Edge computing gateway and sensing unit over-the-air wake-up method, system and storage medium

Abstract

This application relates to the technical field of power transmission and transformation monitoring, and discloses a method, system, and storage medium for over-the-air wake-up of an edge computing gateway and sensor units. The method is based on a sensor unit, a communication node, and an edge computing gateway connected wirelessly in sequence, including: establishing a first and second detection link in stages after receiving a network construction command and double-checking the information; setting a wake-up periodic mode, with the gateway polling and sending request packets to wake up the sensor unit, and the sensor unit going into sleep mode after completing data acquisition and uploading; it also includes steps such as link strength adaptation to power and data length, random time verification of the link, and verification to prevent false wake-ups after sleep mode. This application can avoid packet loss when multiple sensors are woken up simultaneously, ensure the accuracy of monitoring data, optimize equipment energy consumption, extend the battery life of sensor units, and improve the reliability and anti-interference capability of the system in harsh environments.

Description

Technical Field

[0001] This application relates to the technical field of power transmission and transformation monitoring, and in particular to an edge computing gateway and sensor unit over-the-air wake-up method, system and storage medium. Background Technology

[0002] As the core carrier of power grid operation, the operating status of power transmission and transformation equipment directly determines the safety, stability, and reliability of power supply. Conducting online monitoring of most power transmission and transformation equipment, collecting key parameters such as partial discharge and temperature in real time, and promptly identifying potential faults are essential means to ensure the continuous and stable operation of the power grid and reduce the risk of sudden accidents.

[0003] The working environment of power transmission and transformation equipment is relatively harsh, often facing severe conditions such as high temperature, high humidity, strong electromagnetic interference, and wind and sand erosion. Considering that sensors need to meet the requirement of continuous normal operation for five years, most partial discharge sensors choose battery power mode. This requires that the sensors be in an ultra-low power consumption mode when not in operation, in order to maximize battery life and avoid the maintenance burden caused by frequent battery replacements.

[0004] Currently deployed partial discharge sensors generally establish communication connections with edge computing gateways through timed active reporting, transmitting monitoring data to the gateway at preset fixed intervals. However, the gateway cannot flexibly adjust the communication timing according to actual monitoring needs, and cannot accurately control the entire process of sensor registration and activation, parameter sampling, data processing, information uploading, and sleep energy saving, affecting the accuracy of online monitoring data for power transmission and transformation. Summary of the Invention

[0005] To ensure the accuracy of online monitoring data for power transmission and transformation, this application provides a method, system, and storage medium for over-the-air wake-up of edge computing gateways and sensing units.

[0006] In a first aspect, this application provides a method for over-the-air wake-up of an edge computing gateway and a sensing unit, employing the following technical solution:

[0007] An over-the-air wake-up method for an edge computing gateway and a sensing unit, based on a sensing unit, a communication node, and an edge computing gateway connected wirelessly in sequence, includes the following steps:

[0008] Upon receiving the network construction command, the edge computing gateway obtains the corresponding node information, unit information, and detection information based on the network construction command.

[0009] The edge computing gateway establishes a first detection link with the corresponding communication node based on the node information. After the first detection link is established, the unit information and detection information are sent to the corresponding communication node for the first verification of the detection information. The communication node establishes a second detection link with the corresponding sensing unit based on the unit information and performs a second verification of the detection information.

[0010] If both verifications of the detection information pass, the wake-up cycle mode is set. After the wake-up cycle mode is set, the sensing unit is in low-power mode and checks whether a request data packet has been received.

[0011] The edge computing gateway sends a request data packet to the sensing unit through the communication node. After receiving the request data packet, the sensing unit is triggered to wake up and sequentially performs registration, sampling, data processing and uploading to generate sensing data and upload the sensing data to the edge computing gateway through the communication node.

[0012] After receiving the sensor data, the edge computing gateway parses it and sends response data to the sensing unit through the communication node. After receiving the response data, the sensing unit goes into sleep mode.

[0013] After sending response data, the edge computing gateway polls the next sensing unit.

[0014] By adopting the above technical solution, after processing one partial discharge sensor, the system begins polling the next partial discharge sensor, and so on. The edge computing gateway can apply its self-defined master-slave logic to wireless data transmission and reception, effectively solving the problem of packet loss caused by multiple sensors waking up simultaneously. This achieves controllability of device wake-up and data accuracy; saves battery costs and reduces unnecessary power consumption of partial discharge sensors; and effectively controls the wake-up and data transmission and reception mechanisms of multiple partial discharge sensors. Through master-slave control via over-the-air wake-up, the working efficiency of the online monitoring system is improved.

[0015] Optionally, the step of sending the unit information and detection information to the corresponding communication node and performing the first verification of the detection information further includes the following sub-steps:

[0016] The edge computing gateway sends unit information and detection information to the communication node, and the communication node receives the unit information and detection information.

[0017] The communication node searches for the established second detection link based on the unit information. If the second detection link has been established, it generates duplicate response data; otherwise, it waits for the second detection link to be established before generating unit response data.

[0018] If the communication node fails to establish a second detection link corresponding to the unit information after a set waiting period, it generates unit error data; otherwise, the communication node searches the stored information database based on the detection information.

[0019] If the detection information corresponds to the information database, detection response data is generated; otherwise, detection error data is generated.

[0020] If detection response data is generated, the communication node sends single-point data of the first length to the sensing unit through the second detection link corresponding to the unit information, and the communication node sends multi-point data of the second length to multiple adjacent communication nodes through temporary links; wherein, the first length is greater than the second length.

[0021] After receiving single-point data, the sensing unit returns single-point response data; after receiving multi-point data, the communication node returns multi-point response data.

[0022] The communication node packages the generated data into node response data, and then returns the node response data to the edge computing gateway;

[0023] The edge computing gateway receives and parses the node response data;

[0024] If unit error data is parsed, a unit warning is issued; if duplicate response data is parsed, a unit duplicate warning is issued; if detection error data is parsed, a first detection link warning is issued; if detection response data is parsed, the first detection link is successfully established; if single-point response data is parsed, the communication depth is normal; if multi-point response data is parsed, the communication extension is normal.

[0025] By adopting the above technical solution, two types of data are transmitted: continuous, large-volume single-point data and periodic, small-volume multi-point data. Combined with corresponding response mechanisms, this ensures the complete and continuous transmission of key monitoring data between the sensing unit and communication nodes, guaranteeing the accuracy and timeliness of core monitoring information. Furthermore, the streamlined multi-point data enables efficient collaboration and status synchronization between adjacent communication nodes, reducing redundant data transmission and lowering overall power consumption. Simultaneously, response feedback of different data types can verify communication depth and scalability, facilitating timely detection and early warning of link anomalies.

[0026] Optionally, if the communication depth is normal, the following steps are performed:

[0027] The communication node sends a single-point command to the sensing unit through the second detection link.

[0028] The sensing unit responds to a single-point command and returns to repeatedly collect data generated by the sensing target. The length of the collected data is the third length; where the first length > the third length > the second length.

[0029] The communication node receives the collected data; if the format of the collected data matches the preset data format, it returns the normal content of the sensing unit to the edge computing gateway.

[0030] By adopting the above technical solution, two types of data are transmitted: one is a long-cycle, short-volume wake-up command, and the other is a short-cycle, large-volume acquisition data. A hierarchical data transmission mechanism is formed by setting a first-length (longer) single-point data (wake-up related commands), a third-length (medium) acquisition data (multiple-cycle sensor target data), and a second-length (shorter) multi-point data: the longer single-point data ensures the complete and reliable transmission of wake-up commands, guaranteeing the accuracy of sensor unit activation; the medium-length, cyclically acquired, high-frequency data enables intensive monitoring of the sensor target, improving the comprehensiveness and accuracy of data sampling; and the shortest multi-point data maintains node collaboration, reducing redundant energy consumption.

[0031] Optionally, the method further includes the following steps:

[0032] Both the first detection link and the second detection link are wireless links;

[0033] The communication strength detected based on the first detection link is the first link strength, and the communication strength detected based on the second detection link is the second link strength.

[0034] A first relative value is calculated based on the first link strength and a preset first set strength, and a second relative value is calculated based on the second link strength and a preset second set strength.

[0035] The total relative value is calculated based on the first and second relative values.

[0036] Adjust the transmission power of communication nodes according to the negative correlation of the relative values ​​across the entire network;

[0037] Adjust the transmit power of the edge computing gateway based on the negative correlation of the first relative value;

[0038] The transmission power of the sensing unit is adjusted according to the negative correlation of the second relative value.

[0039] By adopting the above technical solution, on the one hand, power is reduced when link strength is sufficient to avoid energy waste; power is increased when link strength is insufficient to ensure data transmission stability, thus achieving a balance between communication quality and energy consumption. On the other hand, adjusting the power of corresponding devices for different links, especially targeted adjustments to battery-powered sensing units, can minimize unnecessary power consumption, effectively extend battery life, and reduce operation and maintenance costs. Simultaneously, adjusting the relative power of communication nodes across the entire path further optimizes the energy distribution of the entire communication link.

[0040] Optionally, the method further includes the following steps:

[0041] The second length is adjusted based on the negative correlation of the overall relative value. The larger the overall relative value, the smaller the second length; the smaller the overall relative value, the larger the second length.

[0042] The first or third length is adjusted according to the negative correlation of the second relative value. The larger the second relative value, the smaller the first or third length; the smaller the second relative value, the larger the first or third length.

[0043] By adopting the above technical solutions, when the link strength is sufficient, the corresponding data length is shortened to reduce redundant transmission and energy consumption; when the link strength is insufficient, the data length is increased to carry more complete verification information or core data, ensuring the effective transmission of key monitoring content.

[0044] Optionally, the method further includes the following steps:

[0045] The edge computing gateway generates first random content based on a first random time, sends the first random content to the communication node, and the communication node saves the first random content.

[0046] The edge computing gateway generates a first read instruction based on a second random time, and reads the first random content stored by the communication node to obtain the first read content.

[0047] Verify the matching between the first random content and the first read content. If the verification fails, issue a data anomaly warning.

[0048] By adopting the above technical solution, random verification of whether the first link remains normal during use can avoid the problem that fixed periodic verification is easily evaded by occasional interference of the link, and can more realistically and timely capture the hidden faults that may exist in the link during non-data transmission periods.

[0049] Optionally, the method further includes the following steps:

[0050] After the communication node saves the random content, it generates the second random content based on the third random time, and sends the second random content to the sensing unit, which then saves the second random content.

[0051] The edge computing gateway generates a second read instruction based on a fourth random time, reads the first random content through the communication node to obtain the second read content, and reads the second random content stored in the sensing unit through the communication node to obtain the third read content. It then verifies the matching of the second read content and the third read content. If the verification fails, it issues a communication link abnormality warning.

[0052] By adopting the above technical solution, the second link is randomly verified during use to ensure that it remains normal. Through the cross-link random verification method, the timeliness blind spots that may exist in the fixed verification cycle can be avoided, and the hidden anomalies of the second link during the sleep-wake interval or low-frequency communication can be accurately captured. At the same time, the collaborative communication capability of the first and second links can be verified simultaneously through the correlation verification of two levels of random content.

[0053] Optionally, after receiving the response data, the sensing unit enters a sleep state, which further includes the following sub-steps:

[0054] After the sensor unit goes into sleep mode, the edge computing gateway reads the collected data in the sensor unit through the communication node without sending a request data packet. If the collected data is successfully read, the sensor unit is marked as having failed to go into sleep mode; otherwise, the sensor unit is marked as having successfully gone into sleep mode.

[0055] After the edge computing gateway traverses multiple sensing units, it records the number of failed sleep cycles and the number of successful sleep cycles.

[0056] The percentage of hibernation failures is calculated based on the number of hibernation failures and the number of successful hibernations.

[0057] If the percentage of sleep failures exceeds a preset percentage, a warning will be issued to the sensor unit, and the transmit power of the edge computing gateway and the transmit power of the communication node will be adjusted according to the negative correlation of the sleep failure percentage.

[0058] By adopting the above technical solution, after the sensing unit goes into sleep mode, the edge computing gateway attempts to read its collected data without sending request data packets. This effectively detects whether the sensing unit is falsely awakened due to electromagnetic interference or other factors and is not truly in sleep mode. By negatively adjusting the transmission power of the gateway and communication node, the power is reduced to decrease electromagnetic interference to the sleep unit, forming a closed-loop anti-interference control mechanism.

[0059] Secondly, this application provides an edge computing gateway and sensing unit over-the-air wake-up system, which adopts the following technical solution:

[0060] An edge computing gateway and sensor unit over-the-air wake-up system includes a processor, wherein the processor performs the steps of the edge computing gateway and sensor unit over-the-air wake-up method as described in any of the preceding claims.

[0061] Thirdly, this application provides a storage medium, which adopts the following technical solution:

[0062] A storage medium storing a program, which, when executed by a processor, implements the steps of the over-the-air wake-up method for edge computing gateways and sensing units described in any of the preceding claims.

[0063] In summary, this application includes at least one of the following beneficial technical effects:

[0064] By establishing the first and second detection links in stages and double-checking the detection information, combined with a link verification mechanism based on random time, random content between the gateway and communication nodes, and between communication nodes and sensing units is randomly generated and verified. This can accurately identify matching anomalies in the early stage of link establishment and dynamically capture hidden faults in the link during operation. At the same time, the two-level random content association verification synchronously ensures the link coordination capability, effectively avoids the blind spots of fixed periodic verification, and ensures the continuous stability of the core communication link of the power transmission and transformation monitoring system.

[0065] For battery-powered sensing units and key communication equipment, a multi-level energy consumption control system is constructed. By dynamically adjusting the transmission power, adjusting the data length as needed, and precisely controlling the sleep state of the sensing unit, the battery life of the sensing unit is extended, the operation and maintenance costs are reduced, and the energy consumption of the entire system is balanced.

[0066] A master-slave logic polling wake-up mechanism is adopted to avoid packet loss caused by multiple sensors waking up simultaneously; hierarchical data transmission and data format verification are used to balance data sampling accuracy and transmission efficiency; at the same time, timely warnings are given when the link is abnormal to prevent data distortion or interruption caused by faulty links, and to ensure the accuracy of monitoring key parameters of power transmission and transformation in harsh environments. Attached Figure Description

[0067] Figure 1 This is a flowchart illustrating the steps of an over-the-air wake-up method for an edge computing gateway and sensing unit.

[0068] Figure 2 This is a diagram showing the sub-steps of sending unit information and detection information to the corresponding communication node and performing the first verification of the detection information.

[0069] Figure 3 This is a flowchart of the steps performed based on normal communication depth. Detailed Implementation

[0070] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0071] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the 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.

[0072] This application discloses an over-the-air wake-up method for an edge computing gateway and a sensing unit, referring to... Figure 1 Based on a sequentially wirelessly connected sensing unit, communication node, and edge computing gateway, the process includes the following steps:

[0073] The edge computing gateway receives the network construction command issued by the power transmission and transformation online monitoring system. Based on the command, it parses out the node information (such as node LoRa address and communication frequency band) corresponding to the communication node to be connected, the unit information (such as sensor device address ADDR and model parameters) corresponding to the sensing unit, and the detection information (such as partial discharge data sampling accuracy and monitoring cycle requirements).

[0074] The edge computing gateway establishes a first detection link (gateway-node link) with the target communication node via LoRa wireless signal based on node information. Once the link is established, it immediately sends the unit information and detection information to the communication node to initiate the first detection information verification. After receiving the information, the communication node establishes a second detection link (node-sensor link) with the corresponding sensing unit through its own LoRa module based on the sensor device address in the unit information. It then performs a second information verification with the sensing unit based on the parameter requirements in the detection information to ensure device compatibility and consistency of monitoring parameters, thus avoiding invalid subsequent data due to device mismatch.

[0075] If both detections pass, meaning the device address matches and the monitoring parameters are consistent, the edge computing gateway sends a wake-up cycle mode configuration command to the sensing unit through the communication node. In this mode, the link detection cycle is enabled first, and the heartbeat wake-up cycle is set to 3 seconds to adapt to the ultra-low power consumption requirements of the sensing unit's battery power supply. In low-power mode, the sensing unit only needs to periodically check whether it has received the heartbeat packet data sent by the gateway, which greatly reduces power consumption in non-working state, and simultaneously sends a customized over-the-air wake-up protocol. The data format is defined as: function code + device address ADDR + time parameter TIME + CRC check bit to ensure the anti-interference and integrity of data transmission. After the wake-up cycle mode is configured, the sensing unit automatically switches to ultra-low power sleep mode, only maintaining the 3-second heartbeat packet detection function, waiting for the request data packet to trigger wake-up.

[0076] When the edge computing gateway needs to acquire partial discharge data, it sends a request data packet conforming to the over-the-air wake-up protocol to the target sensing unit through the communication node. The request data includes the function code, the sensor's ADDR, and other information. After receiving the request data packet within a 3-second heartbeat detection cycle, the sensing unit immediately triggers wake-up and exits low-power mode, performing the following operations: First, it completes registration, changing its working state from sleep to wake-up / activated and reporting the registration status to the communication node. Then, it samples the partial discharge signal according to the sampling accuracy requirements in the detection information, such as continuously collecting multiple partial discharge parameters. The sampled data undergoes noise reduction, filtering, and other data processing to generate standardized sensing data, i.e., partial discharge data. Subsequently, the sensing data is uploaded to the edge computing gateway via the communication node through the second detection link. The entire process relies on LoRa wireless transmission, adapting to the long-distance and anti-interference requirements of power transmission and transformation sites.

[0077] After receiving the sensor data, the edge computing gateway unpacks the data according to the format of the over-the-air wake-up protocol, parses out key information such as local discharge parameters and device status, and after confirming that the data is complete and valid, sends a wireless response packet containing data reception confirmation information to the sensor unit through the communication node. After receiving the response packet, the sensor unit immediately terminates the current workflow, automatically resets and re-enters the ultra-low power sleep mode, resumes the 3-second heartbeat packet detection, and waits for the next wake-up request.

[0078] After the edge computing gateway sends response data to the current sensing unit, it immediately switches to the next sensing unit without waiting for an additional cycle. Through the same process described above, that is, sending request data packets through the corresponding communication node to start the next round of wake-up and data interaction; and so on, the edge computing gateway realizes the polling control of multiple partial discharge sensors through a self-defined master-slave logic. In this case, the gateway is the master control terminal and all sensing units are slave control terminals, avoiding the LoRa channel congestion and data packet loss problems caused by multiple sensors being woken up at the same time.

[0079] By adopting the above technical solutions, on the one hand, the sensing unit only wakes up when it receives a request data packet, and maintains ultra-low power detection with a 3-second cycle for the rest of the time, which greatly reduces unnecessary power consumption, extends battery life, and reduces battery replacement frequency and maintenance costs; on the other hand, relying on the polling control of master-slave logic and the customized over-the-air wake-up protocol, the packet loss problem of multiple sensors waking up at the same time is effectively solved, realizing the controllability of device wake-up and the accuracy of data transmission; at the same time, through hierarchical link establishment and bidirectional verification, CRC check and other designs, the system's anti-interference capability is improved.

[0080] Reference Figure 2 The step of sending unit information and detection information to the corresponding communication node and performing the first verification of the detection information also includes the following sub-steps:

[0081] The edge computing gateway, based on the established LoRa first detection link, packages and sends the parsed unit information and detection information to the target communication node. The unit information includes the sensor device address (ADDR) and model adaptation parameters; the detection information includes partial discharge sampling accuracy, data transmission baud rate, and CRC check rules. After receiving the data packet through its own LoRa module, the communication node first performs basic frame format verification to determine if it conforms to the frame structure of the over-the-air wake-up protocol. If the verification passes, information reception is complete, preparing for subsequent checks and link verification.

[0082] After receiving the information, the communication node first searches the locally stored device-link mapping table based on the sensor device address (ADDR) in the unit information to check whether a second detection link has already been established for the sensing unit, thus avoiding the waste of channel resources caused by duplicate link establishment.

[0083] If the second detection link corresponding to the sensing unit already exists, duplicate response data is immediately generated, including duplicate link establishment identifier, established link parameters, and marking the current link establishment request as a duplicate operation.

[0084] If no established link is found, the second detection link establishment process is initiated. A link establishment request is sent to the target sensing unit via LoRa signal, carrying the model adaptation parameters in the unit information, and the link establishment result is awaited. After the second detection link is successfully established, unit response data is generated, including the link establishment success identifier and the link communication frequency band.

[0085] If no link establishment response is received from the sensor unit after the set waiting time, meaning a second detection link corresponding to the unit information has not been established, then unit error data is generated, including error types such as "sensor unit not responding" and "signal attenuation exceeding limits," thus clarifying the reason for the link establishment failure. The set time is typically 5-10 seconds, depending on the signal attenuation conditions at the power transmission and transformation site, to accommodate long-distance communication delays.

[0086] Detection information matching and verification with the database: If the communication node successfully establishes the second detection link and no unit error data is generated, the local pre-stored detection information database is further retrieved. This database stores the standard detection parameters corresponding to each type of sensor unit preset by the power transmission and transformation monitoring system, such as the partial discharge sampling frequency range, data format specifications, and transmission timeout threshold. The received detection information is then matched item by item with the standard parameters corresponding to the sensor unit model in the database.

[0087] If the detection information is completely consistent with the standard parameters in the database, detection response data will be generated, including a parameter matching success identifier and standard parameter reference values.

[0088] If there are deviations in the detection information, such as sampling accuracy exceeding the standard range or CRC check rules not matching, detection error data will be generated, including error parameter items and standard parameter values.

[0089] Single-point / multi-point data transmission and bidirectional response: If the communication node generates detection response data and confirms the detection information match, then dual-path data transmission verification is initiated to further test the link communication quality and node coordination capabilities.

[0090] Single-point data transmission (node-sensor link verification): The communication node sends a first-length single-point data, typically 50-100 bytes long, to the target sensing unit through the established second detection link, i.e., LoRa directional transmission. This data contains detailed sensor operating parameter configuration instructions, such as partial discharge sampling period, sleep / wake-up threshold, and data encryption key. Core control information must be transmitted completely. After receiving the single-point data, the sensing unit parses and verifies the integrity of the configuration instructions. If there are no errors, it returns single-point response data, including data reception confirmation and the sensor's current status: ready for standby. This response data directly proves that there are no communication failures in the first detection link (gateway-node) and the second detection link (node-sensor), ensuring the smooth operation of the basic link for subsequent data interaction.

[0091] Multi-point data transmission (node-to-node collaborative verification): Communication nodes simultaneously send multi-point data of a second length to multiple adjacent communication nodes (usually 3-5, covering the monitoring area around the target sensing unit) via temporary links and LoRa dynamic networking links, which are temporarily established according to the distribution of nodes on site and do not need to be maintained for a long time. The multi-point data is usually 10-20 bytes long and contains simplified node status information, such as current link strength, remaining power, and whether it is idle. Only the core collaborative parameters are transmitted. After receiving the multi-point data, the adjacent communication nodes quickly verify the connectivity of the temporary link and return multi-point response data, which includes temporary link confirmation and their own idle status. This response data can prove that the communication node has temporary networking and collaborative capabilities, providing support for subsequent multi-node redundancy backup and data relay transmission, and adapting to scenarios with weak signals in some areas of power transmission and transformation sites.

[0092] Response data packaging and back transmission and gateway parsing and early warning: The communication node packages all the data generated in the above process according to the over-the-air wake-up protocol format to generate node response data, including data type identifier, timestamp, and CRC check bit, and transmits it back to the edge computing gateway through the first detection link; all data includes unit response data / repeated response data / unit error data, detection response data / detection error data, single-point response data, and multi-point response data.

[0093] After receiving the node's response data, the edge computing gateway unpacks the data according to preset parsing rules and triggers corresponding early warning prompts or status feedback based on the data type, adapting to the operation and maintenance needs of power transmission and transformation sites.

[0094] If the unit erroneous data is parsed, a unit warning prompt will pop up on the monitoring system interface, such as "Sensor unit ADDR: XXX is not responding, please check the power supply or location of the device", to guide the maintenance personnel to quickly locate the sensor fault;

[0095] If duplicate response data is parsed, a duplicate notification is sent to the unit, such as "Sensor unit ADDR: XXX link already established, no need for repeated operation", to avoid the gateway repeatedly issuing link establishment commands, which would cause channel congestion.

[0096] If incorrect detection data is parsed, a warning message will be triggered in the first detection link, such as "Detection parameters do not match: the sampling accuracy should be ±0.1mV, but it is currently ±0.5mV", reminding the operation and maintenance personnel to correct the detection information configuration.

[0097] If the detection response data is parsed, it indicates that the first detection link has been successfully established, such as "Gateway-Node Link is connected, Detection Parameters Matched", confirming that the core link infrastructure is normal.

[0098] If single-point response data is parsed, it indicates that the communication depth is normal, such as "the node-sensor link is smooth and the parameter configuration is correct", which proves that the data can directly reach the sensing unit and meet the depth monitoring requirements.

[0099] If multiple response data are parsed, it indicates that the communication expansion is normal, such as "adjacent nodes XXX and XXX can temporarily coordinate and support redundant transmission", confirming that the nodes have the ability to expand the network and cope with local link failures.

[0100] By adopting the above technical solution, on the one hand, the first length of single-point data (continuous and large volume) can completely transmit sensor configuration instructions and core monitoring parameters, ensuring that key information between the sensing unit and the communication node is not lost, and guaranteeing the accuracy and timeliness of subsequent partial discharge data sampling and uploading. For example, the detailed sampling period configuration directly affects the density of monitoring data. On the other hand, the second length of multi-point data (periodic small volume) only transmits simplified node status information. While achieving efficient collaboration between adjacent communication nodes (such as temporary relay transmission of data in weak signal areas) and status synchronization (avoiding multiple nodes repeatedly establishing links), it significantly reduces the amount of redundant data transmission, lowers the transmit power consumption of LoRa modules in each node, and adapts to the battery-powered or low-power design requirements of nodes. At the same time, the response feedback corresponding to different data types can accurately divide the verification dimensions of communication depth and communication scalability. Once a link anomaly occurs, the problem link can be quickly located through targeted early warning.

[0101] Reference Figure 3 If the communication depth is normal, then perform the following steps:

[0102] Based on the normal communication depth status feedback, the communication node confirms that the second detection link is in a stable connection state, and then sends a single-point command to the target sensing unit through the LoRa directional link. This command is a customized control command that conforms to the over-the-air wake-up protocol. Its length is a simplified version adapted to the first length, about 30-40 bytes. It includes the command type (such as partial discharge data acquisition command), acquisition parameters (such as the number of cyclic acquisitions of 3-5 times, the duration of each sampling of 200-500ms, and the sampling frequency of 1MHz, adapted to the high-frequency characteristics of partial discharge signals), data upload time limit (such as the data to be returned within 10 seconds after acquisition to avoid data backlog), and command check code (generated based on CRC rules to prevent the command from being tampered with by electromagnetic interference). During the command issuance process, the communication node maintains a stable transmission power of the LoRa signal and refers to the previous link strength adjustment results to ensure that the command accurately reaches the sensing unit and avoids command loss due to strong electromagnetic interference at the power transmission and transformation site.

[0103] After receiving a single-point command, the sensing unit first verifies the command, checking the checksum and the reasonableness of the acquisition parameters. Once confirmed, it immediately switches from low-power standby mode to acquisition mode and initiates a cyclic acquisition process according to the command requirements. For partial discharge monitoring targets in power transmission and transformation equipment, such as transformer bushings and GIS equipment connectors, the built-in partial discharge sensor probe continuously acquires partial discharge signals 3-5 times. Multiple cyclic acquisitions avoid the influence of transient interference on single sampling, improving data reliability. During each acquisition process, the sensing unit performs preliminary filtering on the raw signal (removing 50Hz power interference). The sensor unit performs analog-to-digital conversion (converting analog signals into 16-bit digital signals) and packages the raw data, acquisition timestamp, and signal strength value of each acquisition into a single acquisition segment. After the cyclic acquisition is completed, the sensor unit integrates all single acquisition segments into complete acquisition data. The acquisition data length is the third length, about 20-30 bytes, between the first length (50-100 bytes) and the second length (10-20 bytes). At the same time, a data integrity identifier, such as the acquisition count and the total data volume check value, is added to the end of the acquisition data to ensure that subsequent verification can determine whether the data is complete.

[0104] After the sensing unit generates the acquired data, it transmits it back to the communication node via the second detection link using LoRa directional transmission. During the transmission, a short-frame fast transmission strategy is employed to accommodate medium-sized data volumes of the third frame, reducing signal transmission time in the air and lowering the probability of interference. Simultaneously, a data type identifier, such as partial discharge (PD) acquisition data, is added to the data frame header to facilitate rapid data identification by the communication node. Upon receiving the acquired data, the communication node immediately initiates a data format verification process: retrieving the pre-stored PD data standard format, including the frame structure: a 2-byte frame header + a 1-byte number of acquired segments + a single... Each data segment consists of 5-6 bytes × N + a 4-byte timestamp + a 2-byte checksum. The frame structure, field length, data range, and integrity flag of the received data are checked item by item. For example, the amplitude of a partial discharge signal should be within 0-500mV; exceeding this range is considered abnormal. If the format matches perfectly, it proves that the acquired data has not been tampered with and is complete, and the communication node generates a status flag indicating that the data verification has passed. If there are format deviations, such as missing fields or excessive amplitude, the data is marked as abnormal, and the abnormality type is recorded, such as excessive amplitude or frame structure error, providing a basis for subsequent troubleshooting.

[0105] After the communication node completes the data format verification, if the verification passes, it immediately packages the normal feedback information from the sensing unit into a status data packet. The feedback information includes the sensing unit's ADDR, the acquisition completion time, and the data verification result. The status data packet is approximately 15 bytes long, adapted to the second-length simplified logic. It is then transmitted back to the edge computing gateway through the first detection link. This feedback information allows the gateway to monitor the working status and data acquisition progress of the sensing unit in real time, providing a basis for polling other sensors or initiating the data parsing process. If the verification fails, the communication node adds a data anomaly identifier and anomaly type to the feedback information, reminding the gateway to reissue the acquisition command or troubleshoot the sensing unit for faults such as probe contamination or signal interference sources.

[0106] By adopting the above technical solutions, on the one hand, single-point commands with long cycles (e.g., once every 5-10 minutes) and small data volume (30-40 bytes) can minimize the energy consumption of command transmission while ensuring the accuracy of wake-up and control, adapting to the power supply requirements of sensor batteries, and avoiding the power waste caused by frequent sending of long commands. On the other hand, cyclic data acquisition with short cycles (1-3 seconds for a single acquisition) and medium data volume (20-30 bytes) can capture detailed features of partial discharge signals, such as signal amplitude changes and pulse frequency, through multiple sampling and format verification, thereby improving the comprehensiveness and accuracy of monitoring data and meeting the data accuracy requirements of power transmission and transformation equipment fault early warning, while avoiding transmission delay and increased energy consumption caused by excessive data volume. At the same time, combined with the node collaboration logic of the shortest multi-point data in the early stage, a three-layer transmission system is formed, consisting of long-cycle command control wake-up, short-cycle data strong acquisition, and simplified data to ensure collaboration. Under the premise of ensuring the quality of monitoring data, a balance between energy consumption and efficiency is achieved, effectively coping with the impact of harsh conditions such as high temperature and strong electromagnetic interference at the power transmission and transformation site on data transmission.

[0107] To further adapt to the strong electromagnetic interference and large signal attenuation fluctuations in power transmission and transformation sites, and to ensure stable communication between the first detection link (gateway-node) and the second detection link (node-sensor) under low power consumption, this method also includes a step of dynamically adjusting the transmission power of each device based on link strength. Specifically, the method includes the following steps:

[0108] After the first and second detection links are established, during data transmission, each device initiates a periodic link strength detection mechanism. The detection period is synchronized with the heartbeat wake-up period and is set to 3 seconds to avoid additional power consumption.

[0109] The edge computing gateway sends a strength detection frame to the communication node through the first detection link. It conforms to the over-the-air wake-up protocol, is about 8 bytes long, and contains a detection identifier and a timestamp. After receiving the frame, the communication node returns a strength response frame, which carries the received signal strength value, i.e., the first link strength, in dBm. The gateway takes the average value of multiple reception results to eliminate the impact of instantaneous interference, such as signal fluctuations caused by the start-up and shutdown of power transmission and transformation equipment.

[0110] The communication node sends an intensity detection frame to the sensing unit through the second detection link. After receiving the frame, the sensing unit returns an intensity response frame, which carries the received signal strength value, i.e., the strength of the second link, in dBm. The communication node also takes the average value through multiple samples to ensure the accuracy of the intensity data.

[0111] Considering the signal attenuation characteristics at power transmission and transformation sites, such as local signal weakness areas caused by GIS equipment shielding, the first link strength usually needs to be maintained between -80dBm and -40dBm, and the second link strength needs to be maintained between -85dBm and -45dBm. The sensor unit antenna size is smaller, and the signal coverage range is slightly narrower.

[0112] Link relative value and overall relative value calculation: After each device obtains link strength data, it initiates the relative value calculation process to quantify the link communication quality.

[0113] First relative value calculation: The first set strength is preset and determined according to the designed coverage distance of the first detection link. For example, when the maximum distance between the gateway and the node is 500 meters, the first set strength is set to -60dBm. The first relative value is calculated using the formula: First relative value = (First link strength - First set strength) / First set strength × 100%. If the first link strength is -50dBm, which is better than the set value, then the first relative value is 16.7%, which means the link strength is sufficient. If it is -70dBm, which is weaker than the set value, then the first relative value is -16.7%, which means the link strength is insufficient.

[0114] Second relative value calculation: The second set intensity is preset and determined according to the coverage distance of the second detection link. For example, when the maximum distance between the node and the sensor is 100 meters, the second set intensity is set to -65dBm. The second relative value is calculated through the same logic. For example, when the second link intensity is -60dBm, the second relative value is 7.7%; when it is -75dBm, it is -15.4%.

[0115] Overall relative value calculation: Considering that the communication node is the core relay device of the dual link, the weighted average method is used to calculate the overall relative value. The formula is: Overall relative value = (first relative value × 0.4 + second relative value × 0.6); the battery-powered sensing unit associated with the second link has a higher weight and its energy consumption optimization is prioritized. For example, when the first relative value is 16.7% and the second relative value is 7.7%, the overall relative value is 11.3%.

[0116] Device transmit power negative correlation adjustment: Based on the calculated relative value, each device initiates dynamic adjustment of transmit power, following the negative correlation logic that the more sufficient the intensity, the lower the power; and the less sufficient the intensity, the more the power is appropriately increased. Moreover, the power adjustment range is strictly adapted to the device hardware capabilities and low power consumption requirements.

[0117] Communication node power adjustment: Based on the relative value of the entire link, the initial transmit power of the communication node is set to 17dBm, which is the commonly used power of the LoRa module, covering a range of 500 meters. If the relative value of the entire link is >10%, the link is generally smooth, so the power is reduced by 1dBm per cycle, down to a minimum of 13dBm, covering 300 meters, meeting the node spacing requirements in the power transmission and transformation field and reducing ineffective energy consumption. If the relative value of the entire link is <-10%, the link is generally weak, so the power is increased by 1dBm per cycle, up to a maximum of 20dBm, covering 800 meters, enhancing signal penetration and ensuring relay function.

[0118] Edge computing gateway power adjustment: The power is adjusted according to the first relative value. The initial transmit power of the gateway is set to 18dBm. If the first relative value is >15% and the gateway-node link is smooth, the power is reduced to 15dBm, covering 600 meters. If the first relative value is <-15% and the link is weak, the power is increased to 21dBm, covering 1000 meters. The gateway is powered by AC mains. The power adjustment focuses more on communication stability and takes into account moderate energy saving.

[0119] Sensor unit power adjustment: The initial transmit power of the sensor unit is set to 14dBm based on the second relative value, which is suitable for battery power supply and covers 150 meters. If the second relative value is >8%, the node-sensor link is smooth, so it is reduced to 11dBm, covering 80 meters, minimizing power consumption. For every 1dBm reduction, power consumption is reduced by about 10%, significantly extending battery life. If the second relative value is <-8%, the link is weak, so it is increased to 16dBm, covering 200 meters. However, after each increase, it is maintained for 3 detection cycles (9 seconds) to avoid sudden power consumption due to prolonged high power.

[0120] By adopting the above technical solution, on the one hand, lightweight strength detection with a 3-second cycle captures link fluctuations in real time, such as sudden drops in strength caused by electromagnetic interference, avoiding the problem of high energy consumption in strong links or communication interruption in weak links under fixed power, and accurately balancing communication quality and energy consumption; on the other hand, differentiated adjustment is made for the power supply characteristics of different devices, with sensor units prioritizing energy saving, gateways focusing on stability, and nodes taking into account both relay and energy saving. In particular, for battery-powered sensor units, a narrow range of power adjustment of 11-16dBm is used to meet communication requirements while minimizing ineffective power consumption. Actual tests show that it can extend battery life compared to a fixed power solution; at the same time, the weighted adjustment of the relative value of the entire path for communication nodes ensures a balanced distribution of energy consumption between the two links, avoiding overall communication interruption caused by excessive energy saving in a single link.

[0121] To further optimize the data transmission efficiency of power transmission and transformation wireless communication links, this method, in addition to dynamically adjusting the transmit power, also includes a step of adapting and adjusting the data length according to the relative value of the link. By matching the negative correlation between link strength and data length, the method achieves the dual goals of reducing redundant transmission and ensuring critical data protection. The specific method also includes the following steps:

[0122] Second length adjustment based on the overall relative value: The second length corresponds to multi-point data transmitted between adjacent communication nodes, such as node status synchronization information and temporary coordination instructions. Its core requirement is lightweight coordination, so it is dynamically adjusted according to the overall relative value of the entire link, which reflects the overall quality of the dual links.

[0123] Length baseline setting: The default second length baseline value is 10 bytes, including 2 bytes of node address, 1 byte of status identifier, 1 byte of remaining power, 4 bytes of timestamp, and 2 bytes of simple check code, to adapt to the basic collaboration requirements between nodes.

[0124] Strong link scenario (overall relative value > 10%): When the overall relative value is > 10%, such as 11.3%, it means that the dual links are generally smooth, with little signal attenuation and little interference. The data transmission error rate is low, and there is no need to carry extra redundant information. At this time, the second length is shortened by 2-3 bytes from the baseline value. For example, the millisecond field in the timestamp is removed (retained to the second level), and the check code is simplified to 1 byte, so that the second length is reduced to 7-8 bytes. The shortened data can still completely transmit the core information of the node status, while reducing the transmission time of each frame of data. When the LoRa transmission rate is 9600bps, each reduction of 1 byte can shorten the transmission time by about 0.8ms, reducing the node's transmission power consumption and channel occupancy.

[0125] Weak link scenario (overall relative value < -10%): When the overall relative value is < -10%, such as -12%, it indicates that the dual links are weak and susceptible to electromagnetic interference, leading to data loss. It is determined that the data anti-interference capability needs to be enhanced. At this time, the second length is increased by 3-4 bytes from the baseline value. For example, add node signal strength value (1 byte), add double check code (2 bytes, original check code is retained + new CRC-16 check), and add retransmission flag (1 byte), so that the second length is increased to 13-14 bytes. The added check information can significantly reduce the bit error rate. Double check reduces the bit error rate by about 80% compared with single check. The retransmission flag makes it easy for the receiver to identify whether a retransmission is needed, ensuring that the coordination instructions between nodes are not lost and avoiding network disconnection caused by weak links.

[0126] The first / third length adjustment is based on the second relative value: the first length corresponds to the single-point data transmitted from the communication node to the sensing unit, such as wake-up commands and sampling configurations; the third length corresponds to the collected data returned by the sensing unit, such as partial discharge sampling segments. Both are directly related to the interaction between the sensing unit and the data core, and are therefore adjusted according to the second relative value that reflects the quality of the node-sensor link.

[0127] Length baseline setting: The preset first length baseline value is 40 bytes, including 2 bytes of wake-up command code, 4 bytes of sensor ADDR, 6 bytes of sampling parameters, 8 bytes of encryption key, 8 bytes of timestamp, and 12 bytes of complete check code, to ensure command integrity; the preset third length baseline value is 25 bytes, including 1 byte of sampling count, 5 bytes × 4 times of single sampling data, 4 bytes of timestamp, and 3 bytes of check code, to adapt to the data integration requirements of 4-cycle sampling.

[0128] Strong link scenario (second relative value > 8%): When the second relative value > 8%, such as 9%, it indicates that the node-sensor link is smooth, the sensor unit receives signals stably, and the reliability of judgment commands and data acquisition is high. The length can be appropriately shortened to reduce redundancy.

[0129] First length adjustment: Remove redundant fields in the encryption key and simplify sampling parameters; reduce redundant fields from 8 bytes to 5 bytes, retaining the core encryption factor; simplify sampling parameters such as the sampling duration from 2 bytes integer + 1 byte decimal to 2 bytes integer, so that the first length is reduced to 32-33 bytes.

[0130] The third length adjustment reduces the precision field of a single sampled data (from 5 bytes of 16-bit precision to 4 bytes of 12-bit precision, still meeting the accuracy requirement of ±0.1mV for partial discharge monitoring) and merges the number of acquisitions and the retransmission identifier (shared by 1 byte), reducing the third length to 20-21 bytes.

[0131] The shortened first / third length can still ensure the validity of the command and the accuracy of the collected data, while reducing the time consumed by the sensing unit to receive and process data, and indirectly reducing its power consumption in the wake-up state;

[0132] Weak link scenario (second relative value < -8%): When the second relative value is < -8%, such as -9%, it indicates that the node-sensor link is weak and prone to data segment loss. It is necessary to increase the length to carry complete verification and core data backup.

[0133] First length adjustment: Add sampling parameter backup (repeatedly store key parameters such as sampling frequency, 2 bytes), extend the check code to 16 bytes (including independent check values ​​of each field of the instruction), and add instruction priority identifier (1 byte), so that the first length increases to 43-44 bytes; the backup parameters can avoid sensor unit configuration errors due to the loss of instruction fragments, and the independent check code facilitates quick location of lost fields.

[0134] The third length adjustment includes: adding a backup of the collected data (repeating the last sampled data, 5 bytes), supplementing the sampling environment parameters (such as sensor temperature, 1 byte, used for subsequent data anomaly detection), and extending the checksum to 5 bytes (including overall CRC check + checksum of each sampling segment), thus increasing the third length to 31-32 bytes; the backup sampled data can be used as a supplement when the main data is lost, and the environmental parameters provide a basis for data validity analysis, avoiding the distortion of monitoring data caused by weak links.

[0135] Synergistic Linkage of Length Adjustment and Power Adjustment: Data length adjustment forms a synergistic mechanism with the previously mentioned transmit power adjustment. When the link is weak, such as when the overall relative value is <-10% and the second relative value is <-8%, the transmit power is increased (to enhance signal strength) while the data length is increased to supplement verification information, providing double protection for data transmission. When the link is strong, such as when the overall relative value is >10% and the second relative value is >8%, the transmit power is reduced (to reduce energy consumption) while the data length is shortened to reduce channel occupancy and maximize efficiency. The linkage between the two is uniformly managed by the edge computing gateway. The gateway obtains the overall relative value and the second relative value in real time through the communication node and dynamically issues length adjustment commands to ensure that the data length of each device is synchronously adapted.

[0136] By adopting the above technical solutions, the short-length design for strong links can reduce redundant data transmission and lower the energy consumption of each device and the risk of channel congestion; the long-length design for weak links, through additional verification and data backup, improves the effective reception rate of key data such as sampling configuration and partial discharge data, and avoids monitoring interruption caused by link interference; at the same time, the coordination of data length adjustment and transmission power adjustment significantly enhances the adaptability of the power transmission and transformation online monitoring system in complex communication environments.

[0137] To avoid sporadic electromagnetic interference at power transmission and transformation sites, such as pulse interference generated during equipment start-up and shutdown, and to bypass fixed-period verification, this method also includes a link verification step based on random time to promptly detect latent faults in the first detection link during non-data transmission periods, such as increased signal attenuation or abnormal node storage. Specifically, the method includes the following steps:

[0138] Random content generation and distribution storage: After completing routine data interactions, such as receiving sensor data and issuing wake-up commands, the edge computing gateway initiates the first random time generation mechanism. The first random time adopts a base period and random offset strategy. The base period is set to 3-5 minutes to adapt to the regular data transmission interval of power transmission and transformation monitoring, avoiding frequent verification and increased energy consumption. The random offset range is ±60 seconds. For example, when the base period is 3 minutes, the first random time may be 2 minutes and 10 seconds or 3 minutes and 50 seconds, ensuring that the verification time is irregular and avoiding fixed periods of occasional interference.

[0139] When the first random time is reached, the gateway generates 16 bytes of first random content using an encryption algorithm, such as AES-128. This content contains random characters, timestamp fragments, and a gateway identifier. This content has no actual business significance but is unique and unpredictable. The gateway then encapsulates the first random content into a verification data packet that conforms to the over-the-air wake-up protocol. A random verification identifier of approximately 20 bytes is added to the frame header, and the packet is sent to the target communication node through the first detection link.

[0140] After receiving the verification data packet, the communication node first verifies the frame format and the legality of the identifier. If it is correct, it stores the first random content in the local non-volatile storage area, such as EEPROM, to avoid loss when the node is powered off. At the same time, it records the reception time and storage address, generates implicit feedback of successful storage, and does not immediately send it back, thereby reducing the energy consumption and channel occupation of the verification process.

[0141] Random read command issuance and content retrieval: After the edge computing gateway issues the first random content, it starts the second random time generation mechanism; the second random time is uncorrelated with the first random time, and the basic period is set to 1-2 minutes to ensure that the verification is completed in a short time after the first random content is stored, so as to avoid the node from overwriting the data due to subsequent operations. The random offset range is ±30 seconds. For example, if the first random time is 2 minutes and 10 seconds, the second random time may be 3 minutes and 25 seconds or 4 minutes and 05 seconds.

[0142] When the second random time is reached, the gateway generates a first read instruction, which includes the target communication node address, the storage address identifier of the first random content, and a read timeout threshold of 5 seconds. This instruction is also encapsulated as a verification instruction packet and sent through the first detection link. After receiving the first read instruction, the communication node retrieves the previously saved first random content from the EEPROM according to the storage address identifier, generates a read response packet, which includes the first random content to be read and the storage duration, and sends it back to the gateway through the first detection link.

[0143] If a communication node does not receive the first read instruction within the read timeout threshold, such as due to a momentary link interruption, it will proactively report the status of not receiving the read instruction during subsequent regular data interactions; if a node fails to retrieve stored content, such as due to storage area corruption, it will return an error indicating that the content is lost.

[0144] Content matching verification and anomaly warning handling: After receiving the read response packet, the edge computing gateway initiates the matching verification process; it compares the first random content stored locally with the first read content returned byte by byte, and simultaneously verifies whether the storage time is within a reasonable range, such as not exceeding 3 minutes, to prevent content from being tampered with or overwritten.

[0145] If both are completely consistent and the storage duration is normal, the first detection link is determined to be in a stable state with no hidden faults. The gateway only records the verification results, does not trigger any additional operations, and continues to maintain normal data interaction.

[0146] If verification fails, such as due to content byte mismatch, storage duration exceeding the limit, or receiving a content loss flag, the anomaly handling mechanism is immediately activated: First, a data anomaly warning pops up on the monitoring system interface, including the fault node address, verification time, and fault type, such as content mismatch or storage anomaly; second, three retests are automatically triggered, with the random time offset range for each retest expanded to ±120 seconds to eliminate instantaneous interference. If all three retests fail, a persistent fault is determined in the first detection link, further triggering link repair instructions, such as adjusting the gateway's transmit power or restarting the node's storage area; simultaneously, the fault information is uploaded to the power transmission and transformation monitoring cloud platform, allowing maintenance personnel to remotely view fault details and formulate on-site troubleshooting plans, such as checking whether the node antenna is offset or whether any new obstructions have been added between the gateway and the node.

[0147] By adopting the above technical solution, on the one hand, the random time generation mechanism breaks the predictability of fixed-period verification and avoids occasional interference, such as equipment pulse interference every 5 minutes, which happens to be out of sync with the verification time, making it impossible to avoid monitoring for hidden faults, such as intermittent attenuation of link signals, and the verification results are closer to the actual operating state of the link; on the other hand, by storing and reading meaningless random content, it only occupies a very small amount of channel resources, and the amount of data for a single verification is about 40 bytes, which is far less than conventional sensor data, and does not affect normal data transmission, thus controlling energy consumption while ensuring monitoring accuracy.

[0148] To address the issue that the second detection link (communication node-sensor unit) in power transmission and transformation monitoring often operates in an intermittent sleep-wake state, and that low-frequency communication can easily conceal hidden faults (such as decreased signal reception sensitivity when the sensor is in sleep mode or temporary link interruption at the node), this method uses a cross-link random verification mechanism to simultaneously verify the stability of the second link and the collaborative capability of the two links. The specific method also includes the following steps:

[0149] Second random content generation and sensor unit storage: After successfully saving the first random content sent by the edge computing gateway, the communication node immediately starts the generation of the third random time. The third random time adopts a sleep cycle adaptation strategy. The base cycle is set to 2-3 times the sensor unit heartbeat wake-up cycle, i.e., 6-9 seconds, to avoid triggering verification when the sensor is in a high sleep depth, reduce power consumption, and the random offset range is ±2 seconds. For example, if the base cycle is 6 seconds, the third random time may be 5 seconds or 7 seconds to ensure that the verification timing avoids the sensor's normal wake-up period.

[0150] When the third random time is reached, the communication node generates a 12-byte second random content based on a fragment of the first random content, such as the last 8 bytes of the first random content, using a hash algorithm (such as SHA-256). This second random content includes a node identifier fragment and a generation timestamp, creating a logical relationship between the second random content and the first random content that is related but not duplicated, thus avoiding verification failure caused by the leakage of a single piece of content. Subsequently, the communication node encapsulates the second random content into a sensor verification data packet, adds a node-sensor verification identifier to the frame header, with a length of approximately 15 bytes, and sends it to the sensing unit through the second detection link.

[0151] After receiving a data packet in low-power detection mode, the sensing unit does not need to be fully woken up. It only starts the storage module to save the second random content to a temporary storage area, such as RAM. It is automatically cleared after power failure to avoid long-term occupation of storage space. At the same time, it returns a very simple response to confirm storage, with only 1 byte identifier, to reduce transmission power consumption. After receiving the data, the communication node records the storage status and does not provide real-time feedback to the gateway to maintain the confidentiality of the verification process.

[0152] Cross-link read command issuance and dual content acquisition: Within 1-2 minutes after generating the first read command, the edge computing gateway initiates the generation of the fourth random time. The fourth random time is not fixedly related to the third random time. The basic period is set to 3-4 minutes to ensure that the sensing unit has enough time to store the second random content. The random offset range is ±30 seconds. For example, if the basic period is 3 minutes, the fourth random time may be 2 minutes and 40 seconds or 3 minutes and 20 seconds to avoid conflicts between the verification process and regular data transmission.

[0153] When the fourth random time interval is reached, the gateway generates a second read instruction, containing the target communication node address, the sensor unit device address ADDR, a dual-content read identifier, and a read timeout threshold of 8 seconds, and sends it to the communication node through the first detection link; after receiving the instruction, the communication node initiates cross-link collaborative reading:

[0154] Local read: Retrieve the first random content from its own non-volatile storage area as the second read content.

[0155] Remote reading: A content retrieval command containing a second random content storage address is sent to the sensing unit via the second detection link. The sensing unit responds within the heartbeat detection cycle, retrieves the second random content from the temporary storage area, and transmits it back to the gateway through the communication node. This content is the third read content.

[0156] If the communication node does not receive a feedback from the sensor unit within the timeout threshold, such as when the second link is momentarily interrupted, it returns a sensor no-response flag to the gateway; if the contents of the sensor unit's temporary storage area are lost, such as when there is an unexpected power outage, it returns a content missing flag.

[0157] Dual Content Association Verification and Collaborative Fault Location: After receiving the second read content (first random content) and the third read content (second random content), the edge computing gateway initiates association matching verification. First, it uses a preset algorithm, such as decrypting the second random content, to check whether it contains fragments of the first random content, verifying the logical association between the two. Then, it compares the content integrity byte by byte.

[0158] If the correlation is established and the content is completely consistent, the second detection link is determined to be stable. This also proves that there are no abnormalities in the collaborative communication between the first and second links, and that data can be transmitted completely across links. The gateway only records the verification results and continues to maintain regular monitoring.

[0159] If the verification fails, such as due to invalid correlation, content mismatch, or no response / missing content from the sensor, a communication link abnormality warning will be issued immediately. The warning message will clearly indicate the faulty link, such as a mismatch in the content of the second link or a failure of the dual-link coordination, and will also indicate the address of the faulty node and the sensor unit, as well as the verification time.

[0160] Furthermore, the gateway initiates hierarchical fault localization: if the second read content is correct but the third read content is abnormal, it is determined that the fault exists only in the second link, such as sensor storage failure or weak node-sensor signal; if the second read content is abnormal, it is determined that the fault may exist in the first link or the local storage of the communication node, such as packet loss in the gateway-node link or damage to the node storage area; the hierarchical localization results are synchronously uploaded to the operation and maintenance platform, which facilitates targeted troubleshooting by staff. If only the second link needs to be checked, there is no need to debug the gateway on-site, reducing operation and maintenance costs.

[0161] By adopting the above technical solution, on the one hand, the random time is matched with the sleep cycle to accurately capture the hidden faults in the gap between sensor sleep and wake-up, such as the loss of content caused by the decrease in signal reception sensitivity during sleep, thus avoiding the timeliness blind spot of fixed period verification; on the other hand, the correlation verification of two levels of random content not only verifies the stability of the second link, but also simultaneously verifies the collaborative transmission capability of the first and second links, such as the complete link of data from the gateway to the node and then to the sensor, thus avoiding collaborative faults that are ignored by single link verification.

[0162] Strong electromagnetic interference at power transmission and transformation sites, such as electromagnetic pulses generated by high-voltage equipment discharge, can easily cause dormant sensing units to be falsely triggered and awakened, resulting in unnecessary power consumption. Therefore, after a sensing unit receives response data and enters dormancy, an anti-interference closed loop needs to be formed through dormancy state verification and dynamic power adjustment. This specifically includes the following steps:

[0163] Silent Read Verification After Sleep: After the sensing unit sends a sleep confirmation signal (1-byte identifier) ​​back to the communication node, theoretically all functions except heartbeat detection should be turned off, and only a weak signal detection cycle of 3 seconds should be maintained, with power consumption reduced to below 10μA. At this time, the edge computing gateway starts the silent read mechanism. Without sending any request data packets to avoid actively waking up the sensor, a no-request read instruction is sent to the target sensing unit through the communication node. A sleep verification identifier with a length of about 5 bytes is added to the frame header, and the signal strength is reduced to below the minimum wake-up threshold to ensure that it can only be received by the sensor that is not in sleep mode.

[0164] After the communication node forwards the instruction through the second detection link, it waits for the response from the sensing unit: if the sensing unit is truly in sleep mode, its signal receiving module in sleep mode is in a closed state and cannot respond to the read instruction, the communication node returns no data received to the gateway after a 10-second timeout; if the sensing unit is falsely awakened due to electromagnetic interference and is in a standby state, it will respond to the read instruction and send the stored collected data (such as the partial discharge data collected last time) back to the communication node, which then forwards it to the gateway.

[0165] The gateway determines the sleep state based on whether it receives collected data: if it successfully reads complete collected data and the format conforms to the preset standard, it marks the sensor unit as having failed to sleep; if no data is received or only fragmented data is received, it marks the sleep state as having succeeded, and records the verification time and sensor unit address.

[0166] Multi-sensor sleep state statistics: The edge computing gateway performs the above silent reading verification one by one on all sensing units that have been issued sleep commands, such as 20-50 partial discharge sensors in a monitoring area, according to the polling order. After completing the traversal, it collects two types of data:

[0167] Number of sleep failures: Record the total number of sensor units marked as having failed to sleep, e.g., 5.

[0168] Number of successfully hibernating units: Record the total number of sensor units marked as successfully hibernating, such as 45.

[0169] The percentage of sleep failures is then calculated using the formula: Percentage of sleep failures = (Number of sleep failures / Total number of sensor units) × 100%. For example, 5 / 50 × 100% = 10%. This percentage directly reflects the degree of influence of current environmental electromagnetic interference on the sleep state of the sensor units. The higher the percentage, the stronger the interference, and the more important it is to strengthen anti-interference measures.

[0170] Sleep failure percentage determination and early warning adjustment: The preset sleep failure percentage is set according to the interference level at the power transmission and transformation site, usually between 8% and 12%. For example, if set to 10%, the gateway will compare the calculated sleep failure percentage with the set value.

[0171] If the percentage of hibernation failures is less than or equal to the set percentage (e.g., 8% ≤ 10%): the current environmental interference is considered to be within an acceptable range. No power adjustment is required; simply record the statistical results and continue regular monitoring.

[0172] If the percentage of hibernation failures exceeds the set percentage (e.g., 12% > 10%): Immediately activate the dual-response mechanism.

[0173] A warning message will pop up on the power transmission and transformation monitoring system interface, indicating that the current percentage of sleep failures is 12%, which exceeds the set value by 10%. Please check the electromagnetic interference source in the area. The message will also indicate the specific location of the sleep failure sensor, such as sensor ADDR: XXX near phase A of GIS equipment, to guide maintenance personnel to troubleshoot the interference source, such as loose high-voltage connectors or damaged shielding layers.

[0174] Based on the percentage of sleep failures exceeding the threshold, the transmit power of the edge computing gateway and communication nodes is negatively adjusted; the stronger the interference, the lower the power, thus reducing interference from the device's own signals to the sleep sensor.

[0175] Gateway power adjustment: The initial transmit power is set to 18dBm. For every 1% increase in the percentage of sleep failures beyond the set value, the power is reduced by 0.5dBm, down to a minimum of 15dBm. If it exceeds 2%, the power is reduced from 18dBm to 17dBm.

[0176] Node power adjustment: The initial transmit power is set to 17dBm. For every 1% increase in the percentage of sleep failures beyond the set value, the power is reduced by 0.4dBm, down to a minimum of 14dBm. If it exceeds 2%, the power is reduced from 17dBm to 16.2dBm.

[0177] After power adjustment, the gateway needs to re-perform sleep verification after 30 minutes. If the failure percentage drops below the set value, maintain the current power; if it still exceeds the set value, continue to reduce the power to the minimum value, and upgrade the warning level, such as sending an SMS reminder to the maintenance personnel.

[0178] Coordination of sleep verification and wake-up logic: To avoid silent read verification from accidentally triggering wake-up, the signal strength of the signal without a read request command must be strictly controlled and set below -90dBm, which is lower than the sensor unit wake-up threshold of -85dBm, to ensure that only sensors that are not in sleep mode can receive signals. At the same time, if the sensor unit is marked as sleep failure three times in a row, the gateway will issue a sleep parameter reset command during the next wake-up, such as extending the sleep detection period to 5 seconds and raising the wake-up threshold to -80dBm, to further enhance the sensor's anti-interference capability and reduce the probability of false wake-up.

[0179] By adopting the above technical solutions, on the one hand, silent reading verification can accurately identify false wake-ups caused by electromagnetic interference, avoiding power waste caused by sensor false sleep; on the other hand, based on the negative correlation power adjustment of the failure percentage, signal interference between gateways and nodes is reduced from the source, and combined with early warning and parameter reset, a multi-layer anti-interference guarantee is formed.

[0180] This application also discloses an edge computing gateway and sensor unit over-the-air wake-up system, including a processor, wherein the processor executes the steps of the edge computing gateway and sensor unit over-the-air wake-up method as described in any of the above embodiments.

[0181] This application also discloses a storage medium storing a program that, when executed by a processor, implements the steps of the edge computing gateway and sensing unit over-the-air wake-up method described in any of the above embodiments.

[0182] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. An edge computing gateway and sensor unit over-the-air wake-up method, characterized in that, The application discloses a sensing unit, a communication node and an edge computing gateway based on a sequential wireless connection, and comprises the following steps. An edge computing gateway acquires node information, unit information and detection information according to a network construction instruction. The edge computing gateway establishes a first detection link with a corresponding communication node according to the node information, and sends the unit information and the detection information to the corresponding communication node for a first check of the detection information after the first detection link is established. The communication node establishes a second detection link with a corresponding sensing unit according to the unit information and performs a second check of the detection information. If the detection information passes the two checks, a wake-up cycle mode is set, and after the wake-up cycle mode is set, the sensing unit is in a low-power mode and detects whether a request data packet is received. The edge computing gateway sends a request data packet to the sensing unit through the communication node, and the sensing unit triggers wake-up after receiving the request data packet, and sequentially performs registration, sampling, data processing and uploading to generate sensing data and upload the sensing data to the edge computing gateway through the communication node. The edge computing gateway receives and analyzes the sensing data, and sends a response data to the sensing unit through the communication node, and the sensing unit sleeps after receiving the response data. The edge computing gateway polls the next sensing unit after sending the response data. In the step of sending the unit information and the detection information to the corresponding communication node and performing the first check of the detection information, the following sub-steps are further included. The edge computing gateway sends the unit information and the detection information to the communication node, and the communication node receives the unit information and the detection information. The communication node searches for an established second detection link according to the unit information, generates repeated response data if the second detection link is established, or waits for the second detection link to be established and then generates unit response data. After waiting for a set time, if the communication node does not establish the second detection link corresponding to the unit information, unit error data is generated, otherwise, the communication node searches for a stored information library according to the detection information. If the detection information corresponds to the information library, detection response data is generated, otherwise, detection error data is generated. If the detection response data is generated, the communication node sends single-point data of a first length to the sensing unit through the second detection link corresponding to the unit information, and the communication node sends multi-point data of a second length to adjacent multiple communication nodes through a temporary link; wherein the first length is greater than the second length. The sensing unit returns single-point response data after receiving the single-point data, and the communication node returns multi-point response data after receiving the multi-point data. The communication node packages the generated data into node response data, and then returns the node response data to the edge computing gateway. The edge computing gateway receives and analyzes the node response data. If the unit error data is analyzed, a unit warning prompt is given, if the repeated response data is analyzed, a unit repetition prompt is given, if the detection error data is analyzed, a first detection link warning prompt is given, if the detection response data is analyzed, a first detection link establishment success prompt is given, if the single-point response data is analyzed, a communication depth normal prompt is given, and if the multi-point response data is analyzed, a communication expansion normal prompt is given.

2. The edge computing gateway and sensing unit over-the-air wake-up method of claim 1, wherein, If the communication depth is normal, the following steps are performed: The communication node sends a single-point instruction to the sensing unit through the second detection link; The sensing unit returns cyclically collected acquisition data generated by the sensing target in response to the single-point instruction, and the length of the acquisition data is a third length; wherein the first length > the third length > the second length; The communication node receives the acquisition data; if the format of the acquisition data matches the preset data format, the communication node returns the content of the sensing unit to the edge computing gateway.

3. The edge computing gateway and sensing unit over-the-air wake-up method of claim 1, wherein, The method further includes the following steps: The first detection link and the second detection link are both wireless links; The communication intensity is detected based on the first detection link as a first link intensity, and the communication intensity is detected based on the second detection link as a second link intensity; A first relative value is calculated based on the first link intensity and a preset first set intensity, and a second relative value is calculated based on the second link intensity and a preset second set intensity; A whole-link relative value is calculated based on the first relative value and the second relative value; The transmission power of the communication node is adjusted based on the negative correlation of the whole-link relative value; The transmission power of the edge computing gateway is adjusted based on the negative correlation of the first relative value; The transmission power of the sensing unit is adjusted based on the negative correlation of the second relative value.

4. The edge computing gateway and sensing unit over-the-air wake-up method of claim 3, wherein, The method further includes the following steps: The second length is adjusted based on the negative correlation of the whole-link relative value; the larger the whole-link relative value, the smaller the second length; the smaller the whole-link relative value, the larger the second length; The first length or the third length is adjusted based on the negative correlation of the second relative value; the larger the second relative value, the smaller the first length or the third length; the smaller the second relative value, the larger the first length or the third length.

5. The edge computing gateway and sensing unit over-the-air wake-up method of claim 1, wherein, The method further includes the following steps: The edge computing gateway generates a first random content based on a first random time, and sends the first random content to the communication node; the communication node saves the first random content; The edge computing gateway generates a first read instruction based on a second random time, and reads the first random content saved by the communication node to obtain a first read content; The matching of the first random content and the first read content is verified; if the verification fails, a data exception warning is issued.

6. The edge computing gateway and sensing unit over-the-air wake-up method of claim 5, wherein, The method further includes the following steps: After saving the random content, the communication node generates a second random content based on a third random time, and sends the second random content to the sensing unit; the sensing unit saves the second random content; The edge computing gateway generates a second read instruction based on a fourth random time, reads the first random content through the communication node to obtain a second read content, reads the second random content saved by the sensing unit through the communication node to obtain a third read content, and verifies the matching of the second read content and the third read content; if the verification fails, a communication link exception warning is issued.

7. The edge computing gateway and sensing unit over-the-air wake-up method of claim 3, wherein, After receiving the response data, the sensing unit sleeps, and further includes the following sub-steps: After the sensing unit sleeps, the edge computing gateway reads the acquisition data in the sensing unit through the communication node without issuing a request data packet; if the acquisition data is successfully read, the sensing unit is marked as sleeping failure; otherwise, the sensing unit is marked as sleeping success; After traversing the plurality of sensing units, the edge computing gateway records the number of sleeping failures and the number of sleeping successes; The percentage of sleeping failures is calculated based on the number of sleeping failures and the number of sleeping successes; If the hibernation failure percentage is greater than a preset set percentage, a sensing unit early warning prompt is performed, and the transmission power of the edge computing gateway and the transmission power of the communication node are adjusted according to the negative correlation of the hibernation failure percentage.

8. An edge computing gateway and sensor unit over-the-air wake-up system, comprising: The processor executes the steps of the edge computing gateway and sensing unit air wake-up method according to any one of claims 1-7.

9. A storage medium, characterized by The storage medium stores a program, and the program is executed by the processor to realize the steps of the edge computing gateway and sensing unit air wake-up method according to any one of claims 1-7.

Images