A circuit breaker mechanical wear prediction system based on vibration characteristics

Abstract

The application provides a circuit breaker mechanical wear prediction system based on vibration characteristics, and further comprises a vibration signal acquisition module, a signal processing module, a feature extraction module, a feature screening and fusion module, a control and evaluation module, a correction module and a communication module compared with the prior art. The system acquires the original vibration signal of the circuit breaker in the process of opening and closing, extracts various vibration characteristics after filtering, denoising and time alignment processing, and then obtains the optimal feature combination through the feature screening and fusion algorithm, adopts the random forest to construct the mechanical wear prediction model, introduces the environmental health factor to correct the prediction result, and finally outputs the corrected wear degree prediction value RHI. The system has the advantages of high prediction accuracy, strong environmental adaptability, flexible deployment and the like, solves the problems of unstable performance and large prediction error of the traditional method under complex working conditions, and is suitable for online monitoring and predictive maintenance of the circuit breaker equipment.

Description

Technical Field

[0001] This invention relates to the field of low-voltage electrical appliances, and more particularly to a circuit breaker mechanical wear prediction system based on vibration characteristics. Background Technology

[0002] In power systems, circuit breakers are critical electrical devices widely used for connecting, carrying, and disconnecting circuits. Their primary function is to interrupt current under normal or fault conditions to protect electrical equipment from damage caused by overloads, short circuits, and other faults. Circuit breakers typically use the mechanical contact and separation of moving and stationary contacts to achieve on / off control. When a circuit fault occurs, the circuit breaker can quickly interrupt the current within a specified time, preventing equipment damage or even fires caused by overcurrent, thermal effects, or electric arcs. During frequent opening and closing operations, the moving and stationary contacts not only endure repeated mechanical impacts but also experience electrical stresses such as arc erosion and thermal expansion and contraction, leading to mechanical wear and electrical corrosion. This gradually weakens the contact pressure, affecting the circuit breaker's on / off reliability and long-term stable operation. Therefore, accurately assessing the mechanical wear condition of circuit breakers is crucial for ensuring the safe operation of power systems.

[0003] Our experimental team has long been reviewing and studying a large amount of relevant records and materials on related technologies. Simultaneously, relying on relevant resources and conducting numerous experiments, we discovered existing technologies such as those disclosed in CN110567697B, CN109239591A, CN115508695B, and CN114963953A. For example, one disclosed method for online detection of high-precision circuit breaker contact wear rate includes collecting data on the number of circuit breaker switching cycles under different ambient temperature parameters and voltage values. The collected data is then curve-fitted to obtain temperature and voltage influence factors. Furthermore, the method involves preprocessing and curve fitting of the circuit breaker switching life monitoring data. Line fitting enables contact wear rate detection over a continuous, wide-range current domain. The accuracy of curve fitting is improved through error evaluation. The curve fitting function for circuit breaker contact wear rate data requires further discretization, which must consider both system processor storage space and computational load. Finally, the circuit breaker controller system employs a binary lookup table technique to achieve rapid online lookup calculations for contact wear rate, enabling contact wear rate detection over a continuous, wide-range current domain. This high-precision, continuous, and wide-range online contact wear rate detection technology meets the requirements of low-voltage intelligent power distribution technology for online monitoring of equipment health status, ultimately satisfying the requirements of intelligent operation and maintenance.

[0004] This invention was developed to address the common problems in the field, such as insufficient response to environmental conditions, delayed prediction and lack of trend recognition capabilities, high data processing complexity and poor real-time performance. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of current technology by proposing a circuit breaker mechanical wear prediction system based on vibration characteristics.

[0006] To overcome the shortcomings of the prior art, the present invention adopts the following technical solution: A circuit breaker mechanical wear prediction system based on vibration characteristics, comprising a vibration signal acquisition module, a signal processing module, a feature extraction module, a feature filtering and fusion module, a control and evaluation module, a correction module, and a communication module. The vibration signal acquisition module is used to acquire the original vibration signal of the circuit breaker during the opening and closing process and transmit it to the signal processing module. The signal processing module is used to perform bandpass filtering, noise reduction, and time synchronization alignment on the original vibration signal to obtain a processed signal, and then transmit the processed signal to the feature extraction module. The feature extraction module is used to extract feature data from the processed signal and transmit the feature data to the feature filtering and fusion module. The feature filtering and fusion module is used to filter and fuse features related to the mechanical wear state of the circuit breaker, and transmit the fused feature data to the control and evaluation module. The control and evaluation module is used to predict the mechanical wear of the circuit breaker based on the fused feature data and output a preliminary prediction value. The correction module is used to output the final corrected wear prediction value RHI based on the preliminary prediction value; The communication module is used to send the wear prediction value RHI of the circuit breaker to the user terminal.

[0007] Furthermore, the vibration signal acquisition module acquires the vibration signal through an acceleration sensor installed near the circuit breaker operating mechanism or contact mechanism, and the acceleration sensor signal is preprocessed and output through an analog-to-digital converter (ADC) chip.

[0008] Furthermore, the feature data includes the values ​​of six features of the processed signal, which are numbered 1-6 as follows: peak value P, root mean square value RMS, kurtosis P, dominant frequency fdom, spectral energy Espec, and wavelet packet energy Ewave.

[0009] Furthermore, the feature filtering and fusion module includes a feature filtering unit and a feature fusion unit, wherein the feature filtering unit performs the following steps: S101: Conduct extensive training experiments, using the values ​​of six features from a single training experiment and the actual mechanical wear of the circuit breaker as a sample. Store all samples obtained from the training experiments as a dataset. S102: Generate all non-empty feature subsets containing 1 to 6 features from the six features according to the combination method, generating a total of M=63 feature subsets. All feature subsets are numbered S1, S2, ..., SM, where the m-th feature subset is denoted as Sm, m∈[1,M], and Sm contains k features randomly selected from the six features, where the value of k ranges from 1 to 6. S103: Obtain data for each feature subset from historical data, using the data of each feature subset as input features and the corresponding actual wear level value as the target variable. Train M random forest regression models using the input features and target variable. The random forest regression model trained on the m-th feature subset Sm is denoted as the m-th random forest regression model. S104: Using K test circuit breakers for comparative testing, predict the wear level of each of the K test circuit breakers using M trained random forest regression models, and evaluate the predictive performance of each random forest regression model. S105: Measures the prediction accuracy of M random forest regression models: ; Among them, RS m Let y be the predicted deviation of the m-th random forest regression model, K be the total number of test circuit breakers, i be the i-th test circuit breaker, and y be the predicted deviation of the m-th random forest regression model. ture，i Let y be the actual wear value of the i-th test circuit breaker. m，i Let be the wear level value predicted by the m-th random forest regression model for the i-th test circuit breaker. S106: Select the feature subset with the smallest prediction deviation as the optimal feature subset S, and use the random forest regression model trained with the optimal feature subset S as the input feature as the target module.

[0010] Furthermore, the calculation method for the wear level value (DOW) of the circuit breaker is as follows: , Wherein, Vfir is the current volume of the internal mechanical components of the circuit breaker, and Vorig is the initial volume of the internal components of the circuit breaker. Vfir and Vorig are obtained by industrial CT and non-contact three-dimensional imaging calculation. The internal mechanical components include key components that are prone to structural wear during long-term mechanical operation, such as contact assemblies, closing springs, linkage mechanisms, transmission sliders, and opening springs.

[0011] Furthermore, the feature fusion unit performs the following operational steps: S201: Standardize the values ​​of each feature within the optimal feature subset, and represent the value as the j-th standardized feature in the optimal feature subset. : , Among them, f j μ represents the numerical value of the j-th feature in the dataset within the optimal feature subset, with units depending on the feature type. j Let f be the mean of the j-th feature in the dataset, in units of f. j Consistent, σ j Let f be the standard deviation of the j-th feature in the dataset, in units of f. j Consistent; It is dimensionless; S202: Using the dataset as training input, the fuzzy C-means clustering algorithm is used to analyze the standardized features, resulting in the feature membership matrix U=[u ij Based on this, the fusion weight of each feature in the optimal feature subset is calculated, and with... The fusion weight is represented as w j : , Where Q refers to the total number of samples used for fuzzy clustering analysis in the dataset, expressed in units of samples; u qj This represents the fuzzy membership degree of the q-th sample in the dataset with respect to the j-th feature in the optimal feature subset. It is calculated by the fuzzy C-means clustering algorithm, and its value ranges from [0,1], with dimensionless units. w j This reflects the degree of matching between the feature and the cluster centers in different samples. The larger the value, the higher the consistency or discriminative power of the feature in most samples. Therefore, its weight in the final fused feature vector should also be increased accordingly. j The unit is dimensionless; S203: Fusion weights w obtained in step S202 j Each standardized feature value f in the optimal feature subset S j′ According to the corresponding fusion weight w j We perform a weighted summation to obtain the fusion eigenvalue F: ; Where J is the total number of features in the optimal feature subset, F is the fused feature value, and F is dimensionless, serving as the input variable for the control and evaluation module; S204: Obtain the fusion feature values ​​corresponding to all samples from the dataset, and represent them as F, the fusion feature of the Lth sample. L Let the fusion feature set X represent the fusion features of all samples in the dataset, X={F1,F2…F…}Q Meanwhile, the actual mechanical wear degree of all circuit breakers in the dataset is used as the wear degree set Y, Y={R1,R2…R Q}, where R L The actual mechanical wear of the circuit breaker in the Lth sample.

[0012] The beneficial effects achieved by this invention are: 1. Improve prediction accuracy: By constructing a nonlinear mapping model between fusion features and mechanical wear degree, and training it with the random forest algorithm, the accuracy and stability of circuit breaker wear state prediction are effectively improved, overcoming the problems of strong feature dependence and poor generalization ability in traditional methods.

[0013] 2. Enhanced environmental adaptability: An environmental health factor E, based on factors such as operating time, humidity, and dust exposure, is introduced, and the initial prediction results are corrected by combining wear trend parameters, so that the system still has reliable prediction performance under complex working conditions such as high humidity and high dust.

[0014] 3. Possesses trend recognition capabilities and deployment flexibility: By introducing wear change acceleration and environmental response coefficient, the system can not only assess the current wear level, but also perceive its development trend. Attached Figure Description

[0015] The invention will be further understood from the following description taken in conjunction with the accompanying drawings. The components in the drawings are not necessarily drawn to scale, but rather the emphasis is on illustrating the principles of the embodiments. In different views, the same reference numerals designate corresponding parts.

[0016] Figure 1 This is a modular schematic diagram of the short-circuit breaker mechanical wear prediction module based on vibration characteristics according to the present invention.

[0017] Figure 2 This is a modular schematic diagram of the feature filtering and fusion module of the present invention.

[0018] Figure 3 This is a schematic diagram of a surface experiment for predicting the response of the RHI value under different environmental and wear trend conditions according to the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to its embodiments. It should be noted that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this invention. Other systems, methods, and / or features of this embodiment will become apparent to those skilled in the art after reviewing the following detailed description. Furthermore, the terminology used to describe positional relationships in the accompanying drawings is for illustrative purposes only and should not be construed as limiting this patent. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0020] Example 1: Combined with Appendix Figure 1 Appendix Figure 2 and attached Figure 3 This embodiment constructs a circuit breaker mechanical wear prediction system based on vibration characteristics, including a vibration signal acquisition module, a signal processing module, a feature extraction module, a feature filtering and fusion module, a control and evaluation module, a correction module, and a communication module. The vibration signal acquisition module is used to acquire the original vibration signal of the circuit breaker during the opening and closing process and transmit it to the signal processing module. The signal processing module is used to perform bandpass filtering, noise reduction, and time synchronization alignment on the original vibration signal to obtain a processed signal, and then transmit the processed signal to the feature extraction module. The feature extraction module is used to extract feature data from the processed signal and transmit the feature data to the feature filtering and fusion module. The feature filtering and fusion module is used to filter and fuse features related to the mechanical wear state of the circuit breaker, and transmit the fused feature data to the control and evaluation module. The control and evaluation module is used to predict the mechanical wear of the circuit breaker based on the fused feature data and output a preliminary prediction value. The correction module is used to output the final corrected wear prediction value RHI based on the preliminary prediction value; The communication module is used to send the wear prediction value RHI of the circuit breaker to the user terminal.

[0021] The vibration signal acquisition module acquires the vibration signal through an acceleration sensor installed near the circuit breaker operating mechanism or contact mechanism. The acceleration sensor signal is preprocessed and output through an analog-to-digital converter (ADC) chip.

[0022] The bandpass filter is used to filter out frequency components below 1kHz and above 10kHz in the original vibration signal. The denoising process uses wavelet denoising to remove high-frequency noise or irrelevant interference from the signal. The time synchronization alignment is achieved by using a dynamic time warping algorithm to synchronize vibration signals from different acceleration sensors. The signal processing module is implemented by configuring a digital signal processor (DSP) or an edge computing gateway.

[0023] The feature extraction module is implemented using a signal analysis program developed based on MATLAB or Python. The feature extraction module, feature filtering and fusion module, control and evaluation module, and correction module can be deployed on the same server, or, depending on the actual system architecture and resource configuration needs, deployed on independent edge computing devices or cloud servers. The modules interact with each other through a local area network or industrial communication protocol, exhibiting good system scalability and module independence.

[0024] The feature data includes the values ​​of six features of the processed signal, which are numbered 1-6 as follows: peak value P, root mean square value RMS, kurtosis P, dominant frequency fdom, spectral energy Espec, and wavelet packet energy Ewave.

[0025] The wear level (DOW) value of a circuit breaker is calculated as follows: , Wherein, Vfir is the current volume of the internal mechanical components of the circuit breaker, and Vorig is the initial volume of the internal components of the circuit breaker. Vfir and Vorig are obtained by industrial CT and non-contact three-dimensional imaging calculation. The internal mechanical components include key components that are prone to structural wear during long-term mechanical operation, such as contact assemblies, closing springs, linkage mechanisms, transmission sliders, and opening springs.

[0026] The feature filtering and fusion module includes a feature filtering unit and a feature fusion unit, wherein the feature filtering unit performs the following steps: S101: Conduct extensive experimental training, using the values ​​of six features and the actual mechanical wear of the circuit breaker in one experimental training as a sample, and store all the samples obtained in the experimental training as a dataset; S102: Generate all non-empty feature subsets containing 1 to 6 features from the six features according to the combination method, and generate a total of M=63 feature subsets. All feature subsets are numbered as S1, S2, ..., SM, where the m-th feature subset is denoted as Sm, m∈[1,M], and Sm contains k features randomly selected from the six features, where the value of k ranges from 1 to 6; S103: Obtain the data of each feature subset from the historical data, and use the data of the feature subset as the input feature, and use the corresponding actual wear and tear value as the target variable. Train M random forest regression models through the input feature and the target variable. The random forest regression model trained with the m-th feature subset Sm is denoted as the m-th random forest regression model. S104: Using K test circuit breakers for comparative testing, use M trained random forest regression models to predict the wear degree values ​​of the K test circuit breakers respectively, and evaluate the prediction performance of each random forest regression model; S105: Measures the prediction accuracy of M random forest regression models: ; Among them, RS m Let y be the predicted deviation of the m-th random forest regression model, K be the total number of test circuit breakers, i be the i-th test circuit breaker, and y be the predicted deviation of the m-th random forest regression model. ture，i Let y be the actual wear value of the i-th test circuit breaker. m，i Let be the wear level value predicted by the m-th random forest regression model for the i-th test circuit breaker; S106: Select the feature subset with the smallest prediction deviation as the optimal feature subset S, and use the random forest regression model trained with the optimal feature subset S as the input feature as the target module.

[0027] The feature fusion unit performs the following steps: S201: Standardize the values ​​of each feature within the optimal feature subset, and represent the value as the j-th standardized feature in the optimal feature subset. : , Among them, f j μ represents the numerical value of the j-th feature in the dataset within the optimal feature subset, with units depending on the feature type. j Let f be the mean of the j-th feature in the dataset, in units of f. j Consistent, σ j Let f be the standard deviation of the j-th feature in the dataset, in units of f. j Consistent; It is dimensionless; S202: Using the dataset as training input, the fuzzy C-means clustering algorithm is used to analyze the standardized features, resulting in the feature membership matrix U=[u ij Based on this, the fusion weight of each feature in the optimal feature subset is calculated, and with... The fusion weight is represented as w j : , Where Q refers to the total number of samples used for fuzzy clustering analysis in the dataset, expressed in units of samples; u qj represents the fuzzy membership degree of the qth sample in the dataset on the jth feature in the optimal feature subset. It is calculated by the fuzzy C-means clustering algorithm and its value ranges from [0,1], with the unit being dimensionless. w j This reflects the degree of matching between the feature and the cluster centers in different samples. The larger the value, the higher the consistency or discriminative power of the feature in most samples. Therefore, its weight in the final fused feature vector should also be increased accordingly. j The unit is dimensionless; S203: Fusion weights w obtained in step S202 j Each standardized feature value f in the optimal feature subset S j′ According to the corresponding fusion weight w j We perform a weighted summation to obtain the fusion eigenvalue F: , J is the total number of features in the optimal feature subset, and F is the fused feature value, which is dimensionless and serves as the input variable for the control and evaluation module.

[0028] S204: Obtain the fusion feature values ​​corresponding to all samples from the dataset, and represent them as F, the fusion feature of the Lth sample. L Let the fusion feature set X represent the fusion features of all samples in the dataset, X={F1,F2…F…} Q Meanwhile, the actual mechanical wear degree of all circuit breakers in the dataset is used as the wear degree set Y, Y={R1,R2…R Q}, where R L The actual mechanical wear of the circuit breaker in the Lth sample.

[0029] This invention provides a circuit breaker mechanical wear prediction system based on vibration characteristics, which has significant advantages such as high precision, strong robustness and engineering feasibility. It realizes closed-loop control of the entire process from vibration signal acquisition, signal processing, feature extraction, feature selection and fusion, health assessment, wear correction to remote communication.

[0030] This invention employs a random forest regression method based on multi-feature combination to model the mechanical wear state of circuit breakers. By combining a large amount of experimental data with an optimal feature subset selection mechanism, the generalization ability and prediction accuracy of the model are significantly improved. At the same time, a fuzzy C-means clustering algorithm is introduced to weight the fused features, and combined with the calculation of feature representativeness strength, the discriminative power after feature fusion is effectively improved.

[0031] Example 2: Combined with Appendix Figure 1 Appendix Figure 2 and attached Figure 3 In addition to the content included in the above embodiments, the control and evaluation module also implements the following steps: S301: Receives the fused feature set X and wear set Y output from the feature fusion unit. S302: Using the fused feature set X as input samples and the wear set Y as output samples, a random forest regression algorithm is used to supervise the training of the samples to construct a nonlinear mapping model: Rr=h(Ff), Wherein, the function h(·) is the nonlinear mapping model obtained after training. The nonlinear mapping model can map the fusion feature value of the current test sample to the corresponding mechanical wear degree or comprehensive health index. Ff is the fusion feature of the circuit breaker, and Rr is the preliminary predicted value of the wear degree of the circuit breaker.

[0032] The correction module performs the following steps: S401: To improve the accuracy of health assessments, an environmental health factor E is introduced to adjust the original prediction results. , Where TU is the circuit breaker's service life in hours, TL is the circuit breaker's intended life in hours, HA is the average daily humidity of the current environment in which the circuit breaker is located in (%), DDU is the dust exposure concentration in milligrams per cubic meter, HR is the circuit breaker's rated allowable humidity, DRA is the circuit breaker's rated allowable dust concentration, and A... deg γ represents the wear rate of change, e is the base of the natural logarithm, γ1 is the time load correction factor, γ2 is the humidity exposure correction factor, and γ3 is the dust concentration correction factor. S402: Based on the predicted value Rr and the environmental correction factor E, the final corrected wear prediction value RHI is obtained. , Among them, A deg The calculation formula is as follows: , R -1t R is the predicted wear value measured at a time lagging behind the current time by one monitoring cycle. -2t The wear prediction value is the value measured two monitoring cycles later than the current time. Δt represents the time interval between the two monitoring cycles. α is the environmental correction response coefficient, which is dimensionless and represents the degree of response of environmental factor E to the wear prediction value. β is the wear change trend response coefficient, which is measured per hour and is used to measure the degree of influence of wear change acceleration on the final prediction value.

[0033] The performance evaluation factor E is calculated using a formula that comprehensively quantifies the impact of the current operating environment and historical operating conditions of the circuit breaker on its wear trend. By modeling correction factors based on the actual usage time, humidity exposure level, and dust exposure degree of the circuit breaker, and combining the time load correction coefficient γ1, humidity correction coefficient γ2, and dust concentration correction coefficient γ3, and using the exponential form of the natural logarithm base e to nonlinearly synthesize these effects, the environmental health factor E is finally obtained. The performance evaluation factor E is used to improve the accuracy of mechanical wear prediction and enhance the robustness of the system under complex environmental conditions.

[0034] The coefficients γ1, γ2, and γ3 were obtained by those skilled in the art through extensive experimental measurements, data modeling, and error correction to accurately quantify the impact of environmental and operational factors on the assessment of the circuit breaker's health index. Specifically, these coefficients include: γ1 simulates the operating conditions of circuit breakers under different environmental humidity conditions, records the corresponding mechanical wear process and health index changes, constructs a humidity-degradation relationship model, and compares it with the rated humidity HRA. The influence coefficient of humidity on the health degradation rate is extracted, and the coefficient is calibrated after data fitting and residual analysis so that γ1 can reflect the sensitivity of humidity change to the performance of circuit breakers. By changing the dust particle concentration in the test environment, the influence of foreign matter accumulation in the moving mechanism of the circuit breaker on the mechanical response and wear degree is observed. A nonlinear correlation model between the pollution degree and the health index is constructed by combining the rated dust concentration DRA, and the value of γ2 is determined by fitting, so that γ2 can reasonably reflect the interference intensity of different dust conditions on the system performance. γ3 is calculated by monitoring the health index change trajectory of different circuit breakers in continuous operation cycles over a long period of time. The wear rate decrease between two adjacent tests is used as an indicator of mechanical wear change acceleration. This is compared with the actual upward trend of fault risk, and its influencing factors are fitted. The value of dynamic change correction coefficient γ3 is extracted. γ3 is used to adjust the sensitivity of the health index calculation to the rapid aging trend.

[0035] α is determined by those skilled in the art through prior analysis of multiple circuit breaker samples with different environmental exposures but known wear rates within the same time period, analyzing the influence of performance evaluation factor E on the prediction deviation of actual wear degree, and using the minimum mean square error method to fit and obtain its optimal coefficient value.

[0036] The value of β is established by mapping the wear rate change trend obtained from continuous monitoring among different samples to the actual wear deviation. The optimal value of β is fitted by regression method, so that β can be dynamically adjusted within a preset range according to the type of circuit breaker and operating conditions, thereby improving the sensitivity of the wear trend change to the evaluation results.

[0037] This invention constructs a nonlinear mapping model between fusion characteristics and mechanical wear degree through a control and evaluation module, and combines it with the environmental health factor E and wear change acceleration A introduced in the correction module. deg By dynamically correcting the prediction results, it can achieve high-precision and highly adaptable prediction of the mechanical wear of circuit breakers. It not only effectively improves the prediction accuracy under multiple operating conditions and high-noise environments, but also has the ability to sense wear trends. It solves the problems of poor model generalization, weak environmental adaptability and prediction lag in existing technologies, and has good engineering feasibility and forward-looking operation and maintenance value.

[0038] While the invention has been described above with reference to various embodiments, it should be understood that many changes and modifications can be made without departing from the scope of the invention. That is, the methods, systems, and devices discussed above are examples. Various configurations can be appropriately omitted, substituted, or added to various processes or components. For example, in alternative configurations, methods can be performed in a different order than described, and / or various components can be added, omitted, and / or combined. Moreover, features described with respect to certain configurations can be combined in various other configurations, such as different aspects and elements of the configuration can be combined in a similar manner. Furthermore, the elements therein can be updated as the technology develops; many elements are examples and do not limit the scope of this disclosure or the claims. It should also be understood that after reading the description of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent changes and modifications also fall within the scope defined by the claims of this invention.

Claims

1. A circuit breaker mechanical wear prediction system based on vibration characteristics, characterized in that, The circuit breaker mechanical wear prediction system based on vibration characteristics includes a vibration signal acquisition module, a signal processing module, a feature extraction module, a feature filtering and fusion module, a control and evaluation module, a correction module, and a communication module. The vibration signal acquisition module is used to acquire the original vibration signal of the circuit breaker during the opening and closing process and transmit it to the signal processing module. The signal processing module is used to perform bandpass filtering, noise reduction, and time synchronization alignment on the original vibration signal to obtain a processed signal, and then transmit the processed signal to the feature extraction module. The feature extraction module is used to extract feature data from the processed signal and transmit the feature data to the feature filtering and fusion module. The feature filtering and fusion module is used to filter and fuse features related to the mechanical wear state of the circuit breaker, and transmit the fused feature data to the control and evaluation module. The control and evaluation module is used to predict the mechanical wear of the circuit breaker based on the fused feature data and output a preliminary prediction value. The correction module is used to output the final corrected wear prediction value RHI based on the preliminary prediction value; The communication module is used to send the circuit breaker wear prediction value RHI to the user terminal; The feature filtering and fusion module includes a feature filtering unit and a feature fusion unit. The feature fusion unit performs the following steps: S201: Standardize the values ​​of each feature within the optimal feature subset, and represent the value as the j-th standardized feature in the optimal feature subset. : ， Among them, f j μ represents the numerical value of the j-th feature in the dataset within the optimal feature subset, with units depending on the feature type. j Let f be the mean of the j-th feature in the dataset, in units of f. j Consistent, σ j Let f be the standard deviation of the j-th feature in the dataset, in units of f. j Consistent; It is dimensionless; S202: Using the dataset as training input, the fuzzy C-means clustering algorithm is used to analyze the standardized features, resulting in the feature membership matrix U=[u ij Based on this, the fusion weight of each feature in the optimal feature subset is calculated, and with... The fusion weight is represented as w j : ， Where Q refers to the total number of samples used for fuzzy clustering analysis in the dataset, expressed in units of samples; u qj This represents the fuzzy membership degree of the q-th sample in the dataset with respect to the j-th feature in the optimal feature subset. It is calculated by the fuzzy C-means clustering algorithm, and its value ranges from [0,1], with dimensionless units. w j This reflects the degree of matching between the feature and the cluster centers in different samples. The larger the value, the higher the consistency or discriminative power of the feature in most samples. Therefore, its weight in the final fused feature vector should also be increased accordingly. j The unit is dimensionless; S203: Fusion weights w obtained in step S202 j Each standardized feature value f in the optimal feature subset S j′ According to the corresponding fusion weight w j We perform a weighted summation to obtain the fusion eigenvalue F: ； Where J is the total number of features in the optimal feature subset, F is the fused feature value, and F is dimensionless, serving as the input variable for the control and evaluation module; S204: Obtain the fusion feature values ​​corresponding to all samples from the dataset, and represent them as F, the fusion feature of the Lth sample. L Let the fusion feature set X represent the fusion features of all samples in the dataset, X={F1,F2…F…} Q Meanwhile, the actual mechanical wear degree of all circuit breakers in the dataset is used as the wear degree set Y, Y={R1,R2…R Q }, where R L The actual mechanical wear of the circuit breaker in the Lth sample.

2. The circuit breaker mechanical wear prediction system as described in claim 1, characterized in that, The vibration signal acquisition module acquires the vibration signal through an acceleration sensor installed near the circuit breaker operating mechanism or contact mechanism. The acceleration sensor signal is preprocessed and output through an analog-to-digital converter (ADC) chip.

3. The circuit breaker mechanical wear prediction system as described in claim 1, characterized in that, The feature data includes the values ​​of six features of the processed signal, which are numbered 1-6 as follows: peak value P, root mean square value RMS, kurtosis P, dominant frequency fdom, spectral energy Espec, and wavelet packet energy Ewave.

4. The circuit breaker mechanical wear prediction system as described in claim 1, characterized in that, The feature filtering unit performs the following steps: S101: Conduct extensive training experiments, using the values ​​of six features from a single training experiment and the actual mechanical wear of the circuit breaker as a sample. Store all samples obtained from the training experiments as a dataset. S102: Generate all non-empty feature subsets containing 1 to 6 features from the six features according to the combination method, generating a total of M=63 feature subsets. All feature subsets are numbered S1, S2, ..., SM, where the m-th feature subset is denoted as Sm, m∈[1,M], and Sm contains k features randomly selected from the six features, where the value of k ranges from 1 to 6. S103: Obtain data for each feature subset from historical data, using the data of each feature subset as input features and the corresponding actual wear level value as the target variable. Train M random forest regression models using the input features and target variable. The random forest regression model trained on the m-th feature subset Sm is denoted as the m-th random forest regression model. S104: Using K test circuit breakers for comparative testing, predict the wear level of each of the K test circuit breakers using M trained random forest regression models, and evaluate the predictive performance of each random forest regression model. S105: Measures the prediction accuracy of M random forest regression models: ； Among them, RS m Let y be the predicted deviation of the m-th random forest regression model, K be the total number of test circuit breakers, i be the i-th test circuit breaker, and y be the predicted deviation of the m-th random forest regression model. ture，i Let y be the actual wear value of the i-th test circuit breaker. m，i Let be the wear level value predicted by the m-th random forest regression model for the i-th test circuit breaker. S106: Select the feature subset with the smallest prediction deviation as the optimal feature subset S, and use the random forest regression model trained with the optimal feature subset S as the input feature as the target module.

5. The circuit breaker mechanical wear prediction system as described in claim 4, characterized in that, The wear level (DOW) value of a circuit breaker is calculated as follows: , Wherein, Vfir is the current volume of the internal mechanical components of the circuit breaker, and Vorig is the initial volume of the internal components of the circuit breaker. Vfir and Vorig are obtained by industrial CT and non-contact three-dimensional imaging calculation. The internal mechanical components include key components that are prone to structural wear during long-term mechanical operation, such as contact assemblies, closing springs, linkage mechanisms, transmission sliders, and opening springs.

Images