A leakage current on-line monitoring system and method based on mesh networking

Abstract

The application discloses a leakage current online monitoring system and method based on a Mesh networking, and belongs to the technical field of power Internet of Things. The method comprises the following steps: a monitoring device collects leakage current data and operates as a Mesh network node; the communication quality parameters such as RSSI, LQI and PER are comprehensively evaluated, the weight is dynamically adjusted in combination with the environmental electromagnetic interference intensity, the comprehensive signal quality score is calculated, the network coverage range is adaptively expanded, and a reliable multi-hop transmission path is established. The node with the highest comprehensive signal quality score is dynamically elected as a data aggregation node, the network topology state data is collected, the topology characteristic parameters such as node distribution density, average physical distance and environmental electromagnetic interference intensity are extracted, the application scenarios such as intelligent buildings, industrial plants or low-voltage distribution areas are identified, the hierarchical early warning threshold and the regional alarm condition are adaptively set, and the leakage current early warning based on the scenarios is realized.

Description

Technical Field

[0001] This invention belongs to the field of power Internet of Things technology, specifically relating to an online leakage current monitoring system and method based on Mesh networking. Background Technology

[0002] With the rapid development of power Internet of Things (IoT) technology, electrical safety monitoring has become an important component of smart grid construction. In application scenarios such as low-voltage distribution network areas, industrial plants, and smart buildings, leakage current monitoring, as a key technology for preventing electrical fires and ensuring personal safety, is of great significance for realizing intelligent operation and maintenance and safety management of power systems. Especially in environments with strong electromagnetic interference, such as industrial plants, accurate leakage current detection is of even greater practical importance for ensuring the safe operation of equipment and preventing electrical accidents. Traditional electrical safety management mainly relies on periodic inspections and manual testing, which is insufficient to meet the demands of modern power environments for real-time monitoring and rapid response. Summary of the Invention

[0003] The purpose of this invention is to provide a leakage current online monitoring system and method based on Mesh networking, so as to at least solve or improve one of the problems existing in the prior art.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] In a first aspect, the present invention provides an online leakage current monitoring system based on a mesh network, comprising:

[0006] Multiple monitoring devices are installed near the power distribution line under test. Each monitoring device includes a processing and control module, a leakage current detection module, and a communication module. The leakage current detection module is used to collect leakage current data of the power distribution line under test. The communication module is used to transmit wireless data with other monitoring devices based on a wireless communication protocol. The processing and control module is used to evaluate the communication quality between nodes through the communication module and to control the communication module to adaptively expand the coverage of the Mesh network based on the communication quality in order to establish a multi-hop transmission path.

[0007] The data aggregation node is the monitoring device with the highest overall signal quality score in the Mesh network. The processing and control module of the data aggregation node is also configured to collect network topology status data of the Mesh network. The communication module of the data aggregation node is also configured to receive monitoring data packets containing leakage current data transmitted via multi-hop transmission paths. The processing and control module of the data aggregation node is further configured to extract topology feature parameters based on the network topology status data and identify the current application scenario based on the topology feature parameters. Based on the identified application scenario, it adaptively sets tiered early warning thresholds and regional alarm conditions. Furthermore, it issues early warnings based on the tiered early warning thresholds and regional alarm conditions, as well as the received leakage current data. The topology feature parameters include node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity.

[0008] The aforementioned solution, through dynamic evaluation of communication quality and adaptive expansion of the mesh network by the processing control module, intelligently selects the optimal communication path in scenarios such as industrial plants with strong electromagnetic interference, effectively avoiding interference and ensuring the stability and reliability of leakage current data feedback, thus providing a data foundation for accurate early warning. By dynamically electing the node with the highest comprehensive signal quality score as the data aggregation node, the core hub of the network is ensured to be in optimal communication condition, optimizing the data transmission path, reducing transmission latency and packet loss rate, and improving the response speed and processing efficiency of the entire monitoring system. By extracting topology feature parameters to automatically identify application scenarios and adaptively setting graded early warning thresholds and regional alarm conditions accordingly, the system can provide differentiated and precise early warnings based on the leakage current risk characteristics of different scenarios, reducing the false alarms and missed alarms caused by the traditional fixed threshold mode in complex environments.

[0009] In a second aspect, the present invention provides a method for online monitoring of leakage current based on a mesh network, comprising the following steps:

[0010] Each monitoring device was installed near the power distribution line to be tested;

[0011] Leakage current data of the power distribution line under test is collected in a non-contact manner by using an open-close zero-sequence current transformer in the monitoring equipment.

[0012] Each monitoring device is used as a node in the Mesh network, and the nodes communicate with each other based on a wireless communication protocol. By evaluating the communication quality between nodes, the coverage of the Mesh network is adaptively expanded according to the communication quality to establish a multi-hop transmission path. During the construction of the Mesh network, the node with the highest overall signal quality score in the Mesh network is used as the data aggregation node, which collects the network topology status data of the Mesh network.

[0013] Based on the multi-hop transmission path, the monitoring data packet containing leakage current data is transmitted to the data aggregation node in multiple hops;

[0014] The data aggregation node extracts topology feature parameters based on network topology status data and identifies the current application scenario based on the topology feature parameters; among them, the topology feature parameters include node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity;

[0015] The data aggregation node adaptively sets tiered early warning thresholds and regional alarm conditions based on the identified application scenarios; furthermore, the data aggregation node issues early warnings based on the tiered early warning thresholds, regional alarm conditions, and leakage current data.

[0016] The aforementioned solution organically integrates dispersed monitoring devices through a self-organizing mesh network, achieving full automation from data acquisition and wireless transmission to intelligent analysis. This transforms the traditional, outdated model that relies on manual inspections and wired connections. It significantly reduces manual intervention, lowers maintenance costs, and enables 24 / 7 uninterrupted real-time monitoring, allowing for timely detection of potential hazards and prevention of electrical fires. By evaluating the communication quality between nodes, the system adaptively expands the mesh network's coverage based on communication quality to establish multi-hop transmission paths. This allows the mesh network to proactively perceive the communication environment and dynamically adjust its topology. In complex environments such as industrial plants with strong magnetic interference or severe building obstruction, the system intelligently selects or constructs high-quality communication links, ensuring reliable and stable transmission of leakage current data to the data aggregation node, laying a solid foundation for accurate early warning. The network topology status data is combined with the intensity of environmental electromagnetic interference as characteristic parameters for identifying actual application scenarios.

[0017] Furthermore, each monitoring device is installed near the power distribution line to be tested, including: using a neodymium iron boron permanent magnet magnetic chassis to attach each monitoring device to the surface of the corresponding power distribution box.

[0018] The above solution utilizes a neodymium iron boron permanent magnet magnetic chassis for installation, solving the problems of complex construction, high cost, and safety hazards associated with traditional wired installation methods that require power outages. This installation method eliminates the need for specialized tools and complex wiring, enabling rapid deployment and flexible adjustment of monitoring equipment. It significantly improves installation efficiency and completely eliminates the risk of power outages during installation, making it particularly suitable for situations requiring rapid deployment or where power outages are inconvenient.

[0019] Furthermore, leakage current data of the power distribution line under test is collected in a non-contact manner using an on / off zero-sequence current transformer in the monitoring equipment, including:

[0020] Open the separable iron core of the openable zero-sequence current transformer and clamp the three-phase conductors and neutral wire to be monitored;

[0021] A separable iron core can be closed to form a complete magnetic circuit;

[0022] When a ground leakage fault occurs, a voltage signal proportional to the leakage current is induced on the secondary side of the zero-sequence current transformer.

[0023] The sensed voltage signal is converted and processed, and then encapsulated into a monitoring data packet by combining node identifier and timestamp.

[0024] The above scheme uses a switchable zero-sequence current transformer for non-contact measurement, achieving truly uninterrupted installation and maintenance, and further improving operational safety. By encapsulating the induced signal with node identifiers and timestamps into a data packet, the integrity and traceability of the monitoring data are ensured, providing accurate data support for subsequent fault location and analysis.

[0025] Furthermore, by evaluating the communication quality between nodes, the coverage of the Mesh network is adaptively expanded based on the communication quality to establish multi-hop transmission paths, including:

[0026] Each node collects communication quality parameters within its communication range; these parameters include RSSI signal strength index, LQI link quality index, and PER packet error rate index.

[0027] The intensity of environmental electromagnetic interference is measured by spectrum scanning, and corresponding dynamic weights are set for each communication quality parameter based on the intensity of environmental electromagnetic interference.

[0028] Based on communication quality parameters and corresponding dynamic weights, calculate the comprehensive signal quality score of the link with neighboring nodes;

[0029] A list of neighboring nodes is established and maintained based on the comprehensive signal quality score;

[0030] When the number of neighboring nodes is insufficient, the network coverage is expanded by gradually increasing the search radius, forming a multi-hop transmission path.

[0031] The above scheme overcomes the inaccuracy of relying solely on RSSI in strong electromagnetic interference environments by comprehensively employing multi-dimensional communication quality parameters such as RSSI, LQI, and PER, and dynamically adjusting their weights in conjunction with the intensity of environmental electromagnetic interference. This results in a more comprehensive and accurate evaluation of communication links. Furthermore, by progressively expanding the search radius, the system can resiliently establish effective multi-hop transmission paths even in situations with sparse network nodes or severe signal obstruction, significantly improving network coverage and robustness.

[0032] Furthermore, the node with the highest overall signal quality score in the mesh network is designated as the data aggregation node. This data aggregation node collects network topology status data of the mesh network, including:

[0033] Each node periodically broadcasts status information containing its own comprehensive signal quality score;

[0034] Each node receives and compares status information from other nodes, and jointly identifies the node with the highest overall signal quality score in the network as the data aggregation node.

[0035] Data aggregation nodes collect and maintain topology status data, including the number of neighboring nodes, inter-node communication hops, and network coverage.

[0036] The data aggregation node calculates node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity as topology characteristic parameters based on the collected topology status data.

[0037] The above scheme uses periodic broadcasting and comparison of state information among nodes to jointly confirm the data aggregation node, achieving a decentralized dynamic election mechanism. This ensures that the core network node is always optimal, enhancing the network's adaptability. The data aggregation node collects and calculates specific topological feature parameters, providing a quantitative and reliable decision-making basis for subsequent accurate scene identification.

[0038] Furthermore, based on a multi-hop transmission path, the monitoring data packets containing leakage current data are transmitted to the data aggregation node via multiple hops, including:

[0039] The monitoring data packets are encapsulated into network data packets conforming to the Bluetooth Mesh protocol format, and the source node address, destination node address, packet sequence number and initial TTL hop value are added to the network data packets;

[0040] A controlled flooding routing mechanism is adopted, in which the source node broadcasts network data packets to neighboring nodes in its list of neighboring nodes whose overall signal quality score is higher than a preset threshold.

[0041] When a relay node receives a network data packet, it checks the packet sequence number and source node address of the network data packet. If the current network data packet has not been forwarded and the TTL value is greater than 0, the TTL value is decremented by 1 and the packet is forwarded to a neighboring node that meets the signal quality requirements.

[0042] When the TTL value drops to 0 or the number of consecutive transmission failures reaches a preset number, forwarding stops and network topology reconstruction is triggered.

[0043] The data aggregation node receives multiple network data packets originating from the same monitoring data packet through different paths. The data aggregation node performs deduplication on the multiple network data packets based on the source node address and packet sequence number, retains the network data packet that arrives first, and returns an acknowledgment message to the source node. If the source node does not receive an acknowledgment message within a preset time, it resends the monitoring data packet.

[0044] The above scheme employs controlled flooding routing and restricts forwarding to only high-quality neighbor nodes, effectively avoiding broadcast storms and controlling network traffic while ensuring data transmission reliability. Through TTL and retransmission mechanisms, it prevents packets from endlessly looping in the network and ensures reliable end-to-end transmission. The deduplication mechanism at the data aggregation node avoids resource waste caused by redundant data processing. These mechanisms work together to build an efficient, reliable, and self-healing data transmission network.

[0045] Furthermore, the data aggregation node extracts topology feature parameters based on network topology status data, and identifies the current application scenario based on these parameters, including:

[0046] The three parameters of node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity are normalized; the normalized parameters are then mapped to a three-dimensional coordinate system with node distribution density as the X-axis, average physical distance as the Y-axis, and environmental electromagnetic interference intensity as the Z-axis.

[0047] Based on preset scene judgment rules, the current scene is identified as one of the following: smart building, industrial plant, or low-voltage distribution substation.

[0048] The above scheme maps topological feature parameters to a three-dimensional coordinate system and performs scene recognition according to preset rules. It transforms complex network topology and environmental conditions into quantifiable and clear scene classifications, making the scene recognition process based on evidence and with clear logic. This improves the accuracy and consistency of the recognition results and lays a solid technical foundation for the subsequent adaptive setting of early warning parameters.

[0049] Furthermore, the scene determination rule is as follows:

[0050] When the node distribution density is greater than 0.8 and the average physical distance is less than 0.3, it is determined to be a smart building scenario;

[0051] When the node distribution density is less than 0.3 and the environmental electromagnetic interference intensity is greater than 0.7, it is determined to be an industrial plant scenario;

[0052] When the conditions of node distribution density being greater than 0.8 and average physical distance being less than 0.3 are not met, and the conditions of node distribution density being less than 0.3 and environmental electromagnetic interference intensity being greater than 0.7 are also not met, it is determined to be a low-voltage distribution substation scenario.

[0053] The above solution provides specific and quantifiable scene judgment rules, which accurately characterize the typical features of smart buildings (high density, close proximity), industrial plants (low density, strong interference), and low-voltage distribution transformer areas (general conditions), ensuring the practicality and effectiveness of scene recognition.

[0054] Furthermore, the data aggregation node adaptively sets tiered early warning thresholds and regional alarm conditions based on the identified application scenarios; and the data aggregation node issues early warnings based on the tiered early warning thresholds, regional alarm conditions, and leakage current data, including:

[0055] When the scene is identified as a smart building, the first-level warning threshold is set to 10mA and the second-level alarm threshold is set to 50mA. When any node detects that the leakage current exceeds the corresponding threshold, the area alarm is triggered.

[0056] When the scene is identified as an industrial plant, the first-level warning threshold is set to 100mA and the second-level alarm threshold is set to 300mA. When 5 or more nodes detect leakage current exceeding the corresponding threshold, an area alarm is triggered.

[0057] When the scenario is identified as a low-voltage distribution transformer area, the first-level warning threshold is set to 30mA, the second-level alarm threshold is set to 100mA, and the area alarm is triggered when three or more nodes detect leakage current exceeding the corresponding threshold.

[0058] The above solution sets differentiated, tiered early warning thresholds and regional alarm conditions for different identified scenarios, enabling refined and scenario-specific customization of the early warning strategy. This improves alarm accuracy: low thresholds and rapid alarms are used in smart buildings where personal safety is highly sensitive; high thresholds with multi-node confirmation are used in industrial plants with high interference and potentially high normal leakage current to prevent false alarms; and intermediate values ​​are used in low-voltage distribution transformer areas. This differentiated strategy ensures the system maintains optimal early warning performance in various application environments. Attached Figure Description

[0059] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0060] Figure 1 This is a schematic diagram of a mesh-based online leakage current monitoring system according to some embodiments of this application;

[0061] Figure 2 This is an exemplary flowchart illustrating an online leakage current monitoring method based on Mesh networking, according to some embodiments of this application.

[0062] Figure 3 This is a schematic diagram illustrating an application scenario for leakage current monitoring in a large industrial park power distribution system, based on some embodiments of this application.

[0063] Figure 4 This is a schematic diagram of the Bluetooth Mesh data packet structure according to some embodiments of this application;

[0064] Figure 5 This is a schematic diagram of network topology and transmission path according to some embodiments of this application. Detailed Implementation

[0065] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0066] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used in this invention is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0067] Example 1

[0068] To address the low accuracy of leakage current detection under strong electromagnetic interference, this application provides an online leakage current monitoring system based on mesh networking. This system achieves wide-area wireless coverage through non-contact, uninterrupted installation of open-type zero-sequence current transformers and adaptive mesh network topology expansion based on the Nordic nRF52832 chip. Furthermore, it improves leakage current detection accuracy under strong electromagnetic interference by automatically identifying and setting differentiated warning thresholds according to application scenarios.

[0069] like Figure 1 As shown, a mesh-based online leakage current monitoring system includes:

[0070] Multiple monitoring devices are installed near the power distribution line under test. Each monitoring device includes a processing and control module, a leakage current detection module, and a communication module. The leakage current detection module is used to collect leakage current data of the power distribution line under test. The communication module is used to transmit wireless data with other monitoring devices based on a wireless communication protocol. The processing and control module is used to evaluate the communication quality between nodes through the communication module and to control the communication module to adaptively expand the coverage of the Mesh network based on the communication quality in order to establish a multi-hop transmission path.

[0071] The data aggregation node is the monitoring device with the highest overall signal quality score in the Mesh network. The processing and control module of the data aggregation node is also configured to collect network topology status data of the Mesh network. The communication module of the data aggregation node is also configured to receive monitoring data packets containing leakage current data transmitted via multi-hop transmission paths. The processing and control module of the data aggregation node is further configured to extract topology feature parameters based on the network topology status data and identify the current application scenario based on the topology feature parameters. Based on the identified application scenario, it adaptively sets tiered early warning thresholds and regional alarm conditions. Furthermore, it issues early warnings based on the tiered early warning thresholds and regional alarm conditions, as well as the received leakage current data. The topology feature parameters include node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity.

[0072] In one embodiment, the monitoring device can be installed on the surface of the distribution box and installed and removed using a neodymium iron boron permanent magnet magnetic chassis.

[0073] In one embodiment, the leakage current detection module uses a switchable zero-sequence current transformer for non-contact leakage current detection and uninterrupted power-off installation. By clamping the three-phase conductors and neutral wire with a separable iron core, a voltage signal proportional to the leakage current is induced when a ground leakage fault occurs.

[0074] In one embodiment, the communication module uses the Nordic nRF52832 chipset for networking based on the Bluetooth Mesh protocol.

[0075] In one embodiment, the processing control module assesses communication quality by collecting RSSI signal strength, LQI link quality, and PER packet error rate indicators. It sets weights based on the intensity of environmental electromagnetic interference and calculates a comprehensive signal quality score, establishing a neighbor node list. When there are insufficient neighbor nodes, it expands the search radius and multi-band search to extend the coverage area. The processing control module also performs scene early warning analysis based on the collected leakage current data. It establishes a three-dimensional coordinate system based on node distribution density, average physical distance, and environmental electromagnetic interference intensity to identify scenes. Scenes are classified into smart buildings, industrial plants, or low-voltage distribution substations, and differentiated graded early warning thresholds and regional alarm conditions are set according to scene categories.

[0076] In an optional embodiment, a power supply module is also included, which collects light energy through a monocrystalline silicon solar panel and provides independent power to the system in conjunction with a protective lithium battery.

[0077] Example 2

[0078] Based on the same inventive concept as Embodiment 1, Embodiment 2 of the present invention also provides a method for online leakage current monitoring based on Mesh networking, such as... Figure 2 As shown, steps S1 to S6 are included:

[0079] S1. Install each monitoring device near the power distribution line to be tested.

[0080] In one embodiment, a neodymium iron boron permanent magnet magnetic chassis can be used to attach each monitoring device to the surface of the corresponding distribution box.

[0081] S2. By using the open-close zero-sequence current transformer in the monitoring equipment, leakage current data of the power distribution line under test is collected in a non-contact manner.

[0082] In one embodiment, the switchable zero-sequence current transformer clamps the three-phase conductors and the neutral line in a non-contact manner through a separable iron core. When a ground leakage fault occurs, a voltage signal is induced on the secondary side; the induced voltage signal is proportional to the leakage current.

[0083] Specifically, the separable core of the openable zero-sequence current transformer is opened to clamp the three-phase conductors and neutral wire to be monitored; the separable core is closed to form a complete magnetic circuit; when a ground leakage fault occurs, a voltage signal proportional to the leakage current is induced on the secondary side of the zero-sequence current transformer; the induced voltage signal is converted and processed, and combined with node identifiers and timestamps to encapsulate it into a monitoring data packet.

[0084] As an optional example, the openable zero-sequence current transformer collects line leakage current data in a non-contact manner for online monitoring. Specifically, the separable core of the openable zero-sequence current transformer is opened, clamping the three-phase conductors and neutral wire to be monitored, so that the three-phase conductors and neutral wire pass through the center of the transformer core; the closed core forms a complete magnetic circuit, establishing a non-contact electromagnetic coupling detection circuit; when the line is operating normally, the vector sum of the three-phase current and the neutral wire current is zero, and the secondary output of the zero-sequence current transformer is zero; when a ground leakage fault occurs, the imbalance between the phase wire and neutral wire currents generates a zero-sequence current, inducing a voltage signal proportional to the leakage current on the secondary side of the transformer; when the detected leakage current value is greater than a preset threshold, an anomaly flag is triggered and the timestamp of the anomaly and the leakage current amplitude are recorded; the collected leakage current data is converted from digital to analog and processed, and the data is encapsulated with node ID, timestamp, and anomaly flag as a monitoring data packet.

[0085] Among them, the switchable zero-sequence current transformer refers to a zero-sequence current detection device with an openable structure design. The iron core can be separated and opened for installation, and current detection can be achieved without disconnecting the conductor being tested. In this solution, the switchable zero-sequence current transformer is the core component for leakage current detection. It detects the zero-sequence current in a three-phase four-wire system through the principle of electromagnetic induction, realizing non-contact monitoring of ground leakage faults. Compared with traditional fixed transformers, it has the advantages of convenient installation and no power outage operation.

[0086] Separable core: This refers to the magnetic conductor core of a zero-sequence current transformer, which adopts a detachable ring or square structure and is usually made of silicon steel sheets or ferrite materials with high magnetic permeability. In this solution, the separable core design allows the device to be installed without disconnecting the conductor being tested. By opening the core, clamping the conductor, and then closing it again, a complete magnetic circuit is formed to achieve electromagnetic induction detection.

[0087] The center of the transformer core refers to the hollow area inside the core of the zero-sequence current transformer, which is the passage through which the conductor being tested passes. In this scheme, the three-phase conductors and the neutral wire need to be bundled and pass through the center of the transformer core to ensure that all current-carrying conductors are surrounded by the core. When a ground leakage occurs, the unbalanced current generates a change in magnetic flux in the core, thereby inducing a detection signal in the secondary winding.

[0088] Non-contact electromagnetic coupling detection circuit: This refers to a detection system based on the principle of electromagnetic induction that operates without physical contact, transmitting and detecting current signals through magnetic field coupling. In this scheme, the detection circuit consists of the primary side conductors under test (three-phase lines and neutral line), an iron core magnetic circuit, and a secondary side detection winding. The magnetic field generated by the primary side current is coupled to the secondary side through the iron core, achieving current detection under electrically isolated conditions.

[0089] The secondary side of a zero-sequence current transformer refers to the output winding of the transformer and its connected detection circuit, used to collect and process the induced voltage signal. In this scheme, the secondary side includes a detection winding, signal conditioning circuit, and analog-to-digital converter. When a zero-sequence current appears on the primary side, the secondary side outputs a voltage signal proportional to the leakage current. After amplification, filtering, and digital processing, this provides the data basis for leakage current detection and early warning.

[0090] S3. Each monitoring device is used as a node in the Mesh network, and the nodes communicate with each other based on a wireless communication protocol. By evaluating the communication quality between nodes, the coverage of the Mesh network is adaptively expanded according to the communication quality to establish a multi-hop transmission path. During the construction of the Mesh network, the node with the highest comprehensive signal quality score in the Mesh network is used as the data aggregation node, and the data aggregation node collects the network topology status data of the Mesh network.

[0091] In one embodiment, by evaluating the communication quality between nodes and adaptively expanding the coverage of the mesh network based on the communication quality, a multi-hop transmission path is established, including:

[0092] Each node collects communication quality parameters within its communication range; these parameters include RSSI signal strength index, LQI link quality index, and PER packet error rate index.

[0093] The intensity of environmental electromagnetic interference is measured by spectrum scanning, and corresponding dynamic weights are set for each communication quality parameter based on the intensity of environmental electromagnetic interference.

[0094] Based on communication quality parameters and corresponding dynamic weights, calculate the comprehensive signal quality score of the link with neighboring nodes;

[0095] A list of neighboring nodes is established and maintained based on the comprehensive signal quality score;

[0096] When the number of neighboring nodes is insufficient, the network coverage is expanded by gradually increasing the search radius, forming a multi-hop transmission path as the Mesh network topology.

[0097] In one embodiment, the node with the highest overall signal quality score in the mesh network is designated as the data aggregation node. The data aggregation node collects network topology status data of the mesh network, including:

[0098] Each node periodically broadcasts status information containing its own comprehensive signal quality score;

[0099] Each node receives and compares status information from other nodes, and jointly identifies the node with the highest overall signal quality score in the network as the data aggregation node.

[0100] Data aggregation nodes collect and maintain topology status data, including the number of neighboring nodes, inter-node communication hops, and network coverage.

[0101] The data aggregation node calculates node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity as topology characteristic parameters based on the collected topology status data.

[0102] As an optional example, each monitoring device constructs a mesh network based on the Nordic nRF52832 chipset and Bluetooth Mesh protocol, expanding the mesh network's coverage area through multi-hop relays. Specifically, the system collects communication quality parameters and environmental electromagnetic interference intensity within a preset range of the monitoring devices. Communication quality parameters include RSSI signal strength index, LQI link quality index, and PER packet error rate index. The communication quality parameters are preprocessed to obtain standardized scores. Dynamic weights for each communication quality parameter are set according to the environmental electromagnetic interference intensity. A comprehensive signal quality score is calculated based on the standardized scores and corresponding dynamic weights. When the comprehensive signal quality score exceeds a threshold, a neighbor node list containing node ID, comprehensive signal quality score, and network hop count is established. When the neighbor node list is empty or the number of nodes is less than two, the search radius is progressively expanded to obtain a mesh network topology with expanded coverage.

[0103] The Received Signal Strength Indicator (RSSI) refers to the power intensity of the wireless signal received by the Nordic RF52832 chipset, expressed in dBm. In this scheme, RSSI is used to evaluate the communication distance and signal propagation quality between monitoring nodes; a higher value indicates a stronger signal and better communication quality. In environments with strong electromagnetic interference, RSSI is easily affected by electromagnetic noise generated by metal equipment and frequency converters, leading to significant fluctuations in the measured value. Therefore, it needs to be used in conjunction with other parameters to improve the accuracy of the evaluation.

[0104] Link Quality Indicator (LQI): In the Bluetooth Mesh protocol, LQI is a comprehensive metric used to evaluate data packet transmission quality and link stability, expressed as a value ranging from 0 to 255. In this scheme, LQI reflects the reliability of the communication link between nodes by analyzing parameters such as the integrity of received data packets, symbol error rate, and signal modulation quality. Compared to RSSI, LQI is less sensitive to electromagnetic interference and can more accurately assess communication quality in environments with strong electromagnetic interference; therefore, its weight is increased when strong interference is detected.

[0105] Packet Error Rate (PER): This refers to the ratio of the number of failed data packets to the total number of transmitted data packets within a certain period, expressed as a percentage. In this solution, PER directly reflects the transmission reliability of the communication link by statistically analyzing the packet loss situation of the Nordic nRF52832 chipset in a Bluetooth Mesh network. A lower PER value indicates more stable communication and is an important indicator for evaluating network connection quality, maintaining a relatively stable weight under various environmental conditions.

[0106] Furthermore, dynamic weights for each communication quality parameter are set based on the intensity of environmental electromagnetic interference, including: performing a spectrum scan using the Nordic nRF52832 chipset to measure the noise floor level in the 2.4GHz ISM band used by the Bluetooth Mesh protocol; when the noise floor level is less than or equal to a threshold, it is determined to be a normal environment, and the RSSI signal strength index weight is set to [value missing]. The weight of the LQI link quality index is The weight of the PER error rate indicator is: When the noise floor level exceeds a threshold, it is determined to be a strong magnetic interference environment, and the RSSI signal strength index weight is set to [value missing]. The weight of the LQI link quality index is The weight of the PER error rate indicator is: ;

[0107] The Nordic nRF52832 chipset refers to a low-power Bluetooth system-on-a-chip manufactured by Nordic Semiconductor of Norway. It integrates an ARM Cortex-M4F processor, a 2.4GHz RF transceiver, and a rich set of peripheral interfaces. In this solution, the nRF52832 chipset serves as the core communication module of the monitoring equipment, responsible for executing the Bluetooth Mesh protocol stack, performing spectrum scanning, measuring communication quality parameters, and enabling wireless data transmission and network building between devices.

[0108] Bluetooth Mesh Protocol: This refers to a many-to-many network communication protocol based on Bluetooth Low Energy (BLE) technology. It employs controlled flooding and relay mechanisms to achieve wide-range device interconnection. In this solution, the Bluetooth Mesh protocol is used to build a self-organizing network (Mesh network) among monitoring devices, supporting multi-hop data transmission, expanding network coverage, and providing network self-healing and dynamic topology adjustment capabilities.

[0109] The 2.4GHz ISM band refers to the unlicensed industrial, scientific, and medical band allocated by the International Telecommunication Union, ranging from 2.400 to 2.485 GHz. In this solution, the Bluetooth Mesh protocol operates within this band, using three broadcast channels—No. 37 (2402 MHz), No. 38 (2426 MHz), and No. 39 (2480 MHz)—for device discovery and data transmission. This band is also the operating frequency for devices such as WiFi and microwave ovens, making it prone to interference.

[0110] Noise floor level: refers to the power intensity of ambient electromagnetic noise in the 2.4GHz ISM band, measured using the spectrum scanning function of the Nordic RF52832 chipset, and expressed in dBm. In this scheme, the noise floor level is used to determine the current ambient electromagnetic interference intensity, serving as the basis for dynamically adjusting the weights of communication quality parameters.

[0111] Furthermore, the weight values ​​for normal environments are as follows:

[0112] (RSSI weight): 0.5 or 0.6;

[0113] (LQI weight): 0.2 or 0.3;

[0114] (PER weight): 0.2 or 0.3.

[0115] Weighting values ​​for strong magnetic interference environments:

[0116] (RSSI weight): 0.2 or 0.3;

[0117] (LQI weight): 0.4 or 0.5;

[0118] (PER weight): 0.2 or 0.3.

[0119] Among them, each weight satisfies the normalization condition: , And satisfy , , The relationship.

[0120] Furthermore, when the neighbor node list is empty or the number of nodes is less than 2, the search radius is progressively expanded to obtain a mesh network topology with expanded coverage. This includes setting the initial search parameters for the Nordic nRF52832 chipset and setting the transmit power to... The search radius is set to The search time interval is set to The system performs an omnidirectional search within a 360-degree radius of the current location of the monitoring device, broadcasting Beacon frames and listening for responses from surrounding devices. If the omnidirectional search fails, it switches to a sector-based directional search, narrowing the search angle to a ±45-degree sector area. An omnidirectional search fails when no response is received after the search interval ends. After the search interval ends, the system checks the neighbor node list; if the list is empty or contains fewer than two nodes, the transmission power is set to... The search radius is set to The search time interval is set to The system employs a multi-band search mode, sequentially switching between broadcast channels 37, 38, and 39 in the 2.4GHz ISM band to discover neighbors, thereby increasing the discovery probability and avoiding interference from a single channel. If the system fails to discover a preset number of neighbor nodes after N consecutive rounds of searching, it activates an isolated node survival mode, where the corresponding node periodically broadcasts its own location while simultaneously listening for joining signals from other nodes. This search strategy is repeated until the transmit power reaches the preset maximum value dBmax, the search radius reaches the preset maximum value Rmax, or the number of valid nodes in the neighbor node list reaches two or more, thus forming a Mesh network topology with expanded coverage.

[0121] Beacon frames, in the Bluetooth Mesh protocol, are broadcast data packets used for device discovery and network joining. They contain key parameters such as device identifier, network information, and signal strength. In this scheme, monitoring devices periodically broadcast Beacon frames to announce their presence and simultaneously listen for beacon frames sent by other devices to discover potential neighboring nodes. This is the fundamental communication mechanism for building the Mesh network topology.

[0122] ±45-degree sector area: This refers to a 90-degree sector-shaped coverage area extending 45 degrees to the left and right of the current location of the monitoring device and its current orientation as the central axis. In this solution, when the 360-degree omnidirectional search fails, the system switches to directional search mode, concentrating the search range within the ±45-degree sector area. By increasing the signal power density and search accuracy in this direction, the probability of discovering neighboring nodes is increased.

[0123] Neighbor node list: This refers to the data structure stored in each monitoring device, recording information about neighboring devices that can currently communicate directly. It includes parameters such as node ID, overall signal quality score, network hop count, and connection status. In this solution, the neighbor node list is the core data foundation for building the Mesh network topology, used for path selection, data forwarding, and network maintenance.

[0124] Channels 37, 38, and 39 refer to the three dedicated broadcast channels defined by the Bluetooth Low Energy protocol in the 2.4 GHz ISM band, corresponding to frequencies of 2402 MHz, 2426 MHz, and 2480 MHz, respectively. In this solution, the multi-band search mode avoids the electromagnetic interference of a single channel by polling and switching between these three channels, thereby improving the success rate of device discovery and the reliability of network connections.

[0125] Orphaned node survival mode: This refers to a backup working mode activated when the monitoring device fails to find enough neighboring nodes after multiple rounds of searching. The device periodically broadcasts its own location information and remains in a receiving state, waiting for other nodes to join. In this solution, the orphaned node survival mode ensures that the device can maintain basic network discovery capabilities even under extremely sparse deployment conditions, providing a foundation for subsequent network expansion.

[0126] Extended coverage Mesh network topology: This refers to an adaptive network architecture formed through progressively expanding search radius and multi-mode discovery mechanisms, capable of dynamically adjusting topological connections based on node distribution density. In this scheme, the extended coverage Mesh network topology achieves network expansion from locally dense connections to large-scale sparse coverage, ensuring the diversity of data transmission paths and network robustness.

[0127] Furthermore, the range of transmit power values ​​is as follows:

[0128] (Initial transmit power): -20dBm to -12dBm;

[0129] (Extended transmit power): -8dBm to 0dBm;

[0130] (Maximum transmit power): +4dBm to +8dBm.

[0131] Search radius range:

[0132] (Initial search radius): 10m to 30m;

[0133] (Extended search radius): 50m to 100m;

[0134] (Maximum search radius): 150m to 300m.

[0135] Search time interval range:

[0136] (Initial search time): 5s to 15s;

[0137] T 2 (Extended search time): 20s to 45s;

[0138] (Maximum search time): 60s to 120s.

[0139] S4. Based on the multi-hop transmission path, the monitoring data packet containing leakage current data is transmitted to the data aggregation node in multiple hops.

[0140] In one embodiment, based on a multi-hop transmission path, the monitoring data packet containing leakage current data is transmitted to the data aggregation node via multiple hops, including:

[0141] The monitoring data packets are encapsulated into network data packets conforming to the Bluetooth Mesh protocol format, and the source node address, destination node address, packet sequence number and initial TTL hop value are added to the network data packets;

[0142] A controlled flooding routing mechanism is adopted, in which the source node broadcasts network data packets to neighboring nodes in its list of neighboring nodes whose overall signal quality score is higher than a preset threshold.

[0143] When a relay node receives a network data packet, it checks the packet sequence number and source node address of the network data packet. If the current network data packet has not been forwarded and the TTL value is greater than 0, the TTL value is decremented by 1 and the packet is forwarded to a neighboring node that meets the signal quality requirements.

[0144] When the TTL value drops to 0 or the number of consecutive transmission failures reaches a preset number, forwarding stops and network topology reconstruction is triggered.

[0145] The data aggregation node receives multiple network data packets originating from the same monitoring data packet through different paths. The data aggregation node performs deduplication on the multiple network data packets based on the source node address and packet sequence number, retains the network data packet that arrives first, and returns an acknowledgment message to the source node. If the source node does not receive an acknowledgment message within a preset time, it resends the monitoring data packet.

[0146] As an optional example, a mesh network is used to transmit leakage current data to the data aggregation node via multi-hop transmission. Specifically, each monitoring device encapsulates the generated monitoring data packets into network data packets conforming to the Bluetooth Mesh protocol format, adding the source node address, packet sequence number, TTL hop limit, and data aggregation node address. The data aggregation node is the node with the highest overall signal quality score in the network. Based on the Mesh network topology, a controlled flooding routing mechanism is used for data packet transmission. The source node broadcasts the data packet to nodes in its neighbor list whose overall quality score is greater than a threshold. Each data packet has an initial TTL hop value of 8. After receiving the data packet, each relay node first checks the packet sequence number and source node address. If the corresponding data packet has already been forwarded... If a data packet has already been sent, it is discarded to avoid duplicate transmission. If a data packet has not been forwarded and the TTL hop value is greater than 0, the TTL value is decremented by 1 and forwarded to a neighboring node that meets the signal quality requirements. Forwarding stops when the TTL hop value drops to 0 to prevent infinite loop propagation. When three consecutive data packet transmission failures are detected in the network, the network topology reconstruction mechanism is triggered, and the neighbor discovery process is re-executed. After receiving the same data packets from various paths, the data aggregation node performs deduplication based on the source node address and packet sequence number, retains the first arriving data packet, and sends an acknowledgment message to the source node. If the source node does not receive the acknowledgment message within a preset time, the retransmission mechanism is initiated.

[0147] The network data packet refers to a transmission data unit encapsulated in the Bluetooth Mesh protocol standard format, containing the payload of the monitoring data packet and the control information required for network transmission. In this solution, the network data packet is encapsulated from the original leakage current monitoring data packet (including node ID, timestamp, leakage current value, and anomaly identifier), with added network layer information such as source node address, destination node address (data aggregation node), packet sequence number, and TTL hop limit to ensure that the data can be correctly routed and transmitted in the Mesh network.

[0148] Time To Live (TTL) limits network packets to a maximum number of times they can be forwarded in a mesh network. This prevents packets from looping indefinitely. In this scheme, each network packet has an initial TTL of 8, meaning it can be forwarded a maximum of 8 times. The TTL is decremented by 1 each time a packet passes through a relay node. When the TTL reaches 0, the packet is dropped and no longer forwarded, thus avoiding network congestion and resource waste.

[0149] Controlled flooding routing mechanism: This refers to a data transmission method that adds control conditions and restriction policies to traditional flooding routing, optimizing network performance by setting forwarding conditions. In this scheme, the controlled flooding mechanism includes multiple control conditions: forwarding only to nodes in the neighbor node list whose comprehensive signal quality score is greater than a threshold, avoiding duplicate forwarding by checking packet sequence number and source node address, setting TTL hop limit to prevent infinite propagation, and requiring relay nodes to meet signal quality requirements, etc. Compared with uncontrolled broadcast flooding, the controlled flooding mechanism reduces network load and transmission latency while ensuring data delivery rate.

[0150] S5. The data aggregation node extracts topology feature parameters based on network topology status data and identifies the current application scenario based on the topology feature parameters; among them, the topology feature parameters include node distribution density, average physical distance between nodes and environmental electromagnetic interference intensity.

[0151] In one embodiment, the data aggregation node extracts topology feature parameters based on network topology status data and identifies the current application scenario based on the topology feature parameters, including:

[0152] The three parameters of node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity are normalized; the normalized parameters are then mapped to a three-dimensional coordinate system with node distribution density as the X-axis, average physical distance as the Y-axis, and environmental electromagnetic interference intensity as the Z-axis.

[0153] Based on preset scene judgment rules, the current scene is identified as one of the following: smart building, industrial plant, or low-voltage distribution substation.

[0154] Specifically, the scene determination rules are as follows:

[0155] When the node distribution density is greater than 0.8 and the average physical distance is less than 0.3, it is determined to be a smart building scenario;

[0156] When the node distribution density is less than 0.3 and the environmental electromagnetic interference intensity is greater than 0.7, it is determined to be an industrial plant scenario;

[0157] When the conditions of node distribution density being greater than 0.8 and average physical distance being less than 0.3 are not met, and the conditions of node distribution density being less than 0.3 and environmental electromagnetic interference intensity being greater than 0.7 are also not met, it is determined to be a low-voltage distribution substation scenario.

[0158] As an optional example, the data aggregation node processes the leakage current data received from each monitoring node. Based on the established Mesh network topology, it collects network connection status data over a 30-minute period. This data includes the number of neighboring nodes, inter-node communication hops, and network coverage area. Based on the neighboring node list, it counts the total number of valid nodes in the network and calculates the network coverage area using the search radius parameter. The node distribution density index is calculated by dividing the total number of nodes by the coverage area. Using the collected RSSI signal strength index, the average physical distance between nodes in the network is calculated using free-space propagation loss. The noise floor level in the 2.4GHz ISM band used by the Bluetooth Mesh protocol is used as the electromagnetic interference intensity parameter. The node distribution density, average physical distance, and electromagnetic interference intensity are normalized to establish a three-dimensional coordinate system with node distribution density as the X-axis, average physical distance as the Y-axis, and electromagnetic interference intensity as the Z-axis. Based on this three-dimensional coordinate system, scenarios are classified using preset scenario judgment rules. These scenarios include smart buildings, industrial plants, and low-voltage distribution substations.

[0159] Based on the scenario classification results, set graded early warning thresholds: for low-voltage distribution transformer areas, set a first-level early warning threshold of 30mA and a second-level alarm threshold of 100mA; for industrial plants, set a first-level early warning threshold of 100mA and a second-level alarm threshold of 300mA; for smart buildings, set a first-level early warning threshold of 10mA and a second-level alarm threshold of 50mA.

[0160] The collected leakage current is compared with the warning threshold corresponding to the scenario, and the area alarm conditions are set according to different scenarios: low voltage distribution substation, the area alarm is triggered when 3 or more nodes are abnormal; industrial plant, the area alarm is triggered when 5 or more nodes are abnormal; smart building, the area alarm is triggered when any node is abnormal.

[0161] Network connectivity status data refers to a comprehensive set of parameters reflecting the topology and connectivity characteristics of the Mesh network, including key indicators such as the number of neighboring nodes for each node, the number of communication hops between nodes, and the network coverage. In this solution, network connectivity status data is the fundamental data source for automatic scene identification. By analyzing the network topology change characteristics within 30 minutes, it provides data support for subsequent node distribution density calculations, average physical distance estimations, and scene classification.

[0162] Free-space propagation loss refers to the power attenuation of a wireless signal as it propagates in a vacuum or free space due to increased distance, following a theoretical propagation model that adheres to the inverse square law of distance. In this scheme, the physical distance between nodes within the network is calculated by combining the free-space propagation loss formula with measured RSSI signal strength indicators, providing average physical distance parameters for constructing a three-dimensional coordinate system for scene recognition.

[0163] Average physical distance: refers to the statistical average of the actual physical distances between all adjacent nodes in the network, calculated using RSSI signal strength and free space propagation loss models. In this scheme, the average physical distance serves as the Y-axis parameter of the scene recognition 3D coordinate system, reflecting the spatial distribution characteristics of equipment deployment in different application scenarios. It is an important indicator for distinguishing smart buildings, industrial plants, and low-voltage distribution substations.

[0164] Normalization: This refers to a data preprocessing method that maps parameters with different dimensions and numerical ranges to the [0,1] interval, eliminating the influence of dimensional differences on the analysis results. In this scheme, three heterogeneous parameters—node distribution density (number of nodes / area), average physical distance (meters), and electromagnetic interference intensity (dBm)—are normalized to enable scene classification and comparison within the same coordinate system.

[0165] A three-dimensional coordinate system refers to a spatial coordinate system constructed with node distribution density as the X-axis, average physical distance as the Y-axis, and environmental electromagnetic interference intensity as the Z-axis. In this solution, this coordinate system is the core analysis tool for achieving automatic scene recognition. By mapping the topological characteristics of the monitoring network to different regions in three-dimensional space, it enables automatic classification of three application scenarios: smart buildings, industrial plants, and low-voltage distribution substations.

[0166] Furthermore, based on the three-dimensional coordinate system, the scene is classified using preset scene judgment rules, including: when the node distribution density is greater than 0.8 and the average physical distance is less than 0.3, it is judged as a smart building scene; when the node distribution density is less than 0.3 and the environmental electromagnetic interference intensity is greater than 0.7, it is judged as an industrial plant scene; when neither of the above two conditions is met, it is judged as a low-voltage distribution transformer area scene.

[0167] Smart buildings refer to modern building forms that utilize advanced information technology to achieve intelligent management of building equipment, typically characterized by high-density equipment deployment, short-range communication, and a relatively clean electromagnetic environment. In this solution, the smart building scenario corresponds to areas with a node distribution density greater than 0.8 and an average physical distance less than 0.3, where the most stringent leakage current warning thresholds (10mA Level 1, 50mA Level 2) and the most sensitive area alarm conditions (an alarm is triggered if any node malfunctions) are set.

[0168] Industrial plant: refers to large factory buildings used for industrial production, typically characterized by high-power equipment, strong electromagnetic interference, and relatively sparse monitoring point distribution. In this solution, for industrial plant scenarios, areas with a node distribution density of less than 0.3 and an environmental electromagnetic interference intensity greater than 0.7 are assigned relatively lenient leakage current warning thresholds (100mA Level 1, 300mA Level 2) and higher area alarm conditions (5 or more abnormal nodes).

[0169] Low-voltage distribution area: refers to the coverage area of ​​a low-voltage distribution network powered by a 10kV distribution transformer, typically including power infrastructure such as distribution transformers, distribution boxes, and user terminals. In this solution, the low-voltage distribution area scenario is used as the default scenario for those that do not meet the criteria for smart buildings and industrial plants, and is equipped with a moderate leakage current warning threshold (30mA Level 1, 100mA Level 2) and appropriate area alarm conditions (3 or more abnormal nodes).

[0170] S6. The data aggregation node adaptively sets tiered early warning thresholds and regional alarm conditions based on the identified application scenarios; and the data aggregation node issues early warnings based on the tiered early warning thresholds, regional alarm conditions, and leakage current data.

[0171] In one embodiment, the data aggregation node adaptively sets tiered early warning thresholds and regional alarm conditions based on the identified application scenario; and the data aggregation node issues early warnings based on the tiered early warning thresholds, regional alarm conditions, and leakage current data, including:

[0172] When the scene is identified as a smart building, the first-level warning threshold is set to 10mA and the second-level alarm threshold is set to 50mA. When any node detects that the leakage current exceeds the corresponding threshold, the area alarm is triggered.

[0173] When the scene is identified as an industrial plant, the first-level warning threshold is set to 100mA and the second-level alarm threshold is set to 300mA. When 5 or more nodes detect leakage current exceeding the corresponding threshold, an area alarm is triggered.

[0174] When the scenario is identified as a low-voltage distribution transformer area, the first-level warning threshold is set to 30mA, the second-level alarm threshold is set to 100mA, and the area alarm is triggered when three or more nodes detect leakage current exceeding the corresponding threshold.

[0175] The above scheme utilizes multi-dimensional communication quality parameter fusion evaluation technology to comprehensively evaluate RSSI signal strength, LQI link quality, and PER packet error rate, and dynamically adjusts the weights of each parameter based on the noise floor level of the 2.4GHz ISM band. When a strong interference environment is detected, the RSSI weight is automatically reduced. ) and increase LQI weight ( Leveraging the sensitivity of LQI to link stability and data integrity, it effectively counteracts the impact of electromagnetic interference on a single RSSI parameter, ensuring accurate assessment of communication quality and establishing reliable network connections even in complex industrial environments.

[0176] To address the technical bottleneck of sparse distribution of nodes in low-voltage distribution networks leading to the failure of neighbor discovery algorithms, this solution designs a progressive search radius expansion and a multi-mode adaptive discovery mechanism. When the number of neighbor nodes is insufficient, the system automatically reduces the transmission power from... Upgraded to The search radius is from Expand to Search time from Extended to Simultaneously, multi-band polling search is enabled on three broadcast channels (37, 38, and 39) to avoid discovery failures caused by interference from a single channel. Furthermore, through dynamic switching between omnidirectional search and ±45-degree sector-oriented search, and a fault-tolerance mechanism for isolated node survival mode, the network topology can be gradually established even under extremely sparse distribution conditions, ensuring network connectivity and data transmission reliability in large-scale, low-density deployment scenarios.

[0177] To verify the effectiveness of this invention, a leakage current monitoring project of the power distribution system in a large industrial park is used as an example. Figure 3 As shown, the industrial park covers an area of ​​approximately 150,000 square meters, including 35 workshops, 12 office buildings, and 3 low-voltage distribution substations, with a total of 268 power distribution nodes requiring monitoring. The park contains a large number of operating industrial devices, resulting in a complex electromagnetic environment. The average noise floor level in the 2.4GHz band ranges from -85dBm to -75dBm, and in some areas, due to the operation of high-power frequency converters and induction heating equipment, the environmental electromagnetic interference intensity can reach above -70dBm.

[0178] During the equipment installation phase, the data processing team first collected GPS coordinates and assigned numbers to the 268 distribution boxes. Each monitoring device uses a unified coding rule: park code (3 digits) + area code (2 digits) + device serial number (3 digits), such as "WY1-A1-001" which represents device number 001 in area A1 of industrial park 1.

[0179] The adsorption force data of the NdFeB permanent magnet magnetic chassis was calibrated using a tensile testing instrument. The standard adsorption force is 120N. Under the condition that the surface roughness Ra of the distribution box is 6.3μm, the actual adsorption force remains within the range of 105-115N. The three-dimensional coordinate data of the equipment installation position was collected through a Beidou / GPS dual-mode positioning system with sub-meter accuracy. All coordinate data were uniformly converted to the WGS-84 coordinate system and stored in the equipment information database.

[0180] After physical installation is completed, the system automatically generates a heat map of equipment distribution density. The equipment density in the A1 industrial plant area (12,000 square meters, with 32 devices) is 2.67 devices / thousand square meters; the equipment density in the B2 office building area (3,500 square meters, with 18 devices) is 5.14 devices / thousand square meters; and the equipment density in the C area low-voltage distribution station area (8,000 square meters, with 25 devices) is 3.13 devices / thousand square meters.

[0181] S200, real-time acquisition and processing of power generation performance data for monocrystalline silicon solar panels. Each solar panel has a power output of 15W, under standard test conditions (illuminance 1000W / m²). 2 The battery temperature is 25°C, the open-circuit voltage is 21.6V, and the short-circuit current is 0.85A. In actual operation, the park's geographical location (118.2°E, 31.8°N) has an average annual sunshine duration of 2180 hours, and the average annual power generation of the solar panels is 32.7kWh.

[0182] The protective lithium battery has a capacity of 20Ah, an operating voltage of 3.7V, and an energy storage capacity of 74Wh. Power consumption analysis of the monitoring equipment shows that the Nordic nRF52832 chipset consumes 15mA / 3.3V in Bluetooth Mesh active mode and 2μA / 3.3V in sleep mode; the zero-sequence current transformer acquisition circuit consumes 8mA / 3.3V; and the data processing module consumes 12mA / 3.3V.

[0183] Energy consumption calculation based on a 24-hour working cycle: The device collects data 6 times per hour, with each collection lasting 30 seconds, and operates in low-power mode for the remaining time. The average daily power consumption is: (15+8+12)×0.5×6×24+2×23.5×24=1512+1128=2640μAh≈2.64mAh / day. At this power consumption level, it can operate continuously for approximately 278 days on a full charge, and the solar panels can still provide sufficient supplementary power during cloudy or rainy weather to ensure the system's continuous and stable operation.

[0184] Before the S300 system officially went into operation, it underwent a 72-hour network quality baseline test. Taking device WY1-A1-001 as an example, its 5-meter radius contains 4 neighboring nodes. The raw communication quality data collected over 24 hours is as follows:

[0185] The RSSI signal strength index collected 86,400 sample points, with values ​​ranging from -35dBm to -68dBm, a mean of -48.3dBm, and a standard deviation of 8.7dBm. The LQI link quality index ranged from 85 to 255, with a mean of 178.5 and a standard deviation of 23.4. The PER packet error rate index ranged from 0.2% to 2.8% within a 1-hour statistical window, with a mean of 1.2% and a standard deviation of 0.6%.

[0186] The standardized scoring process uses the Z-score normalization method: This was then mapped to a 0-100 score scale. After processing, the RSSI score was 72, the LQI score was 68, and the PER score was 85.

[0187] A spectrum scan was performed using the Nordic nRF52832 chipset, covering the 2402MHz to 2480MHz band, with a scan resolution of 1MHz and a measurement time of 100ms per frequency point. The spectrum scan results for the A1 industrial plant area showed that the 2420MHz to 2440MHz band was subject to interference from industrial equipment, with an average noise floor level of -72dBm, exceeding the -80dBm threshold, indicating a strong magnetic interference environment.

[0188] Under strong magnetic interference, the weighting parameters are adjusted as follows: (RSSI weight is reduced from 0.4 in normal environment). (The LQI weight is increased from 0.3 in the normal environment). (PER weights remain unchanged). The overall signal quality score is calculated as follows: Score = 72 × 0.2 + 68 × 0.5 + 85 × 0.3 = 14.4 + 34 + 25.5 = 73.9 points.

[0189] Initial search parameter settings: transmit power Search radius Search time interval Device WY1-C3-015 is located at the edge of the distribution area. The initial omnidirectional search only found one neighbor node, which does not meet the requirement of at least two nodes.

[0190] First round of parameter adjustments: Transmit power Search radius Search time interval A sector-shaped directional search was used, with directional scanning conducted in eight directions, including 45 degrees east of north and 45 degrees east of south, with each direction requiring a search time of 7.5 seconds.

[0191] After the multi-band search mode was enabled, the search was conducted alternately between channels 37 (2402MHz), 38 (2426MHz), and 39 (2480MHz). In the second round of searching, device WY1-C3-008 (signal quality score 81) was found at a distance of 18 meters and device WY1-C3-021 (signal quality score 76) at a distance of 22 meters on channel 38, successfully establishing a neighbor list containing 3 nodes. After 6 rounds of progressive searching, the final Mesh network topology contained 247 valid nodes (21 nodes did not join the network due to excessive distance or substandard signal quality), with an average node degree of 3.2, a network diameter of 12 hops, and a connectivity rate of 92.2%.

[0192] Taking device WY1-A1-008 as an example, the S400 monitors a three-phase four-wire power supply line with a rated current of 400A. The switchable zero-sequence current transformer uses a high-permeability silicon steel core with an inner diameter of 120mm and a transformation ratio of 1000:1. During installation, the three-phase conductors (with a cross-sectional area of ​​185mm²) are... 2 ) and neutral line (cross-sectional area of ​​95mm²) 2 Pass through the center of the transformer to ensure that the conductor is symmetrical with the center of the iron core.

[0193] After the instrument transformers were installed, on-site calibration was performed. Using a standard leakage current generator, simulated leakage currents of 5mA, 10mA, 30mA, 50mA, and 100mA were injected, and the secondary output voltage of the instrument transformers was measured. The calibration results showed that 5mA corresponded to an output of 5.2mV, 10mA to 10.1mV, 30mA to 29.8mV, 50mA to 49.9mV, and 100mA to 100.3mV. The linear correlation coefficient r = 0.9997, and the transformation ratio error was less than 0.5%, meeting the measurement accuracy requirements.

[0194] The system uses a 24-bit ΔΣ-type ADC for data acquisition with a sampling frequency of 2kHz, capable of capturing the 50Hz fundamental frequency and its harmonic components. In a specific fault monitoring operation, equipment A1-008 detected an anomaly at 14:32:18 on March 15, 2024: the three-phase currents were 178A, 185A, and 182A respectively, and the neutral current was 8A. Theoretically, the zero-sequence current should be close to zero, but the measured zero-sequence current was 45mA.

[0195] First, a 50Hz power frequency notch filter is applied to remove fundamental frequency interference from the power grid. Then, the RMS value of the leakage current is calculated using the effective value. Finally, a 5-second moving average filter is applied to eliminate transient interference. The processed stable leakage current value is 43.8mA, which exceeds the preset threshold of 30mA. The system automatically triggers an anomaly flag and generates a monitoring data packet containing the node ID (WY1-A1-008), timestamp (20240315143218), leakage current amplitude (43.8mA), and anomaly level (Level 1).

[0196] The S500 monitors data packets by encapsulating them according to the Bluetooth Mesh protocol standard, such as... Figure 4 As shown, the data packet structure includes: a network header (8 bytes), a transport header (4 bytes), application data (32 bytes), and a message integrity check (4 bytes), for a total length of 48 bytes. The source node address is 0x1A08 (corresponding to WY1-A1-008), the data aggregation node address is 0x1C01 (corresponding to the central device in the WY1-C area with the highest signal quality score), the packet sequence number is 0x2F3A, and the initial TTL is set to 8.

[0197] like Figure 5 As shown, the controlled flooding routing mechanism is executed as follows: Device WY1-A1-008 broadcasts data packets to three nodes in its neighbor node list (WY1-A1-005, WY1-A1-012, and WY1-A1-015). The signal quality scores of these nodes are 81, 76, and 79 respectively, all exceeding the transmission threshold of 70. See Table 1 for transmission parameters.

[0198] Table 1 Transmission Parameters

[0199]

[0200] First hop: After receiving the data packet, WY1-A1-005 decrements the TTL by 1 to 7 and forwards it to its two neighboring nodes (WY1-A1-002 and WY1-A2-003); WY1-A1-012 forwards it to its three neighboring nodes; WY1-A1-015 forwards it to its two neighboring nodes.

[0201] Second hop: Each relay node checks the packet sequence number 0x2F3A and source address 0x1A08. If it confirms that the packet is new, it continues to forward it, and the TTL is reduced to 6. At this time, there are 8 packet copies being transmitted simultaneously in the network.

[0202] Hops 3 to 6: Data packets are transmitted to the data aggregation node through different paths. Due to the redundancy of the network topology, there are a total of 5 different paths to the aggregation node, with transmission delays of 125ms for path 1 (4 hops), 178ms for path 2 (5 hops), 142ms for path 3 (4 hops), 203ms for path 4 (6 hops), and 189ms for path 5 (5 hops).

[0203] After receiving the first data packet (path 1, 125ms), the data aggregation node performs deduplication based on the source address and packet sequence number, discarding the subsequent four duplicate data packets. The aggregation node sends an acknowledgment message to the source node, which successfully receives the acknowledgment within the preset 500ms timeout period, completing the data transmission.

[0204] The S600 data aggregation node collected network connection status data from 247 valid nodes within a 30-minute time window. Statistical results show that the average number of neighbors per node is 3.2, the maximum number of hops is 12, and the network coverage area is approximately 150,000 square meters. The node distribution density is calculated as: 247 ÷ 150,000 ≈ 0.00165 nodes / square meter, which is 0.42 after normalization.

[0205] The average physical distance is calculated based on RSSI signal strength, using the free space propagation loss formula: The transmission power The mean value is 4 dBm, the mean RSSI is -52 dBm, and the calculated mean physical distance is 28.5 meters, which is 0.35 after normalization.

[0206] The environmental electromagnetic interference intensity assessment results show that the average noise floor in the A1 industrial plant area is -72dBm (normalized to 0.8), the B2 office building area is -81dBm (normalized to 0.45), and the C area distribution station area is -78dBm (normalized to 0.6).

[0207] After establishing the three-dimensional coordinate system, the coordinates of each region are as follows:

[0208] A1 Industrial Plant: (0.25, 0.32, 0.8), therefore the environmental electromagnetic interference intensity is 0.8 > 0.7, and it is determined to be an industrial plant scenario;

[0209] Office building B2: (0.89, 0.18, 0.45), therefore the density 0.89>0.8 and the distance 0.18<0.3, it is determined to be a smart building scenario;

[0210] Area C distribution station area: (0.42, 0.35, 0.6), therefore it does not meet the first two conditions and is judged as a low-voltage distribution station area scenario;

[0211] Based on the scene classification results, the system automatically sets tiered early warning thresholds:

[0212] A1 Industrial Plant: Level 1 warning 100mA, Level 2 alarm 300mA, area alarm condition is 5 or more abnormal nodes;

[0213] B2 office building: Level 1 warning 10mA, Level 2 alarm 50mA, area alarm condition is any node abnormality;

[0214] C area distribution station: Level 1 warning 30mA, Level 2 alarm 100mA, area alarm condition is 3 or more abnormal nodes.

[0215] On the 38th day of system operation, a typical early warning event occurred in the A1 industrial plant area: device WY1-A1-008 detected a leakage current of 43.8mA, which did not reach the 100mA level 1 early warning threshold, so no alarm was triggered; 5 minutes later, device WY1-A1-015 detected a leakage current of 156mA, triggering a level 1 early warning; subsequently, devices WY1-A1-020, WY1-A1-022, and WY1-A1-028 successively detected leakage currents exceeding 100mA, meeting the area alarm condition of "5 or more abnormal nodes", and the system issued an area alarm signal.

[0216] The entire early warning process, from initial anomaly detection to the issuance of a regional alarm, took 23 minutes, with a data transmission success rate of 99.2% and a false alarm rate below 0.8%. Subsequent troubleshooting confirmed that the area did indeed have a multi-point grounding fault caused by aging cable insulation, verifying the accuracy and effectiveness of the early warning system.

[0217] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0218] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A mesh-based online leakage current monitoring system, characterized in that, include: Multiple monitoring devices are installed near the power distribution line under test. Each monitoring device includes a processing and control module, a leakage current detection module, and a communication module. The leakage current detection module is used to collect leakage current data of the power distribution line under test. The communication module is used to transmit wireless data with other monitoring devices based on a wireless communication protocol. The processing and control module is used to evaluate the communication quality between nodes through the communication module and to control the communication module to adaptively expand the coverage of the Mesh network based on the communication quality in order to establish a multi-hop transmission path. The data aggregation node is the monitoring device with the highest overall signal quality score in the Mesh network. The processing and control module of the data aggregation node is also configured to collect network topology status data of the Mesh network. The communication module of the data aggregation node is also configured to receive monitoring data packets containing leakage current data transmitted via multi-hop transmission paths. The processing and control module of the data aggregation node is further configured to extract topology feature parameters based on the network topology status data and identify the current application scenario based on the topology feature parameters. Based on the identified application scenario, it adaptively sets tiered early warning thresholds and regional alarm conditions. Furthermore, it issues early warnings based on the tiered early warning thresholds and regional alarm conditions, as well as the received leakage current data. The topology feature parameters include node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity.

2. A method for online monitoring of leakage current based on mesh networking, characterized in that, Includes the following steps: Each monitoring device was installed near the power distribution line to be tested; Leakage current data of the power distribution line under test is collected in a non-contact manner by using an open-close zero-sequence current transformer in the monitoring equipment. Each monitoring device is used as a node in the Mesh network, and the nodes communicate with each other based on a wireless communication protocol. By evaluating the communication quality between nodes, the coverage of the Mesh network is adaptively expanded according to the communication quality to establish multi-hop transmission paths. During the construction of the Mesh network, the node with the highest overall signal quality score in the Mesh network is used as the data aggregation node, which collects the network topology status data of the Mesh network. Based on the multi-hop transmission path, the monitoring data packet containing leakage current data is transmitted to the data aggregation node in multiple hops; The data aggregation node extracts topology feature parameters based on network topology status data and identifies the current application scenario based on the topology feature parameters; among them, the topology feature parameters include node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity; The data aggregation node adaptively sets tiered early warning thresholds and regional alarm conditions based on the identified application scenarios; furthermore, the data aggregation node issues early warnings based on the tiered early warning thresholds, regional alarm conditions, and leakage current data.

3. The online leakage current monitoring method according to claim 2, characterized in that, Each monitoring device is installed near the power distribution line to be tested, including: Using a neodymium iron boron permanent magnet magnetic chassis, each monitoring device is attached to the surface of its corresponding distribution box.

4. The online leakage current monitoring method according to claim 3, characterized in that, Leakage current data of the power distribution line under test is collected in a non-contact manner using an on / off zero-sequence current transformer in the monitoring equipment, including: Open the separable iron core of the openable zero-sequence current transformer and clamp the three-phase conductors and neutral wire to be monitored; A separable iron core can be closed to form a complete magnetic circuit; When a ground leakage fault occurs, a voltage signal proportional to the leakage current is induced on the secondary side of the zero-sequence current transformer. The sensed voltage signal is converted and processed, and then encapsulated into a monitoring data packet by combining node identifier and timestamp.

5. The online leakage current monitoring method according to claim 4, characterized in that, By evaluating the communication quality between nodes, the coverage of the Mesh network is adaptively expanded based on the communication quality to establish multi-hop transmission paths, including: Each node collects communication quality parameters within its communication range; these parameters include RSSI signal strength index, LQI link quality index, and PER packet error rate index. The intensity of environmental electromagnetic interference is measured by spectrum scanning, and corresponding dynamic weights are set for each communication quality parameter based on the intensity of environmental electromagnetic interference. Based on communication quality parameters and corresponding dynamic weights, calculate the comprehensive signal quality score of the link with neighboring nodes; A list of neighboring nodes is established and maintained based on the comprehensive signal quality score; When the number of neighboring nodes is insufficient, the network coverage is expanded by gradually increasing the search radius, forming a multi-hop transmission path.

6. The online leakage current monitoring method according to claim 5, characterized in that, The node with the highest overall signal quality score in the mesh network is designated as the data aggregation node. This data aggregation node collects network topology status data of the mesh network, including: Each node periodically broadcasts status information containing its own comprehensive signal quality score; Each node receives and compares status information from other nodes, and jointly identifies the node with the highest overall signal quality score in the network as the data aggregation node. Data aggregation nodes collect and maintain topology status data, including the number of neighboring nodes, inter-node communication hops, and network coverage. The data aggregation node calculates node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity as topology characteristic parameters based on the collected topology status data.

7. The online leakage current monitoring method according to claim 6, characterized in that, Based on a multi-hop transmission path, monitoring data packets containing leakage current data are transmitted to the data aggregation node via multiple hops, including: The monitoring data packets are encapsulated into network data packets conforming to the Bluetooth Mesh protocol format, and the source node address, destination node address, packet sequence number and initial TTL hop value are added to the network data packets; A controlled flooding routing mechanism is adopted, in which the source node broadcasts network data packets to neighboring nodes in its list of neighboring nodes whose overall signal quality score is higher than a preset threshold. When a relay node receives a network data packet, it checks the packet sequence number and source node address of the network data packet. If the current network data packet has not been forwarded and the TTL value is greater than 0, the TTL value is decremented by 1 and the packet is forwarded to a neighboring node that meets the signal quality requirements. When the TTL value drops to 0 or the number of consecutive transmission failures reaches a preset number, forwarding stops and network topology reconstruction is triggered. The data aggregation node receives multiple network data packets originating from the same monitoring data packet through different paths. The data aggregation node performs deduplication on the multiple network data packets based on the source node address and packet sequence number, retains the network data packet that arrives first, and returns an acknowledgment message to the source node. If the source node does not receive an acknowledgment message within a preset time, it resends the monitoring data packet.

8. The online leakage current monitoring method according to claim 7, characterized in that, The data aggregation node extracts topology feature parameters based on network topology status data and identifies the current application scenario based on these parameters, including: The three parameters of node distribution density, average physical distance between nodes, and environmental electromagnetic interference intensity are normalized. The normalized parameters are then mapped to a three-dimensional coordinate system with node distribution density as the X-axis, average physical distance as the Y-axis, and environmental electromagnetic interference intensity as the Z-axis. Based on the preset scene judgment rules, the current scene is identified as one of the following: smart building, industrial plant, or low-voltage distribution substation.

9. The online leakage current monitoring method according to claim 8, characterized in that, The scenario determination rules are as follows: when the node distribution density is greater than 0.8 and the average physical distance is less than 0.3, it is determined to be a smart building scenario; when the node distribution density is less than 0.3 and the environmental electromagnetic interference intensity is greater than 0.7, it is determined to be an industrial plant scenario; when the conditions of node distribution density greater than 0.8 and average physical distance less than 0.3 are not met, and the conditions of node distribution density less than 0.3 and environmental electromagnetic interference intensity greater than 0.7 are also not met, it is determined to be a low-voltage distribution transformer area scenario.

10. The online leakage current monitoring method according to claim 9, characterized in that, The data aggregation node adaptively sets tiered early warning thresholds and regional alarm conditions based on the identified application scenarios; Furthermore, the data aggregation node issues warnings based on tiered warning thresholds, regional alarm conditions, and leakage current data, including: When the scene is identified as a smart building, the first-level warning threshold is set to 10mA and the second-level alarm threshold is set to 50mA. When any node detects that the leakage current exceeds the corresponding threshold, the area alarm is triggered. When the scene is identified as an industrial plant, the first-level warning threshold is set to 100mA and the second-level alarm threshold is set to 300mA. When 5 or more nodes detect leakage current exceeding the corresponding threshold, an area alarm is triggered. When the scenario is identified as a low-voltage distribution transformer area, the first-level warning threshold is set to 30mA, the second-level alarm threshold is set to 100mA, and the area alarm is triggered when three or more nodes detect leakage current exceeding the corresponding threshold.

Images