An online monitoring system and method for intelligent circuit breaker

Abstract

The application discloses an online monitoring system and method of an intelligent circuit breaker, and relates to the field of circuit breaker monitoring. The method comprises the following steps: collecting current, port voltage drop, mechanism acceleration and high-frequency voltage signals, synchronously labeling timing time stamps and monotonic time stamps for each frame of data; calculating contact power micro-perturbation strength and mechanism rebound spectrum difference in a closing steady-state window; calculating an arc precursor energy gate in an opening rising edge window; constructing a multi-domain diagnostic atlas based on a contact entropy flow index, mechanism rebound spectrum difference and arc precursor energy gate, and outputting waveform sections for event root cause determination; generating fragmented summaries and whole-frame checks according to the classification and fragmentation of events, and embedding time stamps and buffer area indexes; performing time sequence rearrangement on cross-link data to ensure data time sequence consistency; maintaining fragmented receiving states and retransmitting missing fragments to complete whole-frame consistency checks. Through data fusion and intelligent fragmentation management, the monitoring accuracy and data integrity are ensured.

Description

Technical Field

[0001] This invention relates to the field of circuit breaker monitoring, specifically to an online monitoring system and method for intelligent circuit breakers. Background Technology

[0002] As crucial protection and control devices in modern power systems, intelligent circuit breakers primarily function to detect and isolate electrical faults, ensuring the safe and stable operation of the system. In smart grids and industrial automation systems, circuit breaker health monitoring is paramount, as equipment failure or performance degradation can impact overall system stability. Existing intelligent circuit breaker monitoring methods typically rely on single monitoring data points, such as current, voltage, or temperature. However, a single signal cannot comprehensively reflect the overall condition of the equipment, especially in complex electrical and mechanical environments where faults are often the result of multiple factors acting together. Therefore, employing a multi-signal fusion approach, analyzing data from multiple dimensions, can more accurately identify the root causes of equipment failures. Traditional data collection and processing methods are often limited to data acquisition and simple analysis, making it difficult to achieve efficient, real-time, and comprehensive equipment health status monitoring. Summary of the Invention

[0003] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide an online monitoring system and method for intelligent circuit breakers to solve the aforementioned technical problems.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an online monitoring method for intelligent circuit breakers, comprising: S1: Collect current signal, port voltage drop signal, mechanism acceleration signal and voltage high frequency signal, synchronously mark the time and monotonic timestamp for each frame of data, obtain the closing steady state window and the opening rising edge window, and store the original waveforms of preset duration before and after the event trigger through the multi-channel ring buffer at the edge end; S2: Calculate the contact power perturbation intensity within the closing steady-state window to obtain the contact entropy flow index, and calculate the difference between the acceleration root spectrum and the reference root spectrum within the mechanism's natural frequency band to obtain the mechanism's springback spectrum difference; S3: Weighted integration and compression mapping of current rate of change and voltage high-frequency envelope are performed within the rising edge window of the interruption to generate arc precursor energy gate; S4: Based on the contact entropy flow index, mechanism springback spectrum difference and arc precursor energy gate, construct a multi-domain diagnostic map, determine the root cause of the event, and output the waveform segment to be retained. S5: Classify and segment waveform segments, generate frame identifiers and segment numbers, calculate segment summaries and full frame verification, and embed timestamps and buffer indices in each segment; S6: Fit the timing and generate a monotonic time base anchor with the monotonic timestamp, and rearrange the time sequence of cross-link data using the monotonic timestamp as a reference; S7: The cloud maintains the segmented reception bitmap of each frame and returns the gap list. The edge end retransmits the missing segments according to the gap list, and the cloud performs consistency verification on the whole frame data.

[0005] The present invention is further configured such that S1 includes: Current signal, port voltage drop signal, mechanism acceleration signal and high-frequency voltage signal are collected in real time at preset sampling frequencies; Each data frame is simultaneously marked with a timing timestamp and a monotonic timestamp. The timing timestamp is provided by an external clock system, and the monotonic timestamp is generated by the internal clock of the device. After closing the circuit, the steady-state window for closing is obtained by monitoring the current signal, port voltage drop signal, mechanism acceleration signal and high-frequency voltage signal; During the switching process, the rising edge window of the switching is obtained by monitoring the current signal and the voltage signal; A multi-channel circular buffer is set at the edge. The buffer capacity is configured to store complete waveform data for a preset duration before and after the event is triggered. The data is written in chronological order and overwritten cyclically to continuously preserve the original waveforms before and after the event.

[0006] The present invention is further configured such that S2 includes: Within the steady-state window of closing, the contact power is calculated based on the current signal and the port voltage drop signal, and the contact current index is generated according to the rate of change of the contact power. Within the preset frequency band of the mechanism, the current power spectrum is calculated based on the mechanism's acceleration signal, and the difference is integrated with the reference root spectrum to obtain the mechanism's rebound spectrum difference.

[0007] The present invention is further configured such that S3 includes: Within the time window of the rising edge of the switching, the rate of change of current and the high-frequency envelope of voltage are calculated based on the current and voltage signals. The current change rate and the high-frequency envelope of the voltage are weighted and integrated, and the result of the weighted integration is compressed and mapped to obtain the arc precursor energy gate.

[0008] The present invention is further configured such that S4 includes: Based on the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate, data fusion is performed to generate a multi-domain diagnostic map. The multi-domain diagnostic map is analyzed according to the preset judgment rules to determine the root cause of the event; Generate a list of original waveform segments to be retained corresponding to the root cause determination results of the event. The list includes current waveforms, port voltage drop waveforms, mechanism acceleration waveforms, and high-frequency voltage signal waveforms.

[0009] The present invention is further configured such that S5 includes: Set a preset quantity threshold and classify events according to the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate; When the values ​​of the contact entropy flow index, the mechanism springback spectrum difference, and the arc precursor energy gate are all greater than the preset high threshold, it is marked as a Class A event, indicating a high-risk anomaly. When any one of the following values ​​is greater than the preset intermediate threshold: contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate, it is marked as a Class B event, indicating a medium-risk anomaly. When the values ​​of the contact entropy flow index, the mechanism springback spectrum difference, and the arc precursor energy gate are all less than or equal to the preset low threshold, it is marked as a Class C event, indicating a low-risk anomaly. The data for each event or period is fragmented according to the event level to generate corresponding group identifiers and fragment numbers; Calculate summary information for each data segment and perform integrity verification on the entire frame of data; Each data segment contains a time stamp, a monotonic timestamp, and a circular buffer index.

[0010] The present invention is further configured such that S6 includes: Collect and use monitoring data within a preset time window before and after the event, fit the time synchronization timestamp with the monotonic timestamp, and obtain the monotonic time base anchor; The combination of monotonic timestamps and monotonic time base anchors is used as a unified sorting benchmark, and cross-link data is rearranged in time according to this benchmark; The monitoring data includes current signals, port voltage drop signals, mechanism acceleration signals, and high-frequency voltage signals.

[0011] The present invention is further configured such that S7 includes: The cloud maintains a bitmap of received fragments for each group of data frames. The bitmap is used to record the reception status of each fragment. Received fragments are marked as received, and unreceived fragments are marked as unreceived. Based on the bitmap, the cloud generates a gap list, which lists all unreceived fragment sequence numbers, and returns the list to the edge. The edge device retransmits the unreceived fragments listed in the gap list; After all fragments have been received, the cloud performs a consistency check on the entire frame of data.

[0012] The present invention is further configured such that the method also includes: displaying the event root cause analysis results and fragment retransmission status through a graphical interface.

[0013] The present invention also provides an online monitoring system for intelligent circuit breakers, the system comprising: Acquisition module: Acquires current signal, port voltage drop signal, mechanism acceleration signal and high-frequency voltage signal, synchronously marks each frame of data with time synchronization and monotonic timestamp, obtains the closing steady state window and the opening rising edge window, and saves the original waveforms before and after the event trigger for a preset duration through a multi-channel ring buffer at the edge end; The first calculation module calculates the contact power perturbation intensity within the closing steady-state window to obtain the contact entropy flow index, and calculates the difference between the acceleration root spectrum and the reference root spectrum within the mechanism's natural frequency band to obtain the mechanism's springback spectrum difference. The second calculation module performs weighted integration and compression mapping on the current rate of change and the high-frequency envelope of voltage within the window of the rising edge of the interruption to generate the arc precursor energy gate. Diagram construction module: Based on the contact entropy flow index, mechanism springback spectrum difference and arc precursor energy gate, construct a multi-domain diagnostic map, determine the root cause of the event, and output the waveform segment to be retained; Data fragmentation module: classifies and fragments waveform segments, generates frame identifiers and fragment sequence numbers, calculates fragment summaries and full frame verification, and embeds timestamps and buffer indices in each fragment; The time-series reordering module: It fits the timing and monotonic timestamps to generate monotonic time base anchors, and uses the monotonic timestamps as a benchmark to reorder the time series of cross-link data; Fragment retransmission module: The cloud maintains the fragmented reception bitmap of each frame group and returns the gap list. The edge end retransmits the missing fragments according to the gap list, and the cloud performs consistency verification on the whole frame data.

[0014] This invention provides an online monitoring system and method for intelligent circuit breakers. The method comprises: S1: acquiring current signals, port voltage drop signals, mechanism acceleration signals, and high-frequency voltage signals; synchronously labeling each frame of data with time synchronization and monotonic timestamps; obtaining the closing steady-state window and the breaking rising edge window; and storing the original waveforms before and after the event trigger for a preset duration using a multi-channel ring buffer at the edge end; S2: calculating the contact power perturbation intensity within the closing steady-state window to obtain the contact entropy current index; and calculating the difference between the acceleration root spectrum and the reference root spectrum within the mechanism's natural frequency band to obtain the mechanism's springback spectrum difference; S3: performing weighted integration and compression mapping on the current change rate and the high-frequency envelope of the voltage within the breaking rising edge window to generate... S4: Based on the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate, construct a multi-domain diagnostic map, determine the root cause of the event, and output the waveform segment to be retained; S5: Classify and segment the waveform segment, generate frame identifiers and segment numbers, calculate segment summaries and whole frame verification, and embed timestamps and buffer indices in each segment; S6: Fit time synchronization and monotonic timestamps to generate monotonic time base anchors, and rearrange the cross-link data timing based on the monotonic timestamps; S7: The cloud maintains the segment reception bitmap of each frame group and returns the gap list. The edge end retransmits the missing segments according to the gap list. The cloud performs consistency verification on the whole frame data. The beneficial effects include: 1. By synchronously labeling and timing high-frequency signals of current, port voltage drop, acceleration, and voltage with monotonic timestamps, it is ensured that data from different signal sources can maintain a consistent time reference, thereby improving the accuracy and precision of real-time monitoring; 2. By calculating the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate, a multi-domain diagnostic map is constructed to achieve data fusion of multiple dimensions such as contacts, mechanism, and arc precursors, comprehensively reflecting the working status of the circuit breaker and providing anomaly detection capabilities; 3. By managing the received bitmap fragments and retransmitting fragments according to the gap list, the integrity and reliability of the data are ensured. The edge end will only retransmit the missing fragments, reducing redundant data transmission and improving data transmission efficiency. The cloud performs consistency verification on the entire frame of data to ensure that all data is received in the correct order and is not lost or tampered with, thus ensuring data integrity.

[0015] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 A flowchart illustrating an online monitoring method for an intelligent circuit breaker, as shown in an exemplary embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the structure of an online monitoring system for an intelligent circuit breaker, as shown in an exemplary embodiment of the present invention. Detailed Implementation

[0017] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0018] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0019] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0020] Example 1:

[0021] An online monitoring method for intelligent circuit breakers, such as Figure 1 As shown, it includes: S1: Collect current signal, port voltage drop signal, mechanism acceleration signal and voltage high frequency signal, synchronously mark the time and monotonic timestamp for each frame of data, obtain the closing steady state window and the opening rising edge window, and store the original waveforms of preset duration before and after the event trigger through the multi-channel ring buffer at the edge end; S2: Calculate the contact power perturbation intensity within the closing steady-state window to obtain the contact entropy flow index, and calculate the difference between the acceleration root spectrum and the reference root spectrum within the mechanism's natural frequency band to obtain the mechanism's springback spectrum difference; S3: Weighted integration and compression mapping of current rate of change and voltage high-frequency envelope are performed within the rising edge window of the interruption to generate arc precursor energy gate; S4: Based on the contact entropy flow index, mechanism springback spectrum difference and arc precursor energy gate, construct a multi-domain diagnostic map, determine the root cause of the event, and output the waveform segment to be retained. S5: Classify and segment waveform segments, generate frame identifiers and segment numbers, calculate segment summaries and full frame verification, and embed timestamps and buffer indices in each segment; S6: Fit the timing and generate a monotonic time base anchor with the monotonic timestamp, and rearrange the time sequence of cross-link data using the monotonic timestamp as a reference; S7: The cloud maintains the segmented reception bitmap of each frame and returns the gap list. The edge end retransmits the missing segments according to the gap list, and the cloud performs consistency verification on the whole frame data.

[0022] The present invention is further configured such that S1 includes: Current signal, port voltage drop signal, mechanism acceleration signal and high-frequency voltage signal are collected in real time at preset sampling frequencies; Each data frame is simultaneously marked with a timing timestamp and a monotonic timestamp. The timing timestamp is provided by an external clock system, and the monotonic timestamp is generated by the internal clock of the device. After closing the circuit, the steady-state window for closing is obtained by monitoring the current signal, port voltage drop signal, mechanism acceleration signal and high-frequency voltage signal; During the switching process, the rising edge window of the switching is obtained by monitoring the current signal and the voltage signal; A multi-channel circular buffer is set at the edge, with the buffer capacity configured to store complete waveform data for a preset duration before and after the event trigger. The data is written in chronological order and cyclically overwritten to continuously preserve the original waveforms before and after the event. Specifically, the edge device collects current signals, port voltage drop signals, mechanical acceleration signals, and high-frequency voltage signals in real time at preset sampling frequencies. Each sensor converts the collected analog signals into digital signals through a high-precision sampling circuit and transmits them to the device's processing unit in real time through a data acquisition card or corresponding interface. The collected signals include current flow, port voltage drop, mechanical motion status, and voltage fluctuations. In each data frame, the edge device synchronously marks two timestamps, one for the synchronization time. The system uses a time stamp and a monotonic timestamp. The time stamp is provided by an external clock system and is usually synchronized with the external clock via a standard time protocol to ensure time consistency across all devices. The monotonic timestamp is generated by the device's internal clock and is unaffected by external clock drift, ensuring the sequential consistency of data frames within the device. After each data frame is acquired, these two timestamps are written to the header of the data frame to ensure the consistency and accuracy of time information. After the closing operation is completed, the circuit breaker's electrical signals will experience brief dynamic fluctuations. To determine the steady-state window after closing, the edge terminal simultaneously acquires current signals, port voltage drop signals, mechanism acceleration signals, and high-frequency voltage signals within a preset time interval after closing, such as 50 milliseconds to 200 milliseconds. Within this time interval, the instantaneous values ​​of the aforementioned signals are continuously monitored. When the fluctuation amplitude of each signal is detected to be lower than the corresponding preset threshold and remains continuously stable within the threshold range, it is determined that the closing steady-state window has been reached. The breaking rising edge time window refers to the interval during which the current and voltage signals undergo drastic changes in a short period of time during the circuit breaker breaking process. Its duration is usually 10 milliseconds to 100 milliseconds. During this period, the rate of change of the current and voltage signals reaches its maximum value. The rate of change of the current and voltage signals is monitored in real time. When the rate of change reaches the set threshold, it is determined that the breaking rising edge time window has been entered. A multi-channel ring buffer is set in the edge device to store the complete waveform data of the preset duration before and after the event trigger. Each signal Each channel has an independent buffer capacity configured to store data for a preset duration before and after an event. For example, when the preset duration is 100 milliseconds, each channel can retain waveform data of 100 milliseconds. Data is written to the buffer in chronological order, and the oldest data is overwritten when the capacity is exceeded to ensure that the latest waveforms before and after the event are always retained. The circular buffer continuously stores the raw data and retains the relevant data for subsequent analysis when the event triggering conditions are met. After the data is written, the data integrity is checked in real time, and the data is segmented and stored before and after the event is triggered. When an event such as closing or opening occurs, the device triggers data acquisition in real time and writes the relevant raw waveform data to the circular buffer to continuously retain data for a preset duration before and after the event.When an event is detected, an event timestamp is marked, and data before and after the event is completely preserved. The start and end times of the closing steady-state window and the opening rising edge window are dynamically adjusted according to the event requirements and used as the time base for subsequent analysis. Synchronously labeled current signals, port voltage drop signals, mechanism acceleration signals, and high-frequency voltage signal data frames, as well as the start and end times of the closing steady-state window and the opening rising edge window, are output for subsequent analysis and processing. Simultaneously, raw waveform data stored in a multi-channel ring buffer at the edge provides data support for subsequent steps.

[0023] The present invention is further configured such that S2 includes: Within the steady-state window of closing, the contact power is calculated based on the current signal and the port voltage drop signal, and the contact current index is generated according to the rate of change of the contact power. Within the preset frequency band of the mechanism, the current power spectrum is calculated based on the mechanism acceleration signal, and the difference is integrated with the reference root spectrum to obtain the mechanism springback spectrum difference. Specifically, the current signal is collected in real time by a current sensor to characterize the current flow during equipment operation; the port voltage drop signal is collected in real time by a port voltage drop sensor to reflect the voltage drop generated when current flows through the circuit breaker port; the mechanism acceleration signal is collected by an acceleration sensor to characterize the motion state of the circuit breaker's mechanical components. These signals are collected at the edge end and simultaneously timestamped, serving as inputs for subsequent contact current index and mechanism springback spectrum difference analysis. Within the closing steady-state window, the contact power is calculated by multiplying the current signal and the port voltage drop signal to characterize the energy consumption of the circuit breaker contact area. Since the current signal and the port voltage drop signal tend to stabilize within the closing steady-state window, the obtained contact power can reflect the power consumption of the circuit breaker in a steady state. The contact current index is calculated based on the rate of change of contact power, and the contact current index is used to reflect the contact area. The system analyzes the perturbation intensity in the control zone; monitors the real-time changes in contact power within the steady-state closing window and calculates the change rate per unit time to obtain the contact power change rate, thereby calculating the contact current index. This contact current index reflects whether there is excessive fluctuation or non-uniformity in the contact area and is output as a characterization index of the stability and health status of the contact area. Within the inherent frequency band of the mechanism, the system calculates the current power spectrum using the collected mechanism acceleration signal and uses it to characterize the energy distribution of the mechanism at different frequencies. By performing frequency domain analysis on the mechanism acceleration signal, the system uses Fast Fourier Transform or other spectral analysis methods to convert the time domain signal into a frequency domain signal and obtains the current power spectrum to reflect the vibration and rebound state of the mechanism within the steady-state closing window. The system integrates the difference between the calculated current power spectrum and the reference root spectrum obtained through long-term operation data analysis to obtain the mechanism rebound spectrum difference, which reflects the degree of abnormality in the mechanism rebound. When the mechanism rebound spectrum difference increases, it indicates that the mechanism rebound behavior deviates from the normal state and there is a mechanical fault or performance degradation.

[0024] The present invention is further configured such that S3 includes: Within the time window of the rising edge of the switching, the rate of change of current and the high-frequency envelope of voltage are calculated based on the current and voltage signals. A weighted integral is performed on the current change rate and the high-frequency envelope of the voltage, and the result of the weighted integral is compressed and mapped to obtain the arc precursor energy gate. Specifically, the voltage signal is acquired in real time by a voltage sensor to reflect the voltage fluctuation during the interruption process. The start and end times of the rising edge time window are used to define the time range of the current and voltage signal changes. The current change rate and the high-frequency envelope of the voltage are calculated within the rising edge time window, where the current change rate represents the amplitude of the current signal change per unit time, which is used to characterize the severity of the current fluctuation and helps to determine whether there is an arc precursor. The high-frequency envelope of the voltage is obtained by extracting the high-frequency components of the voltage signal through a filtering algorithm to characterize the severity of the voltage fluctuation. A high-pass filter can be used to retain the high-frequency fluctuation part in the voltage signal. The output current change rate is used to reflect the severity of the current change, and the output voltage high-frequency envelope is used to reflect the severity of the voltage fluctuation. The intensity of the arc precursor is assessed. A weighted integral is performed on the rate of change of current and the high-frequency envelope of voltage to comprehensively consider their roles in the occurrence of arc precursors. The weighted integral is calculated by pre-setting weights based on the predictive capabilities of different signals to generate a comprehensive energy value. The weighted integral process involves weighting and summing the values ​​of the rate of change of current and the high-frequency envelope of voltage to obtain a weighted integral value. The weights can be adjusted according to the operating characteristics of the equipment to prioritize more important signals. The weighted integral value is then compressed and mapped to standardize it to a fixed range for subsequent processing and to avoid computational instability due to excessively large or small signal values. This compression mapping uses nonlinear algorithms such as the Sigmoid function or logarithmic mapping to transform the weighted integral result to a preset standard range, such as between 0 and 1, ensuring that the arc precursor energy is suitable for subsequent judgment and processing. The output arc precursor energy gate is the compressed and mapped value used to characterize the arc precursor energy and predict the risk of arc occurrence.

[0025] The present invention is further configured such that S4 includes: Based on the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate, data fusion is performed to generate a multi-domain diagnostic map. The multi-domain diagnostic map is analyzed according to the preset judgment rules to determine the root cause of the event; A list of original waveform segments to be retained, corresponding to the root cause determination results of the event, is generated. This list includes current waveforms, port voltage drop waveforms, mechanism acceleration waveforms, and high-frequency voltage signal waveforms. Specifically, the contact entropy current index is obtained by calculating the perturbation intensity of the contact power to characterize the stability and uniformity of the circuit breaker contact area; the mechanism springback spectrum difference is calculated based on the acceleration signal to reflect anomalies in the mechanical state of the equipment, especially springback behavior; the arc precursor energy gate is obtained through current change rate and high-frequency voltage envelope analysis to predict arc precursors and assess arc risk. During data fusion, the data of the contact entropy current index, mechanism springback spectrum difference, and arc precursor energy gate are standardized to normalize to the same dimension. Maintaining the independence and importance of data across all dimensions is crucial. Principal component analysis (PCA) or cluster analysis can be used to merge standardized data into multi-dimensional feature vectors. Each data point represents the state of the circuit breaker at a specific moment. The fusion result forms a multi-domain diagnostic map, which visually reflects the health status of the equipment across various dimensions and guides the subsequent retention of original waveform data. In this embodiment, PCA is used as an example. Standardized multi-dimensional data is input into the PCA algorithm, and a new set of feature dimensions is generated through linear combination. These dimensions are the principal components. Each principal component reflects the key changing features in the original data. The feature vector corresponding to each data point in this feature space characterizes the health status of the equipment across various dimensions at a specific moment. This method effectively integrates the original multi-dimensional data into a multi-domain diagnostic map. Dimensional data is compressed into fewer principal components while retaining key information, providing a reliable basis for subsequent analysis and decision-making. During the root cause determination process, a rule engine is used to analyze multi-domain diagnostic maps to identify the type of equipment failure or abnormal state. The rule engine uses preset judgment rules based on historical data or expert experience to define typical values ​​of the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate under normal and abnormal states. For example, when the contact entropy flow index is greater than a preset threshold, it indicates that the contact area is uneven, suggesting a potential fault, while an arc precursor energy gate greater than a preset threshold indicates an increased probability of arcing. The rule engine integrates the anomalies of data from various dimensions and analyzes and determines the root cause of the event through decision trees or machine learning models. The output of the root cause is used for... This describes the abnormal state or fault type of the equipment, such as uneven current distribution caused by contact wear or high energy fluctuations before an arc occurs. Based on the root cause determination results, the original waveform segments to be retained are selected, with the selection based on the importance of current waveforms, port voltage drop waveforms, mechanism acceleration waveforms, and high-frequency voltage signal waveforms before, during, and after the event. For example, when the pre-arc energy threshold is high, current and voltage waveforms are prioritized for analysis to aid in arc occurrence; when the contact entropy current index is abnormal, port voltage drop and mechanism acceleration waveforms are prioritized for assessment of contact health. Time periods are selected based on event timestamps, typically including specific time periods before and after the event, such as 50 to 200 milliseconds of data to ensure complete event capture.The output list of raw waveform segments to be retained includes current waveforms, port voltage drop waveforms, mechanism acceleration waveforms, and high-frequency voltage signal waveforms along with their time periods to ensure the preservation of critical data for subsequent analysis and diagnosis.

[0026] The present invention is further configured such that S5 includes: Set a preset quantity threshold and classify events according to the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate; When the values ​​of the contact entropy flow index, the mechanism springback spectrum difference, and the arc precursor energy gate are all greater than the preset high threshold, it is marked as a Class A event, indicating a high-risk anomaly. When any one of the following values ​​is greater than the preset intermediate threshold: contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate, it is marked as a Class B event, indicating a medium-risk anomaly. When the values ​​of the contact entropy flow index, the mechanism springback spectrum difference, and the arc precursor energy gate are all less than or equal to the preset low threshold, it is marked as a Class C event, indicating a low-risk anomaly. The data for each event or period is fragmented according to the event level to generate corresponding group identifiers and fragment numbers; Calculate summary information for each data segment and perform integrity verification on the entire frame of data; Each data segment contains a timing timestamp, a monotonic timestamp, and a circular buffer index. Specifically, events are classified according to preset threshold values, and the raw waveform data is fragmented. The input data includes the contact entropy current index, mechanism springback spectrum difference, and arc precursor energy gate. The raw waveform segment list generated using the aforementioned steps includes event-related current waveforms, port voltage drop waveforms, mechanism acceleration waveforms, and high-frequency voltage signal waveforms. During the event classification process, three thresholds—high, medium, and low—are preset for the contact entropy current index, mechanism springback spectrum difference, and arc precursor energy gate based on the equipment's historical operating data and working characteristics. The high threshold indicates a high-risk event, signifying a serious equipment anomaly; the medium threshold indicates a medium-risk anomaly; and the low threshold indicates a low-risk event. The process begins by identifying low-risk states. Then, the data for each event is assessed. If the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate values ​​are all greater than the high threshold, the event is marked as a high-risk (Level A) event. If any indicator exceeds the intermediate threshold but does not exceed the high threshold, the event is marked as a medium-risk (Level B) event. If all indicators are less than or equal to the low threshold, the event is marked as a low-risk (Level C) event. After event classification, the data in the original waveform segment list is fragmented according to the event occurrence time and level. Key waveform data for each event is extracted and saved according to preset durations before and after the event, ensuring that waveform data for high-risk events is completely preserved, data for medium-risk events is partially preserved, and data for low-risk events is... Selective retention reduces storage burden; the final output includes a level label for each event and a list of corresponding fragmented original waveform segments, supporting subsequent equipment status analysis, fault diagnosis, and historical data management. During data fragmentation, the retention time window for waveform data for each event is first determined based on the event level. For high-risk (Level A) events, waveform data from a longer period before and after the event is selected to fully capture event characteristics; for medium-risk (Level B) events and low-risk (Level C) events, data from a shorter period is selected to save storage resources. Subsequently, the waveform data within the selected time window is divided into multiple data segments according to a preset time. Each data segment contains current waveforms, port voltage drop waveforms, mechanism acceleration waveforms, and high-frequency voltage signal waveforms over a continuous time period. The system ensures that the segmentation method can completely preserve key event data without losing important signal features; a unique group identifier is generated for each data segment to identify all related segments under the same event, and a sequence number is assigned to each segment to ensure that the data segments maintain the correct order and integrity during subsequent reassembly; after each segment is generated, a data digest is calculated, which is obtained by converting the data segment into a fixed-length simplified representation, which can be generated using a hash algorithm or compression algorithm, in order to verify the consistency of data during data transmission and storage; the entire frame of data undergoes integrity verification before storage and transmission, using checksums or hash values ​​to check all data segments to ensure that the data has not been tampered with or lost, and to support the correct reassembly and analysis of the data segments of each event in the future;The final output includes multiple data segments corresponding to each event, group identifiers and fragment numbers for each segment, as well as summary information and frame integrity verification results for each data segment. This provides a reliable data foundation for subsequent device status analysis, fault diagnosis, and historical data management. After data processing for each event, the data segments are time-stamped, embedded with buffer indexes, and their integrity is verified. Each data segment is first generated with a time stamp provided by an external clock system to identify the accurate time of data acquisition, ensuring consistency of data collected from different devices. Simultaneously, a monotonic timestamp is generated for each data segment, generated by the device's internal clock to ensure the temporal consistency of data segments within the device and prevent external clock drift from affecting the data order. Each data segment also records a circular buffer index to identify its storage location within the edge device's circular buffer, ensuring consistency in the timing and storage location of data within the buffer.

[0027] The present invention is further configured such that S6 includes: Collect and use monitoring data within a preset time window before and after the event, fit the time synchronization timestamp with the monotonic timestamp, and obtain the monotonic time base anchor; The combination of monotonic timestamps and monotonic time base anchors is used as a unified sorting benchmark, and cross-link data is rearranged in time according to this benchmark; The monitoring data includes current signals, port voltage drop signals, mechanism acceleration signals, and high-frequency voltage signals. Specifically, monitoring data is collected within a preset time window before and after the event to provide information on the circuit breaker's performance under different operating states. Each data point is accompanied by a timing timestamp and a monotonic timestamp to identify the actual time of the data collection and the relative time within the device. A fixed time window is set before and after the event, for example, 50 to 200 milliseconds before and after. The edge device continuously collects and stores the above four types of signals within this time window, providing basic data for subsequent timestamp fitting and timing reordering. Subsequently, the timing timestamp and monotonic timestamp are fitted using interpolation algorithms or least squares methods to minimize the time deviation between them, thereby obtaining a monotonic time base anchor. This monotonic time base anchor is used to adjust the synchronization between the device's internal time and the external timing system, eliminating the impact of clock drift or rollback on data timing. The fitting process generates a global time base by collecting the mapping relationship between the timing timestamp and the monotonic timestamp, enabling different devices or... Data from different sensors can be arranged in a unified time order, ensuring correct rearrangement of cross-link data. The final output includes monitoring data within the time window before and after the event, and a monotonic time base anchor obtained through fitting. During cross-link data time-series rearrangement, data collected from different links or devices may have inconsistent time sequences. Therefore, the time stamp and monotonic timestamp are first converted to a unified time base using the monotonic time base anchor, ensuring correct sorting of cross-link data under the same time reference. Subsequently, data from different links or sensors are compared and adjusted, each data point is mapped to a monotonic time base anchor, and the data is rearranged according to the unified time base. This entire process is implemented through software algorithms to ensure data synchronization between links and avoid timing errors caused by device time asynchrony. Finally, the rearranged data is aligned with a global time base, providing a consistent time-series basis for subsequent analysis and calculation, thus generating the cross-link data time-series rearrangement result, ensuring all data are arranged in a unified time order and maintain time-series consistency.

[0028] The present invention is further configured such that S7 includes: The cloud maintains a bitmap of received fragments for each group of data frames. The bitmap is used to record the reception status of each fragment. Received fragments are marked as received, and unreceived fragments are marked as unreceived. Based on the bitmap, the cloud generates a gap list, which lists all unreceived fragment sequence numbers, and returns the list to the edge. The edge device retransmits the unreceived fragments listed in the gap list; After all data segments have been received, the cloud performs a consistency check on the entire frame. Specifically, the cloud receives data frames generated and transmitted by the edge device. Each data frame contains multiple data segments, including current signals, port voltage drop signals, mechanism acceleration signals, and high-frequency voltage signals. Each segment corresponds to a specific time period and is assigned a segment number according to the order of event occurrence. At the same time, the cloud obtains the reception status information of each segment to record whether the segment has been successfully received. Subsequently, for each group of data frames, the cloud maintains a segment reception bitmap, where each bit in the bitmap represents the reception status of a segment. The data frame is divided into segments, with received segments marked as 1 and unreceived segments marked as 0. The bitmap size matches the number of segments in the data frame. The cloud updates the bitmap in real time based on the received data segments to ensure the accuracy of the reception status of each segment. Finally, a segment reception bitmap is generated, clearly showing the reception status of each segment, thus ensuring the integrity of the data frame transmission. In the process of generating the gap list based on the bitmap, the cloud first analyzes the segment reception bitmap, identifies all unreceived segments, and lists the sequence numbers of these segments to form a gap list. Each item in the gap list corresponds to an unreceived segment. The missing fragments are then returned to the edge device. The edge device identifies the missing fragments based on the list and retrieves their data from the circular buffer or storage for retransmission. The retransmission process follows the fragment sequence numbers listed in the missing fragment list. Since only unreceived fragments are retransmitted, successfully received fragments are avoided, thus improving data transmission efficiency. Finally, the retransmission operation guided by the missing fragment list ensures the integrity of each data frame and generates a retransmitted fragment record for subsequent verification. After all fragments are received and retransmitted in the cloud, a consistency check is performed on the entire frame. First, all fragment data is integrated, and a checksum is calculated. This checksum can be obtained through a checksum or hash algorithm and is used to confirm that all fragments were received in the correct order and have not been lost or tampered with. If the calculated checksum matches the expected value, the entire frame is complete and error-free. If the checksum is inconsistent, it indicates an error occurred during transmission or storage, requiring an error handling mechanism to be triggered and a retransmission requested, thus ensuring the integrity and reliability of the data frame. Finally, the result of the entire frame consistency check is recorded in the cloud for subsequent data processing and analysis.

[0029] The present invention is further configured such that the method includes: displaying the event root cause analysis results and fragmented retransmission status through a graphical interface; specifically, displaying the event root cause analysis results and fragmented retransmission status in the graphical interface first involves preparing input data, including the root cause analysis results for each event and the fragmented reception status and retransmission process status for each group of data frames; the event root cause analysis results display the event risk level through a dashboard or bar chart, while also displaying detailed information on the root cause analysis, such as the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate; the fragmented retransmission status displays the reception status of each data frame through a progress bar or pie chart, and displays the gap list and retransmission fragment sequence number through a table or list, with the interface updating in real time to ensure that the user sees the latest information. Status; the root cause of an event can be displayed through pie charts or bar charts showing the proportion of various events, and users can click to view detailed information. A pop-up window displays specific analysis indicators and anomalies. A graphical interface enables interactive functions, allowing users to filter and sort events, such as viewing specific risk levels or specific types of events, and manually trigger fragment retransmission, pause, or cancel operations, controlled via interface buttons. The edge device is responsible for data collection and preliminary processing, and transmits event information and fragment status to the cloud. The cloud maintains the fragment reception bitmap and generates a gap list to feed back to the edge device. The graphical interface receives updates from the cloud in real time through API or data push mechanisms, enabling dynamic display and interactive operations, ensuring that users can clearly understand the event processing progress and the status of data frame reception and retransmission.

[0030] Example 2:

[0031] Please see Figure 2 An exemplary online monitoring system for a smart circuit breaker includes: Acquisition module: Acquires current signal, port voltage drop signal, mechanism acceleration signal and high-frequency voltage signal, synchronously marks each frame of data with time synchronization and monotonic timestamp, obtains the closing steady state window and the opening rising edge window, and saves the original waveforms before and after the event trigger for a preset duration through a multi-channel ring buffer at the edge end; The first calculation module calculates the contact power perturbation intensity within the closing steady-state window to obtain the contact entropy flow index, and calculates the difference between the acceleration root spectrum and the reference root spectrum within the mechanism's natural frequency band to obtain the mechanism's springback spectrum difference. The second calculation module performs weighted integration and compression mapping on the current rate of change and the high-frequency envelope of voltage within the window of the rising edge of the interruption to generate the arc precursor energy gate. Diagram construction module: Based on the contact entropy flow index, mechanism springback spectrum difference and arc precursor energy gate, construct a multi-domain diagnostic map, determine the root cause of the event, and output the waveform segment to be retained; Data fragmentation module: classifies and fragments waveform segments, generates frame identifiers and fragment sequence numbers, calculates fragment summaries and full frame verification, and embeds timestamps and buffer indices in each fragment; The time-series reordering module: It fits the timing and monotonic timestamps to generate monotonic time base anchors, and uses the monotonic timestamps as a benchmark to reorder the time series of cross-link data; Fragment retransmission module: The cloud maintains the fragmented reception bitmap of each frame group and returns the gap list. The edge end retransmits the missing fragments according to the gap list, and the cloud performs consistency verification on the whole frame data.

[0032] It should be noted that the online monitoring system for an intelligent circuit breaker provided in the above embodiments and the online monitoring method for an intelligent circuit breaker provided in the above embodiments belong to the same concept. The specific ways in which each module and unit performs its operation have been described in detail in the method embodiments, and will not be repeated here. In practical applications, the online monitoring system for an intelligent circuit breaker provided in the above embodiments can be assigned to different functional modules as needed, that is, the internal structure of the system can be divided into different functional modules to complete all or part of the functions described above, and this is not a limitation here.

[0033] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An online monitoring method for intelligent circuit breakers, characterized in that, include: S1: Collect current signal, port voltage drop signal, mechanism acceleration signal and voltage high frequency signal, synchronously mark the time and monotonic timestamp for each frame of data, obtain the closing steady state window and the opening rising edge window, and store the original waveforms of preset duration before and after the event trigger through the multi-channel ring buffer at the edge end; S2: Calculate the contact power perturbation intensity within the closing steady-state window to obtain the contact entropy flow index, and calculate the difference between the acceleration root spectrum and the reference root spectrum within the mechanism's natural frequency band to obtain the mechanism's springback spectrum difference; S3: Weighted integration and compression mapping of current rate of change and voltage high-frequency envelope are performed within the rising edge window of the interruption to generate arc precursor energy gate; S4: Based on the contact entropy flow index, mechanism springback spectrum difference and arc precursor energy gate, construct a multi-domain diagnostic map, determine the root cause of the event, and output the waveform segment to be retained. S5: Classify and segment waveform segments, generate frame identifiers and segment numbers, calculate segment summaries and full frame verification, and embed timestamps and buffer indices in each segment; S6: Fit the timing and generate a monotonic time base anchor with the monotonic timestamp, and rearrange the time sequence of cross-link data using the monotonic timestamp as a reference; S7: The cloud maintains the segmented reception bitmap of each frame and returns the gap list. The edge end retransmits the missing segments according to the gap list, and the cloud performs consistency verification on the whole frame data.

2. The online monitoring method for an intelligent circuit breaker according to claim 1, characterized in that, S1 includes: Current signal, port voltage drop signal, mechanism acceleration signal and high-frequency voltage signal are collected in real time at preset sampling frequencies; Each data frame is simultaneously marked with a timing timestamp and a monotonic timestamp. The timing timestamp is provided by an external clock system, and the monotonic timestamp is generated by the internal clock of the device. After closing the circuit, the steady-state window for closing is obtained by monitoring the current signal, port voltage drop signal, mechanism acceleration signal and high-frequency voltage signal; During the switching process, the rising edge window of the switching is obtained by monitoring the current signal and the voltage signal; A multi-channel circular buffer is set at the edge. The buffer capacity is configured to store complete waveform data for a preset duration before and after the event is triggered. The data is written in chronological order and overwritten cyclically to continuously preserve the original waveforms before and after the event.

3. The online monitoring method for an intelligent circuit breaker according to claim 1, characterized in that, S2 includes: Within the steady-state window of closing, the contact power is calculated based on the current signal and the port voltage drop signal, and the contact current index is generated according to the rate of change of the contact power. Within the preset frequency band of the mechanism, the current power spectrum is calculated based on the mechanism's acceleration signal, and the difference is integrated with the reference root spectrum to obtain the mechanism's rebound spectrum difference.

4. The online monitoring method for an intelligent circuit breaker according to claim 1, characterized in that, S3 includes: Within the time window of the rising edge of the switching, the rate of change of current and the high-frequency envelope of voltage are calculated based on the current and voltage signals. The current change rate and the high-frequency envelope of the voltage are weighted and integrated, and the result of the weighted integration is compressed and mapped to obtain the arc precursor energy gate.

5. The online monitoring method for an intelligent circuit breaker according to claim 1, characterized in that, S4 includes: Based on the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate, data fusion is performed to generate a multi-domain diagnostic map. The multi-domain diagnostic map is analyzed according to the preset judgment rules to determine the root cause of the event; Generate a list of original waveform segments to be retained corresponding to the root cause determination results of the event. The list includes current waveforms, port voltage drop waveforms, mechanism acceleration waveforms, and high-frequency voltage signal waveforms.

6. The online monitoring method for an intelligent circuit breaker according to claim 1, characterized in that, S5 includes: Set a preset quantity threshold and classify events according to the contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate; When the values ​​of the contact entropy flow index, the mechanism springback spectrum difference, and the arc precursor energy gate are all greater than the preset high threshold, it is marked as a Class A event, indicating a high-risk anomaly. When any one of the following values ​​is greater than the preset intermediate threshold: contact entropy flow index, mechanism springback spectrum difference, and arc precursor energy gate, it is marked as a Class B event, indicating a medium-risk anomaly. When the values ​​of the contact entropy flow index, the mechanism springback spectrum difference, and the arc precursor energy gate are all less than or equal to the preset low threshold, it is marked as a Class C event, indicating a low-risk anomaly. The data for each event or period is fragmented according to the event level to generate corresponding group identifiers and fragment numbers; Calculate summary information for each data segment and perform integrity verification on the entire frame of data; Each data segment contains a time stamp, a monotonic timestamp, and a circular buffer index.

7. The online monitoring method for an intelligent circuit breaker according to claim 1, characterized in that, S6 includes: Collect and use monitoring data within a preset time window before and after the event, fit the time synchronization timestamp with the monotonic timestamp, and obtain the monotonic time base anchor; The combination of monotonic timestamps and monotonic time base anchors is used as a unified sorting benchmark, and cross-link data is rearranged in time according to this benchmark; The monitoring data includes current signals, port voltage drop signals, mechanism acceleration signals, and high-frequency voltage signals.

8. The online monitoring method for an intelligent circuit breaker according to claim 1, characterized in that, S7 includes: The cloud maintains a bitmap of received fragments for each group of data frames. The bitmap is used to record the reception status of each fragment. Received fragments are marked as received, and unreceived fragments are marked as unreceived. Based on the bitmap, the cloud generates a gap list, which lists all unreceived fragment sequence numbers, and returns the list to the edge. The edge device retransmits the unreceived fragments listed in the gap list; After all fragments have been received, the cloud performs a consistency check on the entire frame of data.

9. The online monitoring method for an intelligent circuit breaker according to claim 1, characterized in that, The method also includes displaying the root cause analysis results and fragment retransmission status through a graphical interface.

10. An online monitoring system for an intelligent circuit breaker, used to implement the online monitoring method for an intelligent circuit breaker as described in any one of claims 1-9, characterized in that, include: Acquisition module: Acquires current signal, port voltage drop signal, mechanism acceleration signal and high-frequency voltage signal, synchronously marks each frame of data with time synchronization and monotonic timestamp, obtains the closing steady state window and the opening rising edge window, and saves the original waveforms before and after the event trigger for a preset duration through a multi-channel ring buffer at the edge end; The first calculation module calculates the contact power perturbation intensity within the closing steady-state window to obtain the contact entropy flow index, and calculates the difference between the acceleration root spectrum and the reference root spectrum within the mechanism's natural frequency band to obtain the mechanism's springback spectrum difference. The second calculation module performs weighted integration and compression mapping on the current rate of change and the high-frequency envelope of voltage within the window of the rising edge of the interruption to generate the arc precursor energy gate. Diagram construction module: Based on the contact entropy flow index, mechanism springback spectrum difference and arc precursor energy gate, construct a multi-domain diagnostic map, determine the root cause of the event, and output the waveform segment to be retained; Data fragmentation module: classifies and fragments waveform segments, generates frame identifiers and fragment sequence numbers, calculates fragment summaries and full frame verification, and embeds timestamps and buffer indices in each fragment; The time-series reordering module: It fits the timing and monotonic timestamps to generate monotonic time base anchors, and uses the monotonic timestamps as a benchmark to reorder the time series of cross-link data; Fragment retransmission module: The cloud maintains the fragmented reception bitmap of each frame group and returns the gap list. The edge end retransmits the missing fragments according to the gap list, and the cloud performs consistency verification on the whole frame data.

Images