussian processors with respect to the parameter domain of the spectral data; The spectral data is then processed using the first response curve to obtain the outputs of three first non-orthogonal Gaussian processors. The outputs of the three first non-orthogonal Gaussian processors are determined as the first processing result.
5. The method for predicting contact erosion of a circuit breaker as described in claim 4, characterized in that, The three first non-orthogonal Gaussian processors have different center wavelengths and are evenly distributed within a preset wavelength range.
6. The method for predicting contact erosion of a circuit breaker as described in claim 1, characterized in that, The second processing result obtained by processing the wavelength domain parameters through a second non-orthogonal Gaussian processor is as follows: Obtain the second response curves of the three second non-orthogonal Gaussian processors in the parameter domain where the wavelength domain parameters are located; wherein the peak times of the three second non-orthogonal Gaussian processors are different; The wavelength domain parameters are then processed using the second response curve to obtain the outputs of three second non-orthogonal Gaussian processors. The outputs of the three second non-orthogonal Gaussian processors are determined as the second processing result.
7. A method for predicting contact erosion of a circuit breaker as described in any one of claims 1 to 6, characterized in that, The spectral data of the electric arc between the circuit breaker contacts collected includes: Sensor data of electric arc between circuit breaker contacts is collected using fiber optic sensors. The spectral data is obtained by analyzing the sensor data using a spectrometer.
8. A prediction system for circuit breaker contact erosion, characterized in that, It includes a spectral data acquisition module, a wavelength domain parameter extraction module, a time domain parameter extraction module, and a prediction module; among which, The spectral data acquisition module is used to acquire the spectral data of the electric arc between the circuit breaker contacts; The wavelength domain parameter extraction module is used to process the spectral data through a first non-orthogonal Gaussian processor to obtain a first processing result; and to extract wavelength domain parameters based on the first processing result; wherein the first non-orthogonal Gaussian processor is a wavelength domain processor; The time-domain parameter extraction module is used to process the wavelength-domain parameters through a second non-orthogonal Gaussian processor to obtain a second processing result; and to extract time-domain parameters based on the second processing result; wherein the second non-orthogonal Gaussian processor is a time-domain processor; The prediction module is used to obtain the mass loss parameters corresponding to the circuit breaker contacts; select solution parameters from the time domain parameters and the wavelength domain parameters; substitute the mass loss parameters and the solution parameters into a preset linear regression model to obtain the contact mass loss of the circuit breaker contacts, thereby realizing the prediction of circuit breaker contact erosion.
9. A terminal device, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the method for predicting circuit breaker contact erosion as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device in which the computer-readable storage medium is located to perform the method for predicting circuit breaker contact erosion as described in any one of claims 1 to 7.