Remote monitoring system of leakage circuit breaker integrated with internet of things

Abstract

The application relates to the technical field of leakage protection monitoring, and discloses a leakage circuit breaker remote monitoring system integrated with an Internet of Things. A leakage feature modeling module of the system acquires real-time monitoring data, extracts current waveform feature parameters, identifies abnormal waveform modes and first occurrence time sequences, and generates an original leakage feature template; a network path generation module sorts monitoring node topologies based on the template, establishes device connection paths, and constructs an Internet of Things topology path; a path anomaly analysis module extracts node sequences based on the topology path, filters abnormal transmission paths to obtain an abnormal path set; a risk level judgment module collects terminal node risk labels, matches to obtain a risk level label group; and a topology structure output module counts associated nodes, divides abnormal paths, establishes a mapping relationship, and generates a remote monitoring topology table. The system improves the accuracy and efficiency of leakage monitoring and is suitable for complex power network environments.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of leakage protection monitoring, in particular to a leakage circuit breaker remote monitoring system integrated with Internet of Things. BACKGROUND

[0002] With the continuous development and intelligent upgrading of power systems, as a key device to ensure power safety, the monitoring demand of leakage circuit breakers is increasingly prominent. The traditional leakage circuit breaker monitoring method relies on local manual inspection or simple alarm of a single node, which is difficult to achieve comprehensive coverage and real-time monitoring of large-scale and distributed power networks. In actual application, due to the wide distribution and complex environment of power network nodes, leakage faults often have suddenness and concealment, and it is difficult to quickly capture fault characteristics by relying on local monitoring, which may lead to delayed fault troubleshooting and even electrical fire, electric shock and other safety accidents.

[0003] The existing monitoring system mostly uses isolated monitoring nodes, lacks effective analysis of the correlation between nodes, and is difficult to build a complete leakage fault propagation path. When a leakage fault occurs in a certain area, the system cannot accurately trace the fault source and the scope of influence, so the maintenance personnel need to check each node one by one, which not only increases the maintenance cost, but also prolongs the fault processing time. In addition, due to the use of different communication protocols and data formats by different monitoring nodes, the data compatibility is poor, and it is difficult to realize efficient integration and sharing of information, which further restricts the intelligent level of leakage fault monitoring.

[0004] Under the background of the gradual popularization of Internet of Things technology, although some monitoring systems begin to introduce Internet of Things technology to realize data transmission, there are still obvious deficiencies in the accurate modeling of leakage characteristics, the dynamic generation of topological paths, and the scientific determination of risk levels. Most systems can only realize simple state data acquisition and uploading, lack deep analysis of leakage waveform characteristics, and are difficult to distinguish between normal fluctuations and abnormal leakage patterns, resulting in frequent false positives and false negatives. Moreover, in the construction of topological paths, a fixed connection method is mostly used, which cannot dynamically adjust the path structure according to real-time leakage characteristics, affecting the pertinence and efficiency of data transmission. The existence of these problems makes it difficult for the existing leakage circuit breaker monitoring system to meet the high-precision and high-efficiency requirements of modern power networks for safety monitoring, and an intelligent monitoring scheme integrated with Internet of Things technology is needed to solve the above bottlenecks. SUMMARY

[0005] The present application relates to the technical field of leakage protection monitoring, in particular to a leakage circuit breaker remote monitoring system integrated with Internet of Things.

[0006] To achieve the above-mentioned purpose, the present application provides a leakage circuit breaker remote monitoring system integrated with Internet of Things, which comprises:

[0007] The leakage characteristic modeling module obtains real-time monitoring data of the circuit breaker, extracts current waveform characteristic parameters, identifies abnormal waveform patterns and first occurrence time, maps and generates a characteristic parameter sequence, and constructs an original leakage characteristic template;

[0008] The network path generation module, based on the original leakage characteristic template, topologically sorts the monitoring nodes, establishes a device connection path from the master node to the terminal node according to the sorting result, collects the communication addresses of all associated nodes in the path and the transmission protocols of adjacent nodes, and constructs an Internet of Things topology path;

[0009] The path anomaly analysis module, according to the Internet of Things topology path, extracts the circuit breaker node sequence, compares the public nodes and counts the terminal node alarm frequency, filters the abnormal transmission path, and obtains an abnormal path set;

[0010] The risk level determination module, based on the terminal nodes in the abnormal path set, collects the risk labels associated with the terminal nodes, sorts them by trigger frequency, and matches the path endpoint risk identification to obtain a circuit breaker risk level label group;

[0011] The topology structure output module, according to the circuit breaker risk level label group, counts the monitoring area nodes corresponding to each label, divides the abnormal path to the corresponding node, establishes a mapping relationship structure between the node and the abnormal path, and generates a remote monitoring topology table.

[0012] Preferably, the specific analysis process of the path anomaly analysis module includes:

[0013] Set the monitoring period, calculate the risk coefficient by comparing the number of high-risk paths in the period with the total number of transmission paths, and if the risk coefficient exceeds the preset risk threshold, generate a first-level alarm signal;

[0014] If the risk coefficient does not exceed the preset risk threshold, calculate the delay deviation value by taking the absolute value of the difference between the path response delay value and the median of the preset delay range, and mark the path stability coefficient as a stability detection value by comparing it with the preset stability threshold;

[0015] Calculate the delay evaluation value by averaging all path delay deviation values in the monitoring period, and calculate the stability evaluation value by averaging all path stability detection values in the monitoring period;

[0016] Calculate the comprehensive risk value by numerically calculating the risk coefficient, delay evaluation value and stability evaluation value, and if the comprehensive risk value exceeds the preset comprehensive threshold, generate a second-level alarm signal, and if the comprehensive risk value does not exceed the preset comprehensive threshold, generate a normal operation signal.

[0017] Preferably, the leakage characteristic modeling module includes:

[0018] The waveform parameter extraction submodule obtains real-time monitoring data of the circuit breaker, performs a multi-scale decomposition operation on the current waveform, extracts a frequency spectrum feature set of each waveform, records a timestamp position of each feature in the waveform, compares a relationship between a first occurrence time of an abnormal pattern in the feature sequence and a waveform period, classifies according to the circuit breaker device, and obtains an abnormal feature time sequence distribution result;

[0019] The feature segment reconstruction submodule extracts a data segment corresponding to the abnormal pattern in the original waveform according to the abnormal feature time sequence distribution result, intercepts data based on a time interval of the abnormal pattern in the waveform, constructs a feature segment set according to a position of the intercepted data of each abnormal pattern, and recombines the feature segment set in combination with a device identifier to which the data belongs, to obtain an abnormal feature segment set;

[0020] The feature template generation submodule counts a trigger frequency value of all abnormal patterns based on the abnormal feature segment set, performs time sequence alignment processing on the feature segment set based on an original time sequence arrangement of the abnormal patterns in the monitoring data, and integrates feature segments of multiple abnormal patterns in the same device in sequence according to a first trigger time.

[0021] Preferably, the network path generation module comprises:

[0022] The node hierarchical ordering submodule orders all nodes according to their hierarchical priority values based on the original leakage feature template in combination with a network hierarchical identifier of a monitoring node, topologically rearranges the nodes in order from a core layer to an access layer, and establishes a node sequence index table;

[0023] The path offset detection submodule obtains a set of adjacent device addresses in the node sequence according to the node sequence index value, records a communication direction and a transmission time delay parameter of each pair of adjacent nodes, and calculates a topological offset strength value of the Internet of Things path;

[0024] The node relationship extraction submodule collects node addresses and communication protocol types in all connection relationships according to the Internet of Things path data, and constructs a node relationship mapping table according to the device connection relationship.

[0025] Preferably, the risk level determination module comprises:

[0026] The risk label collection submodule collects a set of risk classification labels corresponding to each terminal node based on the terminal nodes in the abnormal path set, indexes and maps the abnormal path and the terminal risk label, and generates a path end risk label group;

[0027] The label frequency counting submodule performs trigger frequency counting operation on all risk labels based on the path endpoint risk label group, records the number of occurrences of each label in the abnormal path set, and sorts the labels according to the frequency to obtain a sorted risk label sequence.

[0028] The risk level matching submodule performs priority matching on the label set corresponding to the terminal node in the abnormal path according to the sorted risk label sequence, filters the path risk level matched with the highest label item in the sorting sequence in each path, and integrates the risk level identifiers of all abnormal paths.

[0029] Preferably, the topology output module comprises:

[0030] The monitoring node extraction submodule collects the regional monitoring nodes corresponding to each label according to the circuit breaker risk level label group, records the abnormal path numbers and the number of devices associated with each node, and establishes a mapping index of risk labels and monitoring nodes;

[0031] The path collection submodule divides the corresponding abnormal paths into each monitoring node according to the risk level label based on the mapping index value, establishes a bidirectional association structure of path numbers and node addresses, and extracts the path list managed by each node;

[0032] The topology generation submodule integrates the monitoring nodes and subordinate abnormal path numbers according to the node path management list, outputs the node address, associated risk level and total path quantity, and generates a remote monitoring topology table.

[0033] Preferably, the system further comprises a response delay calibration module:

[0034] The time interval from receiving an instruction to executing an action of the circuit breaker is collected and marked as a response duration reference value;

[0035] The maximum and minimum values of all response durations in the monitoring period are obtained, and a response fluctuation coefficient is calculated;

[0036] A transmission delay compensation value is obtained through device communication quality analysis, and a delay calibration parameter is generated in combination with the response duration reference value and the response fluctuation coefficient;

[0037] When the response fluctuation coefficient exceeds a preset fluctuation threshold, a path optimization instruction is sent to the network path generation module.

[0038] Preferably, the device communication quality analysis comprises:

[0039] The signal strength and data packet loss rate between the monitoring nodes are collected, and if the signal strength is lower than a preset strength threshold or the packet loss rate is higher than a preset packet loss threshold, the link is determined as a low-quality channel;

[0040] The duration ratio of the low-quality channel in the statistical monitoring period is marked as a channel degradation coefficient;

[0041] The channel degradation coefficient is compared with a preset channel threshold value, a transmission delay compensation value is generated, and the delay calibration parameter is updated.

[0042] Preferably, the system further comprises an impedance test triggering module:

[0043] When the risk level determination module outputs a high-risk label, a multi-frequency band impedance test is started;

[0044] A frequency sweep test signal is injected into the circuit breaker associated with the high-risk node;

[0045] The attenuation characteristics of the test signal in the node network are collected;

[0046] The deviation of the measured attenuation characteristics from the theoretical attenuation model is compared, and an impedance anomaly coefficient is generated.

[0047] Preferably, the theoretical attenuation model is constructed by:

[0048] The theoretical signal attenuation curve is calculated according to the node impedance parameters in the Internet of Things topology path, and the environmental interference factor is superimposed to correct the theoretical signal attenuation threshold range;

[0049] When the measured attenuation characteristics exceed the threshold range, the node is marked as an impedance anomaly point.

[0050] Compared with the prior art, the beneficial effects of the present application are:

[0051] The integrated Internet of Things leakage circuit breaker remote monitoring system forms a complete leakage monitoring system through the cooperative work of multiple modules, and exhibits many advantages in practical application. The leakage characteristic modeling module can deeply mine real-time monitoring data of the circuit breaker, extract current waveform characteristic parameters, and identify abnormal waveform patterns and the first occurrence time. The original leakage characteristic template generated thereby provides a precise basis for subsequent fault analysis, helps to more accurately distinguish between normal operation state and leakage fault state, and reduces the misjudgment caused by inaccurate feature judgment.

[0052] The network path generation module performs topology sorting on the monitoring nodes based on the original leakage characteristic template, and then establishes a device connection path from the master node to the terminal node, and collects related communication addresses and transmission protocols to construct an Internet of Things topology path. This dynamically generated topology path can be adjusted according to real-time leakage characteristics, making data transmission more targeted, avoiding the blindness of data transmission under traditional fixed paths, improving the efficiency of information transmission, and also enhancing the adaptability of the system to complex network environments, ensuring smooth interaction of data between nodes.

[0053] The path anomaly analysis module extracts the circuit breaker node sequence according to the Internet of Things topology path, screens the abnormal transmission path by comparing the public nodes and counting the alarm frequency of the terminal nodes, and the obtained abnormal path set can clearly reflect the possible propagation path of the electric leakage fault, which helps to quickly lock the key node where the fault occurs, provides a clear clue for fault tracing, and facilitates maintenance personnel to more targetedly troubleshoot and shorten the fault positioning time.

[0054] The risk level determination module collects the associated risk labels of the terminal nodes in the abnormal path set and matches the path endpoint risk identification after sorting according to the triggering frequency, and the circuit breaker risk level label group formed can intuitively present the risk degree of different nodes. This enables monitoring personnel to reasonably allocate resources according to the risk level, and preferentially process the fault of high-risk nodes, thereby optimizing the fault processing order and improving the overall fault handling efficiency.

[0055] The topology structure output module counts the nodes of the monitoring area corresponding to each label according to the risk level label group, divides the abnormal path to the corresponding node and establishes a mapping relationship structure, and generates a remote monitoring topology table to provide a clear visual interface for monitoring personnel, which can comprehensively master the distribution and association of electric leakage faults in the entire power network, help to grasp the monitoring situation as a whole, realize global management of electric leakage faults, and improve the safety monitoring level of the entire power network.

[0056] Through the organic combination of various modules, the system realizes intelligent processing of the whole process from electric leakage feature capture, path construction, anomaly analysis to risk determination and topology presentation, effectively integrates the Internet of Things technology and electric leakage monitoring requirements, and makes the electric leakage fault monitoring more accurate and efficient, which can better adapt to the complex and variable operating environment of modern power networks. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 The working principle diagram of the integrated Internet of Things electric leakage circuit breaker remote monitoring system described in the application;

[0058] Figure 2 The flowchart of the alarm signal generated by the path anomaly analysis module;

[0059] Figure 3 The path anomaly analysis result graph;

[0060] Figure 4 The network topology heat map;

[0061] Figure 5 The flowchart of feature extraction of the electric leakage feature modeling module;

[0062] Figure 6 The electric leakage feature modeling process graph;

[0063] Figure 7 is a feature parameter analysis graph;

[0064] Figure 8 is a flow chart for generating a response delay calibration module parameter. DETAILED DESCRIPTION

[0065] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work are within the protection scope of the present application.

[0066] Please refer to Figure 1 The present application provides an integrated Internet of Things leakage circuit breaker remote monitoring system, which comprises:

[0067] The leakage feature modeling module acquires circuit breaker monitoring data in real time, extracts a frequency spectrum feature set of a current waveform by using a multi-scale decomposition technology, records a timestamp position of a first occurrence of an abnormal pattern, and constructs an abnormal feature time sequence distribution result containing a device identifier. The network path generation module establishes a device connection path from a master node to a terminal node based on a topological sorting algorithm, acquires a communication address and a transmission protocol of each node to form an Internet of Things topological path. The path abnormality analysis module filters out an abnormal transmission path set by comparing a public node and a statistical alarm frequency. The risk level determination module sorts a risk label trigger frequency of a terminal node, matches a risk identifier of a path endpoint to generate a level label group. The topological structure output module finally divides the abnormal path into a corresponding monitoring node according to a risk level, establishes a node-path mapping relationship, and generates a remote monitoring topological table.

[0068] Embodiment 1: refer to Figure 2 The detailed work flow of the path abnormality analysis module. This module dynamically evaluates the Internet of Things topological path through a periodic monitoring mechanism, and realizes accurate identification of abnormal transmission paths by using a multi-level early warning strategy. The monitoring period is preset to 5 minutes in the system initialization stage, and this time window is determined to be able to balance the relationship between real-time performance and system load after multiple debugging. When the monitoring period is started, the data acquisition unit reads the state records of all active transmission paths from the distributed database, including path number, node sequence, communication quality indicators and other basic information.

[0069] The risk coefficient calculation process first screens out path records with high-risk characteristics. The system marks the path as a high-risk path according to the preset evaluation rule library if it meets any of the following conditions: the path contains more than three consecutive alarm nodes, the path terminal node triggers an emergency alarm within the last hour, and the average transmission delay of the path exceeds twice the threshold value. The statistical module calculates the ratio of the number of high-risk paths to the total number of paths in real time, and this ratio directly reflects the overall security status of the network as the risk coefficient. When the risk coefficient exceeds the preset threshold value of 0.35, the early warning generation unit immediately triggers a level one alarm signal, which is broadcast to all associated subsystems through the message queue.

[0070] For the case where the risk coefficient does not reach the threshold value, the system enters the secondary evaluation process. The delay evaluation subsystem first extracts the response delay data of each path from the path quality database, calculates the absolute deviation of the actual delay value of each path from the median value of the preset ideal delay range, and generates an evaluation index reflecting the overall delay state of the network.

[0071] The stability evaluation subsystem processes path stability data synchronously. The system maintains a sliding window to record the stability coefficient of each path for the last 20 communications, which is calculated by weighting parameters such as packet arrival rate, error rate, and retransmission rate. The evaluation process compares the current stability coefficient with the preset baseline value to generate a standardized stability detection value. The system performs mean value operation on the stability detection values of all paths to output an evaluation index representing the overall stability of the network.

[0072] The three-layer weighted algorithm is used to process the three core indicators in the comprehensive risk evaluation stage. The risk coefficient accounts for 50% of the weight, reflecting the overall situation of network security; the delay evaluation value accounts for 30% of the weight, reflecting the abnormal degree of transmission efficiency; and the stability evaluation value accounts for 20% of the weight, reflecting the fluctuation of communication quality. The system normalizes and sums the three indicators to generate a comprehensive risk value ranging from 0 to 1. This value is compared with the preset threshold value of 0.6 to trigger corresponding operations: when it exceeds the threshold value, a level two alarm signal is generated, indicating that there is a potential risk in the network; when it is below the threshold value, a normal operation signal is output, maintaining the current monitoring state.

[0073] The alarm signal processing unit adopts a differentiated response strategy. When a first-level alarm is triggered, the system automatically performs the following operations: suspend all non-critical data transmission tasks, immediately start the backup communication path, and send a highest priority notification to the operation and maintenance personnel. When a second-level alarm is triggered, the system takes relatively moderate measures: limit the bandwidth occupancy of application data transmission exceeding 70%, enhance the monitoring frequency of related paths, and generate a to-be-checked task list for the operation and maintenance queue. In the normal running state, the system continues to maintain the regular monitoring mode, while periodically updating the path quality evaluation database.

[0074] The dynamic parameter adjustment mechanism continuously optimizes the evaluation parameters. The system records the correspondence between the results of each risk evaluation and the actual fault occurrence, and automatically adjusts the following parameters through machine learning algorithms: calculation weight of risk coefficient, trigger threshold of each level of alarm, and reference value of ideal delay range. The parameter adjustment process follows the gradual principle, with an update amplitude not exceeding 10% of the original value, ensuring the stability of the system evaluation standard.

[0075] The data visualization component displays the analysis results in real time. The monitoring interface uses a heat map to display the risk level of each path, with color depth representing the risk level. Trend charts are also provided to show the historical change curves of risk coefficient, delay evaluation value, and stability evaluation value. Operation and maintenance personnel can view the detailed evaluation data of any path through the interactive interface, including the changes of various indicators in the last 10 cycles.

[0076] The abnormal path diagnosis function is automatically started after the alarm is triggered. The system performs in-depth detection on the marked abnormal paths, including: running state check of path node devices, physical connection test of transmission lines, and integrity verification of communication protocols. The detection results generate detailed diagnosis reports, marking possible problem sources such as device hardware failure, excessive line interference, and protocol configuration errors. The diagnosis report and risk evaluation data together constitute a complete path abnormality analysis result, providing a reference basis for subsequent maintenance decisions.

[0077] The system maintenance module ensures the reliability of the analysis process. Daily database maintenance tasks are performed in the early morning, including index reconstruction, data compression, and historical record archiving. The evaluation algorithm is checked for integrity every week to verify the accuracy of each indicator calculation. The evaluation criteria in the rule library are updated every month to adapt to changes in network topology and the impact of device updates. All maintenance operations record detailed log information to support subsequent audit tracking needs.

[0078] Reference Figure 3, which shows the working results of the path anomaly analysis module in detail. The left column chart shows the number distribution of paths with different risk levels, with high-risk paths (red) accounting for 15%, medium-risk paths (orange) accounting for 25%, low-risk paths (yellow) accounting for 30%, and normal paths (green) accounting for 30%. The middle line chart shows the risk coefficient trend of the last 10 monitoring periods, with a peak of 0.42 in the 5th period, triggering a level 1 alarm. The right pie chart shows the distribution of different alarm types, with communication delay alarms accounting for 45%, stability problem alarms accounting for 30%, and node failure alarms accounting for 25%. These data intuitively reflect the overall security status and abnormal distribution of the network, providing an important basis for operation and maintenance decisions.

[0079] Referring to Figure 4 , which shows the IoT topology heat map generated by the network path generation module. In the figure, nodes represent circuit breaker devices, and lines represent communication paths. The color of the nodes reflects the risk level of the device (red for high risk and green for normal), and the thickness of the lines represents the communication quality (thick lines represent high quality and thin lines represent low quality). The background color of the heat map represents the transmission delay of the path, with red areas indicating a delay of more than 120ms and green areas indicating a delay of less than 60ms. The figure clearly shows three main abnormal clusters: a high-risk device aggregation area in the upper left corner, a communication bottleneck area in the middle, and a low-quality path concentration area in the lower right corner. Topology analysis shows that high-risk devices are mainly distributed in the network edge area, and the communication quality in the core area is generally good.

[0080] Example 2: Referring to Figure 5 , the complete workflow of the leakage feature modeling module. This module converts raw current waveform data into structured feature templates through multi-stage processing, providing standardized input data for subsequent analysis. In the system initialization stage, the sampling rate of the current sensor array is configured to 10kHz to ensure that high-frequency components in the waveform can be captured. After the analog signals collected by the sensors are converted by a 24-bit ADC, digital waveform data streams enter the preprocessing link.

[0081] The waveform parameter extraction process uses a multi-resolution analysis method to process the current waveform. The system establishes a three-level decomposition architecture to meet the feature extraction requirements of different time scales. The first level processes millisecond-level waveform fluctuations, using sliding window Fourier transform to extract fundamental and harmonic components; the second level analyzes second-level trend changes, using discrete wavelet transform to calculate energy distribution features; the third level processes minute-level long-term evolution, identifying slow drift patterns through moving average algorithm. Each decomposition level outputs a corresponding set of spectral features, including amplitude, phase, harmonic distortion rate, and 12 other parameters. The timestamp recording unit accurately marks the position of each feature point in the original waveform, with an accuracy of microseconds.

[0082] An abnormal pattern recognition link establishes a dynamic threshold detection mechanism. The system maintains a ring buffer to store waveform feature data of the last 30 minutes as a reference benchmark for normal working conditions. The real-time collected feature parameters are compared with the benchmark data, and when a deviation of more than three standard deviations occurs, an abnormal flag is triggered. The time sequence analysis unit records the first occurrence time of each abnormal pattern and calculates the relative time difference from the start of the waveform cycle. The device classification processor stores the abnormal feature distribution results in the corresponding device archive according to the unique identification code of the circuit breaker.

[0083] The feature fragment reconstruction stage uses adaptive window technology to process abnormal data. The system dynamically adjusts the size of the intercepted window according to the duration characteristics of the abnormal pattern. For transient pulse-type abnormalities, a fixed 20ms capture window is set; for sustained oscillation-type abnormalities, the window length is automatically expanded to 1.5 times the duration of the abnormality. The data interception process uses a Hanning window function for edge smoothing to reduce spectral leakage effects. Each intercepted fragment is accompanied by complete metadata information, including device ID, abnormal type code, timestamp index, etc.

[0084] The feature template generation process realizes the spatio-temporal alignment of multi-source data. The frequency statistics unit maintains a device-based type counter that records the number of triggers for each type of abnormal pattern. The time sequence alignment processor uses the dynamic time warping algorithm to solve the non-uniform distribution problem of different abnormal patterns on the time axis. For multiple abnormal feature fragments detected by the same device, the system arranges them in the order of the first trigger time. The template integration engine packages the processed feature data into a standard format data block, including feature vectors, time sequence indexes, and associated metadata.

[0085] The data quality control mechanism runs throughout the entire processing flow. The input data verification link checks the integrity and continuity of the waveform sampling and discards abnormal records with missing data. Real-time verification is implemented during feature extraction, and data re-sampling is triggered when parameter values outside the physical range are detected. The abnormal fragment reconstruction stage performs cross-validation to ensure that the intercepted data window completely contains the target feature pattern. After template generation, format compliance checks are performed before writing to the feature database.

[0086] The distributed storage architecture supports large-scale data processing. The system uses sharding technology to divide the feature database into multiple logical units by device type. Each shard deploys an independent processing node to implement parallel computing of feature extraction and template generation. The data synchronization service maintains consistency between shards to ensure that global queries can obtain complete feature information. The storage engine implements data compression and index optimization to balance access speed and storage space consumption.

[0087] The dynamic updating mechanism keeps the feature templates up-to-date. The system sets a dual updating strategy: scheduled updating performs full feature reorganization every 30 minutes; event-triggered updating initiates local reconstruction immediately when detecting new abnormal patterns. The template version control system records the content changes of each update, supporting historical version tracing and rollback. The aging data automatic archiving process transfers feature templates that exceed the valid period to secondary storage.

[0088] The visual monitoring interface provides observability of the processing process. The operation and maintenance console displays the spectrum graph of each level of waveform decomposition in real time, with different colors marking the positions of abnormal feature points. The feature distribution heat map shows the aggregation of each type of abnormal pattern on the time axis. The template structure viewer presents the organizational relationship of feature data in the form of a tree graph, supporting drilling to view the data content of any detail level.

[0089] The abnormal diagnosis auxiliary function enhances the practicality of the system. The feature comparison tool allows similarity analysis of real-time collected waveform data with historical templates. The pattern tracking function draws the evolution trajectory of a specific abnormal feature in the time dimension. The correlation analysis engine discovers the potential relationship between different feature parameters, forming an interactive relationship network graph.

[0090] The system maintenance module ensures the stability of the processing flow. Daily storage system health checks are performed to verify data integrity and access performance. Weekly accuracy calibration of feature extraction algorithms is performed to adjust the decomposition parameters to adapt to changes in device aging characteristics. Monthly evaluation of the effectiveness of the template structure is performed to optimize data organization and improve query efficiency. All maintenance operations record detailed operation logs to form a complete system operation trajectory document.

[0091] The fault-tolerant processing mechanism deals with various abnormal situations. When the network is interrupted, local caching is enabled to continue data collection, and after the connection is restored, the backlog tasks are processed synchronously. When hardware fails, it automatically switches to a backup processing node to ensure uninterrupted feature extraction process. When data is abnormal, the self-repair process is triggered to recover damaged information through redundancy checking. The system resource monitoring module dynamically adjusts the processing load to prevent performance degradation caused by overload.

[0092] As Figure 6, which shows the three-level decomposition process of the leakage characteristic modeling module. The upper waveform chart shows the original current waveform and its abnormal feature point positions, with three main abnormal areas marked. The middle three subplots correspond to the feature extraction results at the millisecond, second, and minute levels, respectively: the millisecond-level analysis shows high-frequency harmonic distortion characteristics; the second-level analysis shows energy distribution changes; and the minute-level analysis presents a slow drift trend. The lower heat map shows the distribution density of various abnormal patterns on the time axis, where pulse-type abnormalities are mainly concentrated in the first half, and oscillation-type abnormalities are evenly distributed. The statistical data at the bottom of the chart shows that the system detected 42 abnormal feature points, with an average of 50 devices.

[0093] As shown in Figure 7 , the key parameters extracted by the leakage characteristic modeling module are analyzed in detail. The upper scatter plot matrix shows the correlation of the four core feature parameters: the harmonic distortion rate is negatively correlated with the fundamental amplitude, and the energy entropy has no obvious correlation with the pulse frequency. The middle box plot compares the distribution of feature parameters of devices with different risk levels, and the harmonic distortion rate of high-risk devices is significantly higher than that of other devices. The lower parallel coordinate plot shows the combination relationship of multi-dimensional feature parameters, where high-risk devices are mainly distributed in the area with high harmonic distortion rate, high energy entropy, and high pulse frequency. The chart analysis shows that the harmonic distortion rate is the most significant feature for distinguishing high-risk devices and can be used as a key indicator for early warning.

[0094] Embodiment 3: The cooperative workflow of the network path generation module and the risk level determination module. This implementation scheme builds the Internet of Things topology through multi-dimensional data processing and establishes an accurate risk assessment system. The system initialization stage loads network configuration parameters, including node level definition table, communication protocol whitelist, and transmission delay benchmark value. After the core processing engine starts, it first performs the node level sorting process, reads the network level identification field of each monitoring device, which contains the position information of the device in the network architecture, such as core layer device marked as CL-1 and access layer device marked as AL-3.

[0095] The level sorting algorithm uses an improved topology traversal method to establish a node priority queue. The processing process starts from the core layer device and expands outward layer by layer according to the network traffic forwarding direction, assigning a unique sequence index value to each node. The index calculation considers device type, processing capacity, and connection number, etc., forming a weighted node sequence index table. The path generation unit scans the physical connection relationship between adjacent devices based on the index table, records the communication direction attributes of each pair of nodes, including uplink, downlink, or peer-to-peer connection. The transmission delay monitor collects path delay data in real time, and after sliding average filtering, calculates the path stability index:

[0096]

[0097] wherein, path offset intensity value, sampling times, is the time delay value of the measurement, time delay reference value, is the maximum time delay threshold value allowed. The formula calculates the result for evaluating the transmission quality stability of the path, and the larger the value indicates the more obvious the path fluctuation.

[0098] The node relationship mapping process uses a graph database to store network topology information. The system creates an edge record for each connection relationship, including the source node address, target node address, communication protocol type, and transmission characteristic parameters. The dynamic update mechanism periodically scans network state changes, and when detecting new devices or connection disconnections, triggers topology structure reconstruction. The path optimizer analyzes the connectivity of the current topology, identifies potential redundant paths and single-point failure risks, and proposes structure adjustment suggestions.

[0099] After the risk level determination module is started, the terminal node list is first extracted from the abnormal path set. The risk label collector queries the device attribute database to obtain the risk classification label associated with each terminal node. The label definition follows the hierarchical standard, dividing the risk into three categories: electrical safety, data security, and operation safety, each category has several subcategories. The frequency statistics unit establishes a counting mechanism based on a time window to count the number of occurrences of each label in the last 24 hours. The sorting processor performs multi-condition sorting on the label set, mainly based on the trigger frequency, and secondarily considers the risk category priority.

[0100] The risk level matching process realizes the accurate association of paths and labels. The system establishes a risk assessment file for each abnormal path, recording all the nodes passed by the path and their risk attributes. The matching algorithm starts from the terminal node and traces the risk propagation path in reverse along the path to identify the risk source with the largest impact range. The level marker assigns a comprehensive risk level identifier to the path based on the matching result, using a five-level representation method from R1 (lowest) to R5 (highest). The risk association analyzer discovers the risk transmission relationship between different paths and constructs a global risk propagation network graph.

[0101] The data persistence layer adopts a hybrid storage strategy. Topology structure data is stored in a graph database, supporting complex relationship queries; node attribute information is saved in a document database, facilitating field extension; real-time monitoring data uses a time series database to optimize time series access performance. The storage engine realizes automatic sharding and load balancing to ensure system response speed during large-scale data processing.

[0102] The visualization subsystem provides multi-dimensional monitoring views. The topology map displays network nodes and connection relationships in a force-directed graph, with different colors and shapes distinguishing device types and risk levels. The path tracking view shows the detailed composition and risk distribution of selected paths. The latency heat map visually presents the communication quality status of each region in the network. The risk trend chart plots the time distribution of various risk events.

[0103] The system maintenance mechanism includes daily inspection and dynamic tuning. Database integrity checks and storage space monitoring are performed daily, and topology optimization and risk rule library updates are performed weekly. The performance tuning module automatically adjusts the number of processing threads and data cache size based on the running load. The fault recovery system automatically starts the backup processing flow when an exception is detected and notifies the operation and maintenance personnel.

[0104] The security control module implements strict access management. User permissions are subdivided to the data field level, controlling the access range of different roles to topology information and risk data. Operation audit logs all configuration changes and data processing actions, supporting post-tracing. Data transmission uses end-to-end encryption to prevent information leakage caused by network monitoring.

[0105] The fault-tolerant processing mechanism ensures system reliability. When the network is interrupted, local caching is enabled to continue data collection, and data synchronization is performed after the connection is restored. When a node fails, the communication path is automatically rerouted to maintain monitoring continuity. When data processing is abnormal, the retry mechanism is triggered, and after exceeding the number of retries, it is transferred to the manual processing queue. The resource monitor dynamically adjusts system load to prevent performance degradation caused by overload.

[0106] The version control system manages the evolution of algorithms and rules. Each update creates a new version branch, retaining the rollback ability of historical versions. The change impact assessment tool analyzes the possible chain reaction caused by modification, assisting in decision-making on the update timing. The test verification environment simulates real network conditions to verify the stability and accuracy of the new version before deployment.

[0107] This implementation realizes the transformation from raw network topology data to structured risk assessment results through the above processing flow, forming a complete Internet of Things security monitoring solution. The system design takes into account the complexity of the actual deployment environment, balances the relationship between processing precision and execution efficiency, and adapts to the application needs of networks of different sizes.

[0108] Example 4: see Figure 8, the collaborative working mechanism of the topology output module and the response delay calibration module is described, and the system implementation process is illustrated through specific application scenarios. A regional power distribution network is deployed with the monitoring system, including 12 substation nodes and 56 terminal circuit breaker devices, forming a three-layer Internet of Things topology. The system first loads the circuit breaker risk level label group when running, which contains device number, risk level, and geographic location, etc.

[0109] The monitoring node extraction process reads the regional configuration information from the central database. The system identifies 3 monitoring partitions in the region, each with an independent monitoring server. The data correlation engine performs the following operations: scan the device location code in the risk level label group, match the corresponding monitoring partition; query the number of devices currently managed by each monitoring node; calculate the load balancing index and determine the home node of the new device. Table 1 shows the monitoring node allocation of some devices in the region.

[0110] Table 1: Device monitoring node allocation table.

[0111]

[0112] The path aggregation process is based on the above allocation results. The system creates an independent management list for each monitoring node, recording all the abnormal path information it is responsible for. The path number uses a composite key form, containing the source node ID, target node ID, and timestamp. A bidirectional index structure maintains the mapping relationship between path number and device address, supporting two query modes: from path to device and from device to path. When a high-risk device CB-5120 is added, the system automatically divides it into the least loaded south node and updates the relevant index.

[0113] The topology generation process uses an incremental update strategy. The monitoring interface displays the current topology table in real time, including node address, risk level distribution, and path load, etc. The data synchronization service ensures that all monitoring nodes obtain the latest topology view. When 3 abnormal paths are added to the east node, the system triggers the following operations: recalculate the total number of paths for this node; evaluate the risk level distribution changes; push the update message to the relevant operation and maintenance terminal.

[0114] The response delay calibration module is activated in the circuit breaker action test scenario. The system sends a closing command to the selected device CB-2057, and a high-precision timer records the time delay data from the command issuance to the complete opening of the contact. The test is repeated 10 times to obtain the baseline response time data set. The data analysis unit calculates the time delay fluctuation range of the device and establishes a normal performance baseline curve. When the response delay of a certain operation exceeds the baseline value by 15%, the system marks the operation as abnormal.

[0115] The communication quality analysis is carried out for a specific transmission link. The communication link between the node 192.168.1.10 and the device CB-1024 experiences signal strength fluctuations. The system collects the following parameters: signal strength sample values per minute, packet loss number, and retransmission request number. The quality evaluator identifies that the signal strength is below -78 dBm during the afternoon period, and calculates the channel degradation coefficient for this period as 0.38. The compensation value generator outputs the transmission delay compensation parameters accordingly, and adjusts the time delay judgment threshold of this link.

[0116] The path optimization instruction is triggered under certain conditions. When the average time delay of the paths managed by the west zone node exceeds 300 ms, the system detects that the response fluctuation coefficient is consistently high. The optimization engine analyzes the current topology structure and finds that there is a redundant path that passes through the core layer. The system automatically generates an optimization scheme: change the communication path of CB-2057 to directly connect to the regional aggregation node, skipping the core layer transit. The monitoring data after the implementation of the scheme shows that the response time delay of this device is reduced to 65% of the original level.

[0117] The dynamic load balancing mechanism handles the pressure of the monitoring node. When the number of paths managed by the south zone node reaches the preset upper limit of 20, the system starts the rebalancing program. The evaluator analyzes the current load rate, processing capacity, and network bandwidth of each node, and selects to temporarily migrate the 4 paths of CB-3098 to the east zone node. The migration process maintains service continuity and ensures that monitoring data is not lost. The load balancing status board displays the resource usage rate of each node in real time, assisting the operation personnel in decision-making.

[0118] The data persistence adopts a hierarchical storage strategy. Real-time monitoring data is saved in the in-memory database, supporting high-frequency read and write operations. Historical topology records are stored in the time-series database, managed by time partitioning. Device configuration information is stored using a document database, facilitating the extension of attribute fields. The backup service periodically compresses and archives key data to the distributed file system.

[0119] The exception handling process includes a multi-level response mechanism. When it is detected that the monitoring node 192.168.2.15 is unresponsive, the system performs the following operations in sequence: marks the node as unavailable, temporarily transfers the devices managed by it to other nodes, attempts to automatically restart the node service, and notifies the operation team for intervention. After the fault is recovered, data consistency checks are performed to ensure that the operation records during the transfer period are synchronized completely.

[0120] The visualization interface provides an interactive monitoring experience. The topology map uses different colors to mark the ranges of each partition, and dynamically displays the path flow status. The device detail panel displays complete risk level history records and recent operation logs. The path tracking view supports clicking on any path to view detailed quality indicators. The time delay trend chart compares the performance data before and after compensation.

[0121] Version control manages the history of configuration changes. Each topology adjustment generates a change record, including modification content, operator, and impact assessment. The rollback function allows recovery to any historical version, ensuring the safety of system configuration. The change approval process requires that important modifications must be confirmed twice.

[0122] This implementation shows the working mode of the system in the actual running environment through the specific processing flow described above. From device monitoring to path optimization adjustment, each link is closely connected to form a closed-loop management. The system design fully considers the complexity of the field environment, ensuring processing accuracy while considering execution efficiency, and can adapt to the monitoring needs of Internet of Things of different scales. Through rich monitoring views and detailed data records, operation and maintenance personnel can fully master the network state and make timely decisions.

[0123] Example 5: Complete workflow of impedance test triggering module, the system automatically starts the multi-frequency impedance analysis program when detecting high-risk nodes. When the risk level determination module outputs a risk label of R4 and above, the central controller sends a test preparation instruction to the area where the target circuit breaker is located. The test signal generator uses a programmable logic device to generate a sweep frequency signal from 10 Hz to 1 MHz, and the frequency interval is dynamically adjusted according to the device type. For oil-immersed circuit breakers, logarithmic intervals are used, and for vacuum circuit breakers, linear intervals are used.

[0124] The signal injection link uses photoelectric coupling technology to achieve electrical isolation. The test signal is boosted to an appropriate level through an amplification circuit and is connected to the circuit breaker control loop through a special test interface. The injection process strictly follows the safety specifications of the device to ensure that the signal amplitude does not exceed 30% of the rated operating voltage. The synchronous acquisition unit is deployed at key monitoring points in the target node network and uses high-precision ADC circuits to record signal waveforms. The sampling duration of each test frequency point is inversely proportional to the frequency, and more cycle data is collected in the low frequency band to improve the signal-to-noise ratio.

[0125] The signal attenuation analysis process establishes a multi-dimensional evaluation system. The time domain analysis module calculates the amplitude ratio and phase difference between the received signal and the transmitted signal, and extracts the attenuation characteristics of the fundamental and harmonic components. The frequency domain processing unit performs Fourier transform on the collected signal to generate amplitude-frequency response curves and phase-frequency response curves. The feature extractor identifies key parameter points from the curves, including resonance frequency, three-decade bandwidth, and group delay fluctuation. These parameters form the measured attenuation feature dataset, which is used for subsequent comparison and analysis.

[0126] Theoretical attenuation model construction considers the physical characteristics of the equipment. The system database stores the equivalent circuit parameters of various types of circuit breakers, including coil inductance, contact resistance, ground capacitance, and other basic data. The model generator retrieves the corresponding parameters according to the target device model and calculates the signal transmission function under ideal conditions. The environmental compensation module introduces real-time data collected by temperature and humidity sensors to modify the theoretical model. The final generated theoretical attenuation curve contains upper and lower boundary thresholds, forming an allowable fluctuation range.

[0127] Impedance anomaly detection uses a frequency-by-frequency comparison strategy. The analysis engine compares the measured data with the upper and lower boundaries of the theoretical model, marking the frequency points that exceed the threshold. The anomaly evaluator calculates the deviation integral value and generates an impedance anomaly coefficient ranging from 0 to 1. When the coefficient exceeds 0.15, the system determines that there is an impedance anomaly at the target node. The diagnostic reasoning module combines the abnormal frequency distribution characteristics to preliminarily determine the possible fault types, such as contact oxidation, insulation degradation, or mechanical deformation.

[0128] Multiple safety measures are implemented during the test process. The overcurrent protection circuit monitors the injected current in real-time and immediately cuts off the test signal in case of abnormality. The electromagnetic shielding device reduces the impact of external interference on the measurement results. The test interval controller prevents continuous high-frequency testing from causing overheating of the equipment. The permission verification mechanism ensures that only authorized personnel can initiate manual test instructions.

[0129] The results processing system implements abnormal data classification management. Confirmed impedance anomaly records are stored in the device health file and the historical anomaly pattern database is updated. The warning generator sends different levels of warning notifications based on the severity of the anomaly, and immediately triggers the on-site maintenance process for sudden and severe anomalies. The trend analyzer tracks the changing trend of the test results of the same device multiple times to identify potential faults that develop slowly.

[0130] The visualization interface provides rich test analysis views. The signal spectrum graph compares the measured curve with the theoretical boundary, with abnormal frequency bands highlighted in red. The impedance parameter table displays detailed measurement data for each frequency point. The historical comparison graph superimposes the changing process of the test results of the same device over time. The diagnostic suggestion panel lists possible fault causes and inspection points.

[0131] The system maintenance mechanism ensures the reliability of the test. The regular calibration program verifies the accuracy of the signal generator and the acquisition unit and adjusts the deviation parameters. The test case library saves the characteristic data of typical fault patterns to assist in diagnostic analysis. The software update module keeps the signal processing algorithm up-to-date with the latest research progress. The hardware diagnostic function automatically detects the connection state and component performance of the test circuit.

[0132] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting; it is not intended to exclude myriad other embodiments of the present application that other inventors can develop based on the same general inventive concepts embodied by the described embodiments. That is, although the present application is described in terms of particular embodiments and illustrative figures, it should be apparent that the scope of the present application is not limited to these specific embodiments.

[0133] While the embodiments of the application have been shown and described herein, it will be understood by those of ordinary skill in the art that various changes, modifications, alternatives, and variations can be made to the embodiments without departing from the spirit and scope of the application, which is defined by the claims and their equivalents.

Claims

1. A remote monitoring system for residual current circuit breakers integrating the Internet of Things, characterized in that, include: The leakage current feature modeling module acquires real-time monitoring data of the circuit breaker, extracts current waveform feature parameters, identifies abnormal waveform patterns and their first occurrence sequence, maps and generates feature parameter sequences, and constructs the original leakage current feature template. The network path generation module sorts the monitoring nodes in the topology based on the original leakage current feature template, establishes a device connection path from the master node to the terminal node according to the sorting result, collects the communication addresses of all associated nodes and the transmission protocols of adjacent nodes in the path, and constructs the Internet of Things topology path. The path anomaly analysis module extracts the circuit breaker node sequence based on the IoT topology path, compares common nodes and counts the alarm frequency of terminal nodes, filters abnormal transmission paths, and obtains an abnormal path set. The risk level determination module collects risk tags associated with terminal nodes based on the terminal nodes in the abnormal path set, sorts them by trigger frequency, matches the risk identifier of the path endpoint, and obtains the circuit breaker risk level tag group. The topology output module, based on the circuit breaker risk level label group, counts the monitoring area nodes corresponding to each label, assigns abnormal paths to the corresponding nodes, establishes a mapping relationship structure between nodes and abnormal paths, and generates a remote monitoring topology table. The risk level determination module includes: The risk label collection submodule collects the risk classification label set corresponding to each terminal node based on the terminal nodes in the abnormal path set, performs index mapping between abnormal paths and terminal risk labels, and generates a path endpoint risk label group. The tag frequency statistics submodule performs a trigger frequency statistics operation on all risk tags based on the path endpoint risk tag group, records the number of times each tag appears in the abnormal path set, and sorts them according to frequency to obtain a sorted risk tag sequence. The risk level matching submodule performs priority matching on the tag set corresponding to the terminal node in the abnormal path according to the sorted risk tag sequence, filters the path risk level in each path that matches the highest tag item in the sorted sequence, and integrates the risk level identifiers of all abnormal paths.

2. The remote monitoring system for residual current circuit breakers integrating the Internet of Things as described in claim 1, characterized in that, The specific analysis process of the path anomaly analysis module includes: Set a monitoring period, calculate the risk coefficient by the ratio of the number of high-risk paths to the total number of transmission paths within the period, and generate a first-level alarm signal if the risk coefficient exceeds the preset risk threshold. If the risk coefficient does not exceed the preset risk threshold, the absolute value of the difference between the path response delay value and the median of the preset delay range is used to obtain the delay deviation value, and the ratio of the path stability coefficient to the preset stability threshold is marked as the stability detection value. The delay assessment value is obtained by averaging the delay deviation values ​​of all paths within the monitoring period, and the stability assessment value is obtained by averaging the stable detection values ​​of all paths within the monitoring period. A comprehensive risk value is obtained by numerically calculating the risk coefficient, delay assessment value, and stability assessment value. If the comprehensive risk value exceeds the preset comprehensive threshold, a level 2 alarm signal is generated; if the comprehensive risk value does not exceed the preset comprehensive threshold, a normal operation signal is generated.

3. The remote monitoring system for residual current circuit breakers integrating the Internet of Things as described in claim 1, characterized in that, The leakage current feature modeling module includes: The waveform parameter extraction submodule acquires real-time monitoring data of the circuit breaker, performs multi-scale decomposition on the current waveform, extracts the spectral feature set of each waveform, records the timestamp position of each feature in the waveform, compares the relationship between the first occurrence time of the abnormal mode in the feature sequence and the waveform period, classifies it according to the circuit breaker equipment, and obtains the time-series distribution results of abnormal features. The feature segment reconstruction submodule extracts the data segments corresponding to the abnormal patterns in the original waveform based on the abnormal feature time-series distribution results, truncates the data based on the time interval of the abnormal patterns in the waveform, constructs a feature segment set based on the position of the truncated data for each abnormal pattern, and reassembles it in combination with the device identifier to which the data belongs, to obtain an abnormal feature segment set. The feature template generation submodule calculates the trigger frequency of all abnormal modes based on the set of abnormal feature fragments, performs time sequence alignment processing on the feature fragment set based on the original time sequence of the abnormal modes in the monitoring data, and integrates the feature fragments of multiple abnormal modes in the same device according to the first trigger time.

4. The remote monitoring system for residual current circuit breakers integrating the Internet of Things according to claim 1, characterized in that, The network path generation module includes: The node hierarchy sorting submodule, based on the original leakage current feature template and combined with the network hierarchy identifier of the monitoring node, sorts all nodes according to their hierarchy priority value, and performs topology rearrangement of the nodes in the order from the core layer to the access layer, and establishes a node sequence index table. The path offset detection submodule obtains the set of adjacent device addresses in the node sequence based on the node sequence index value, records the communication direction and transmission delay parameters of each pair of adjacent nodes, and calculates the topology offset strength value of the IoT path. The node relationship extraction submodule collects the node addresses and communication protocol types in all connection relationships based on the IoT path data, and constructs a node relationship mapping table based on the device connection relationships.

5. The remote monitoring system for residual current circuit breakers integrating the Internet of Things according to claim 1, characterized in that, The topology output module includes: The monitoring node extraction submodule collects the regional monitoring nodes corresponding to each label based on the circuit breaker risk level label group, records the abnormal path numbers associated with each node and the number of devices associated with each node, and establishes a mapping index between risk labels and monitoring nodes. Based on the mapping index value, the path aggregation submodule divides the corresponding abnormal paths into each monitoring node according to the risk level label, establishes a bidirectional association structure between path number and node address, and extracts the path list managed by each node. The topology generation submodule integrates the monitoring nodes and their subordinate abnormal path numbers according to the path list, outputs the node address, associated risk level and total number of paths, and generates a remote monitoring topology table.

6. The remote monitoring system for residual current circuit breakers integrating the Internet of Things according to claim 1, characterized in that, The system also includes a response delay calibration module: The time interval between receiving a command and executing an action by the circuit breaker is collected and marked as the response time baseline value; Obtain the maximum and minimum values ​​of all response durations within the monitoring period, and calculate the response fluctuation coefficient; The transmission delay compensation value is obtained by analyzing the device communication quality, and the delay calibration parameters are generated by combining the response duration benchmark value and the response fluctuation coefficient. When the response fluctuation coefficient exceeds the preset fluctuation threshold, a path optimization command is sent to the network path generation module.

7. The remote monitoring system for residual current circuit breakers integrating the Internet of Things according to claim 6, characterized in that, The device communication quality analysis includes: The signal strength and data packet loss rate between monitoring nodes are collected. If the signal strength is lower than the preset strength threshold or the packet loss rate is higher than the preset packet loss threshold, the communication link between the monitoring nodes is determined to be a low-quality channel. The percentage of duration of low-quality channels within the statistical monitoring period is denoted as the channel degradation coefficient. The channel degradation coefficient is compared with the preset channel threshold to generate a transmission delay compensation value and update the delay calibration parameters.

8. The remote monitoring system for residual current circuit breakers integrating the Internet of Things according to claim 1, characterized in that, The system also includes an impedance test trigger module: When the risk level determination module outputs a high-risk label, the multi-band impedance test is initiated. Inject frequency sweep test signals into circuit breakers associated with high-risk nodes; Acquire test signal attenuation characteristics in node network; By comparing the deviation between the measured attenuation characteristics and the theoretical attenuation model, an impedance anomaly coefficient is generated.

9. The remote monitoring system for residual current circuit breakers integrating the Internet of Things according to claim 8, characterized in that, The theoretical attenuation model is constructed as follows: The theoretical signal attenuation curve is calculated based on the node impedance parameters in the IoT topology path, and the theoretical signal attenuation threshold range is corrected by superimposing environmental interference factors. When the measured attenuation characteristics exceed the threshold range, the node is marked as an impedance anomaly point.

Images