A method and system for online monitoring of the state of a return cable of a distributed master-slave architecture

Abstract

The application relates to a backflow cable state online monitoring method and system of a distributed master-slave architecture, and belongs to the technical field of backflow cable monitoring. The monitoring method comprises the following steps: receiving a preconfigured device topology relationship table; controlling a slave device to generate an initial characteristic signal; scanning a current channel state parameter of a backflow cable, dynamically selecting a communication bandwidth and a modulation mode; modulating the initial characteristic signal, generating a carrier characteristic signal and injecting the carrier characteristic signal into the backflow cable; controlling a master device to receive the carrier characteristic signal on the backflow cable, and extracting a device ID and a current timestamp in the carrier characteristic signal; comparing to generate a first abnormal flag bit and a second abnormal flag bit, and fusing to generate a comprehensive abnormal decision code; querying a logical association relationship in the device topology relationship table, determining a fault section identifier, generating a state report and uploading the state report to a remote monitoring platform. The application realizes intelligent and non-contact real-time monitoring of a connection state of a backflow cable in a high-voltage cable line.

Description

Technical Field

[0001] This application relates to the field of return cable monitoring technology, and in particular to a method and system for online monitoring of return cable status using a distributed master-slave architecture. Background Technology

[0002] With the increasing density of underground cable lines in cities, a large number of high-voltage cables using single-core structures require cross-connection of metal sheaths or direct grounding at both ends due to their electromagnetic induction characteristics. This is to limit sheath circulating current and ensure equipment safety. In such grounding systems, the return cable, as a key conductor connecting the metal sheath of the cable terminal head to the grounding box or protector, plays a crucial role in fault current discharge and overvoltage protection. The integrity of its physical connection directly affects the safe and stable operation of the entire cable line and even the regional power grid.

[0003] However, since return cables are mostly laid in open environments such as outdoor cable wells or tunnel entrances, they are exposed to adverse conditions such as humidity, corrosion, and frequent human activity for a long time. This makes them prone to problems such as loose joints, poor contact due to oxidation, and mechanical damage. If these problems are not detected in time, they may cause local overheating and insulation deterioration, or even open circuits in the sheath, causing the metal sheath to be subjected to excessively high induced voltage and break down the main insulation. This seriously threatens the reliability of power supply and may cause power safety problems.

[0004] Currently, maintenance personnel need to regularly visit the site to visually inspect connection points for surface abnormalities such as rust, overheating, discoloration, and missing fasteners, and use infrared thermometers to scan temperatures to identify hot spots. However, this traditional method suffers from significant time lag and high labor costs, making it difficult to achieve continuous 24 / 7 monitoring. This is especially true for line nodes located in remote mountainous areas, across river bridges, or deep underground utility tunnels, where inspections are often only completed once every few months, failing to meet the basic requirements of modern smart grids for critical equipment to be "observable, measurable, and controllable." More significantly, when a return cable experiences a hidden break or partial strand disconnection, there may be no obvious signs of damage, making it extremely difficult to detect using conventional methods. The problem is often only traced and located after the system activates protection mechanisms or experiences a power outage, by which time actual losses have already occurred.

[0005] Furthermore, while some systems have attempted to introduce electrical parameter monitoring devices based on DC resistance testing or low-frequency AC current detection to indirectly assess connection status by measuring the continuity of the return path, such solutions typically require the insertion of sampling elements or the additional deployment of signal injection points into the existing grounding loop. This not only alters the original electrical topology and adds new potential fault points, but also necessitates power outages during construction, severely impacting power supply continuity. These methods are difficult to implement and economically inefficient. Moreover, these methods generally only determine the continuity of the entire section, lacking spatial resolution and the ability to precisely locate the specific fault location. Troubleshooting in complex, multi-branched grounding networks is inefficient, still requiring significant manual intervention, and fails to fundamentally address the problem of slow maintenance response times. Summary of the Invention

[0006] To achieve intelligent, non-contact, real-time monitoring of the return cable connection status in high-voltage cable lines, this application provides a distributed master-slave architecture method and system for online monitoring of return cable status.

[0007] Firstly, this application provides an online monitoring method for the status of the return cable in a distributed master-slave architecture, employing the following technical solution:

[0008] A method for online monitoring of the backhaul cable status in a distributed master-slave architecture, the monitoring method comprising:

[0009] Receive a pre-configured device topology table; wherein the topology table includes a list of directly subordinate slave devices bound to the master device and the logical association between the master device and neighboring devices in adjacent segments;

[0010] Control the slave devices in the direct slave list to generate an initial feature signal carrying the device ID and the current timestamp;

[0011] Scan the current channel status parameters of the return cable, and dynamically select the communication bandwidth and modulation method based on the signal-to-noise ratio of the channel status parameters;

[0012] The initial characteristic signal is modulated based on the selected communication bandwidth and modulation method to generate a carrier characteristic signal, which is then injected into the return cable through a non-contact coupler.

[0013] The control host device receives the carrier characteristic signal on the return cable through a non-contact coupler and extracts the device ID and current timestamp from the carrier characteristic signal;

[0014] Based on the matching result between the device ID and the current timestamp and the device topology relationship table, a first anomaly flag is generated;

[0015] Calculate the signal strength attenuation rate of the continuously received carrier characteristic signal, and generate a second abnormal flag bit based on the comparison result of the signal strength attenuation rate and the preset attenuation threshold;

[0016] The first and second anomaly flags are combined to generate a comprehensive anomaly decision code;

[0017] Based on the comprehensive anomaly decision code and the device ID of the slave device, query the logical association in the device topology table to determine the fault segment identifier;

[0018] Based on the comprehensive anomaly decision code and fault segment identifier, a status report is generated and uploaded to the remote monitoring platform.

[0019] By adopting the above technical solutions, a device topology table containing cross-segment logical associations was established, realizing a leap from single-point monitoring to networked sensing. Channel scanning and adaptive modulation technologies enhanced communication reliability in complex electromagnetic environments. Non-contact coupling injection and spread spectrum modulation balanced installation convenience with signal penetration. The fusion analysis of two-factor anomaly criteria (ID matching and timestamp verification, signal attenuation trend) improved the accuracy and sensitivity of fault identification. Finally, relying on topology reasoning and neighbor verification mechanisms, engineering-level precise location of faulty sections was achieved, enhancing the observability and controllability of the return cable's operating status.

[0020] Secondly, this application provides a distributed master-slave architecture online monitoring system for return cable status, employing the following technical solution:

[0021] A distributed master-slave architecture online monitoring system for return cable status, the monitoring system comprising:

[0022] The topology table acquisition module is used to receive a pre-configured device topology table; wherein, the topology table includes a list of directly subordinate slave devices bound to the master device and the logical association between the master device and neighboring devices in adjacent segments;

[0023] The slave control module is used to control the slave devices in the direct slave list to generate an initial feature signal carrying the device ID and the current timestamp;

[0024] The communication configuration module is used to scan the current channel status parameters of the return cable and dynamically select the communication bandwidth and modulation method according to the signal-to-noise ratio of the channel status parameters.

[0025] The signal modulation module is used to modulate the initial characteristic signal based on the selected communication bandwidth and modulation method, generate a carrier characteristic signal, and inject it into the return cable through a non-contact coupler;

[0026] The host control module is used to control the host device to receive the carrier characteristic signal on the return cable through a non-contact coupler, and to extract the device ID and current timestamp from the carrier characteristic signal;

[0027] The first anomaly detection module is used to generate a first anomaly flag based on the matching result between the device ID and the current timestamp and the device topology relationship table;

[0028] The second anomaly detection module is used to calculate the signal strength attenuation rate of the continuously received carrier characteristic signal, and generate a second anomaly flag bit based on the comparison result of the signal strength attenuation rate and the preset attenuation threshold.

[0029] The comprehensive anomaly decision module is used to merge the first anomaly flag bit and the second anomaly flag bit to generate a comprehensive anomaly decision code.

[0030] The fault segment determination module is used to query the logical association relationship in the device topology table based on the comprehensive anomaly decision code and the device ID of the slave device to determine the fault segment identifier.

[0031] The status report upload module is used to generate a status report and upload it to the remote monitoring platform based on the comprehensive anomaly decision code and fault section identifier.

[0032] In summary, this application achieves at least one of the following beneficial technical effects: by constructing a device network with logical topological relationships and combining adaptive communication mechanisms with multi-dimensional anomaly criterion fusion analysis, intelligent, non-contact, real-time monitoring of the return cable connection status in high-voltage cable lines is realized. This application not only solves the problems of low efficiency and delayed response in traditional manual inspections, but also overcomes the technical bottlenecks of unreliable communication and ambiguous fault location in existing monitoring methods under complex electromagnetic environments. Attached Figure Description

[0033] Figure 1 This is a first flowchart illustrating the online monitoring method for the backflow cable status of a distributed master-slave architecture according to one embodiment of this application.

[0034] Figure 2 This is a second flowchart illustrating the online monitoring method for the backflow cable status of a distributed master-slave architecture according to one embodiment of this application.

[0035] Figure 3 This is a third flowchart illustrating the online monitoring method for the status of the return cable in a distributed master-slave architecture, according to one embodiment of this application.

[0036] Figure 4 This is a schematic diagram of the fourth process of the online monitoring method for the status of the return cable in a distributed master-slave architecture according to one embodiment of this application.

[0037] Figure 5 This is a schematic diagram of the fifth step of the online monitoring method for the status of the return cable in a distributed master-slave architecture according to one embodiment of this application.

[0038] Figure 6 This is a schematic diagram of the sixth step in the online monitoring method for the status of the return cable in a distributed master-slave architecture according to one embodiment of this application.

[0039] Figure 7 This is a schematic diagram of the seventh process of the online monitoring method for the status of the return cable in a distributed master-slave architecture according to one embodiment of this application.

[0040] Figure 8 This is a schematic diagram of the system architecture deployment of one embodiment of this application.

[0041] Figure 9 This is a topology diagram of the deployment of a monitoring unit on a return cable according to one embodiment of this application.

[0042] Explanation of reference numerals in the attached diagram: 101, Coupling module; 102, Power line carrier communication module; 103, Master unit; 104, Slave unit; 105, Communication module; 106, Database server; 107, Monitoring platform. Detailed Implementation

[0043] To make the purpose, technical solution, and advantages of this application clearer, the following description is provided in conjunction with the appendix. Figures 1-9 The present application will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.

[0044] This application discloses an online monitoring method for the status of the return cable in a distributed master-slave architecture.

[0045] Reference Figure 1 A method for online monitoring of the backflow cable status in a distributed master-slave architecture, the monitoring method including:

[0046] Step S101: Receive the pre-configured device topology table;

[0047] The topology table contains a list of directly subordinate slave devices bound to the master device and the logical association between the master device and neighboring devices in adjacent segments.

[0048] The topology table is not simply a list of devices; it includes the binding mapping between the host and its directly subordinate slaves, as well as the logical association information between the host and adjacent segment boundary devices (such as neighboring hosts or end slaves). This cross-segment "neighbor" association design breaks the closed nature of traditional single-point monitoring units, enabling multiple monitoring units to form a continuous spatial logical link, providing data support for subsequent cross-regional collaborative diagnosis.

[0049] For example, in long-distance power transmission lines, a master unit is set up every kilometer to manage several slave units. Each master unit also records the identity of the nearest downstream device, thus forming a relay-like topology chain. When a segment of the signal is interrupted, the specific fault section can be inferred by comparing the status of the preceding and following nodes, greatly improving spatial resolution. This topology table is usually entered by maintenance personnel during the installation and commissioning phase using local configuration tools and synchronized to all relevant master units. It is also uploaded to a remote monitoring platform for visualization and global scheduling.

[0050] Step S102: Control the slave devices in the direct slave list to generate an initial feature signal carrying the device ID and the current timestamp;

[0051] The initial characteristic signal contains two key pieces of information: the device's unique ID and the current timestamp. The device ID is the core field for identification, typically encoded using a MAC address or serial number to ensure uniqueness across the entire monitoring network. The timestamp reflects the specific moment the signal was sent and can be used to determine if the transmission delay exceeds a reasonable range. These two parameters together constitute a dual verification basis for the authenticity and timeliness of the signal source.

[0052] Furthermore, after generating a signal, the slave device does not continuously transmit but enters deep sleep to reduce power consumption. It is only triggered by a timer to execute a complete signal injection process once every preset wake-up period (e.g., every 5 minutes). This intermittent working mechanism significantly extends battery life in field deployment scenarios. It is worth noting that this process can operate autonomously without relying on external communication commands, demonstrating the system's robustness and decentralized nature.

[0053] Step S103: Scan the current channel status parameters of the return cable, and dynamically select the communication bandwidth and modulation method based on the signal-to-noise ratio of the channel status parameters;

[0054] To ensure reliable signal transmission in the non-ideal channel of the return cable, the system performs a channel scan before each communication to obtain the current channel state parameters and dynamically adjust the communication parameters accordingly. As part of the power system, the return cable is constantly exposed to strong electromagnetic interference, resulting in highly time-varying channel characteristics, especially during thunderstorms or when high-power equipment is started or stopped nearby, causing severe fluctuations in background noise. Therefore, traditional carrier communication with fixed bandwidth and modulation methods is highly susceptible to bit errors or even complete failure.

[0055] In this embodiment, the operating frequency band from 0.7MHz to 12MHz is divided into multiple subcarriers. The background noise power and instantaneous signal-to-noise ratio (SNR) of each subcarrier are measured, and a continuous subcarrier segment that meets the minimum SNR threshold and has the lowest noise level is selected as the actual communication bandwidth. For example, if a high SNR and low interference are detected in the 3.5MHz to 6.8MHz frequency band, this segment is preferentially selected for communication. Simultaneously, based on the average SNR value of this segment, a modulation scheme such as BPSK (low-rate, high-interference immunity), QPSK, or 16-QAM (high-rate but requiring a higher SNR) is chosen. This channel-aware adaptive strategy effectively avoids interference peaks and valleys, improves signal penetration and demodulation success rate, and ensures that a basic communication link can be maintained even under harsh operating conditions.

[0056] Step S104: Modulate the initial characteristic signal based on the selected communication bandwidth and modulation method, generate a carrier characteristic signal, and inject it into the return cable through a non-contact coupler;

[0057] The system modulates the initial characteristic signal using the selected communication bandwidth and modulation scheme to generate the final carrier characteristic signal. This process involves not only conventional digital modulation but also enhancement measures such as spread spectrum coding.

[0058] Specifically, the device ID and timestamp are first encoded using binary bit interleaving to shuffle the original bit order and mitigate the impact of burst errors, thus improving error correction capabilities. Then, the encoded stream is spread-spectrum modulated using a chirp waveform, extending the narrowband signal to a wider frequency band for transmission. Chirp signals possess excellent autocorrelation characteristics and resistance to multipath fading, making them particularly suitable for channel environments with uneven impedance and severe reflections, such as power lines. Their frequency varies linearly with time, with a duration of a fixed microsecond-level pulse Tμs (e.g., 1024μs), enabling high-gain despreading at the receiver through a matched filter, allowing recovery of the original information even under extremely low signal-to-noise ratio conditions. This approach is similar to a key technology in LoRa communication, giving the system powerful long-distance weak signal detection capabilities, making it particularly suitable for field conditions with enclosed grounding boxes and severe signal attenuation.

[0059] Subsequently, the generated carrier characteristic signal is injected into the return cable through a non-contact coupler. The coupler is essentially a high-frequency current transformer, consisting of a high-permeability ferrite core and a secondary coil wound around it. During installation, there is no need to cut or damage the original cable structure; it is simply snapped onto the outer surface of the return cable to achieve electromagnetic induction coupling. In transmitting mode, the electrical signal output from the slave device excites the coupler coil to generate an alternating magnetic field, which in turn induces a common-mode current in the return cable conductor, causing the characteristic signal to propagate along the cable. In receiving mode, the opposite occurs: the host-side coupler picks up the carrier current flowing on the cable and converts it into a voltage signal for subsequent processing. This non-intrusive design avoids physical modifications to the grounding system, ensuring power safety while significantly reducing construction difficulty and cost, making it particularly suitable for upgrading existing, operational power lines.

[0060] Step S105: The host device receives the carrier characteristic signal on the return cable through a non-contact coupler and extracts the device ID and current timestamp from the carrier characteristic signal;

[0061] The host device, after being woken up at a predetermined time, also receives the carrier characteristic signal from the return cable via a non-contact coupler and extracts the device ID and current timestamp from it. Since the signal may undergo multiple reflections, attenuation, and noise superposition during transmission, the receiving end needs to use high-performance digital signal processing algorithms for demodulation and diffraction, including automatic gain control (AGC), carrier synchronization, frame synchronization, and channel equalization, to ensure correct data reconstruction.

[0062] Step S106: Generate the first anomaly flag bit based on the matching result between the device ID and the current timestamp and the device topology relationship table;

[0063] Specifically, once decoding is successful, the host immediately initiates a matching and verification process: comparing the extracted device ID with the list of slave IDs that the host should receive in the locally stored topology table. If the ID of a responding slave cannot be recognized, it is determined to be "unmatched," and a first anomaly flag is generated. This may be due to the slave losing power, hardware failure, or a broken return cable, preventing the signal from reaching the host. Furthermore, even if the ID matches successfully, if the parsed timestamp differs from the host's local system time by more than a preset threshold (e.g., ±30 seconds), it is also considered an anomaly, triggering the first anomaly flag. This mechanism effectively identifies problems such as timeout retransmissions and out-of-order reception caused by severe multipath effects or relay failures, preventing the misjudgment of normal signals as valid responses.

[0064] Step S107: Calculate the signal strength attenuation rate of the continuously received carrier characteristic signal, and generate a second abnormal flag bit based on the comparison result of the signal strength attenuation rate and the preset attenuation threshold.

[0065] The host continuously calculates the signal strength attenuation rate to assess channel quality and line health. Since the slave transmits at a constant power and is located in a fixed position each time, the signal strength received by the host should theoretically fluctuate stably within a certain range. The system performs trend analysis on the RSSI values ​​of N consecutive valid receptions (e.g., N=5) to calculate the attenuation ratio relative to the historical average. When this attenuation rate continuously exceeds a preset threshold (e.g., -15dB), it indicates gradual degradation in the signal path, such as poor local contact due to connector oxidation, loose screws, or damaged shielding. Although this has not yet caused a complete disconnection, it has already affected communication quality, and a second anomaly flag is generated. This indicator is an important supplement to the early warning of "soft faults" (complete interruption) and helps to achieve early detection of potential problems and preventative maintenance.

[0066] Step S108: Combine the first anomaly flag bit and the second anomaly flag bit to generate a comprehensive anomaly decision code;

[0067] This step is not a simple "OR" logic, but rather employs a hierarchical weighting mechanism to reflect the severity and causes of different fault types. When both the first and second anomaly flags are true, it means that not only is the signal lost but also significantly attenuated, most likely due to physical disconnection or severe corrosion, belonging to the highest risk level. The system assigns a corresponding code (such as "0xFF") to trigger an emergency alarm. If only the first anomaly flag is true, it indicates that the signal has completely disappeared without any prior obvious attenuation trend, possibly due to sudden shearing or equipment failure, and is marked as "equipment communication anomaly." Conversely, if only the second anomaly flag is true, it is judged as a slowly deteriorating fault and classified as "line attenuation anomaly." This classification mechanism allows the backend to take differentiated handling strategies for different types of events; for example, the former is immediately dispatched for emergency repair, while the latter is included in the regular inspection plan, improving the scientific allocation of operation and maintenance resources.

[0068] Step S109: Based on the comprehensive anomaly decision code and the device ID of the slave device, query the logical relationship in the device topology table to determine the fault segment identifier;

[0069] Traditional monitoring systems often only report "a slave device is not responding," but struggle to pinpoint "which section of cable is causing the problem." This application introduces a "neighbor verification" mechanism: when the host detects a slave device signal loss, it proactively initiates a status verification request to the corresponding neighbor device in the topology table (usually the first device in the next segment). If the neighbor device reports normal reception, the problem is limited to the current host and the malfunctioning slave device; however, if the neighbor device also reports an anomaly, there may be cross-segment concurrent faults or a common connection point issue (such as a loose shared grounding busbar).

[0070] Through this two-way verification logic, the system can not only distinguish between single-point faults and regional faults, but also accurately output standardized fault segment ID formats. For example, "H01:S03" represents the segment between host H01 and its subordinate slave S03, while "H01:N-H02" represents the boundary segment between host H01 and its neighboring host H02. This structured identification method facilitates rapid parsing by the monitoring platform and highlights the fault location on the Geographic Information System (GIS) map.

[0071] Step S110: Generate a status report and upload it to the remote monitoring platform based on the comprehensive anomaly decision code and fault section identifier.

[0072] The system integrates comprehensive anomaly decision codes and fault segment identifiers, encapsulates them into a structured status report, and uploads it to the remote monitoring platform via a wireless communication protocol. This status report adheres to lightweight and highly compatible design principles, employing the MQTT protocol as the primary transmission channel due to its low overhead, support for publish / subscribe modes, and suitability for unstable network environments. The message structure includes header fields (protocol version, message type), payload fields (fault segment ID, anomaly code, timestamp), and a CRC-16 checksum to ensure data integrity and traceability. Upon receiving the message, the monitoring platform can automatically trigger alarm notifications (SMS, email, app push), generate work orders, link video surveillance to retrieve on-site footage, and even deploy drones for investigation, forming a complete intelligent operation and maintenance closed loop.

[0073] In the above implementation, a device topology table containing cross-segment logical associations is established, realizing a leap from single-point monitoring to networked sensing; channel scanning and adaptive modulation technologies enhance communication reliability in complex electromagnetic environments; non-contact coupling injection and spread spectrum modulation balance installation convenience and signal penetration; the fusion analysis of two-factor anomaly criteria (ID matching and timestamp verification, signal attenuation trend) improves the accuracy and sensitivity of fault identification; and finally, relying on topology reasoning and neighbor verification mechanisms, engineering-level precise location of faulty segments is achieved, improving the observability and controllability of the return cable's operating status.

[0074] Reference Figure 2 As one implementation of S103, the steps of scanning the current channel state parameters of the return cable and dynamically selecting the communication bandwidth and modulation method based on the signal-to-noise ratio of the channel state parameters include:

[0075] Step S201: Divide the communication frequency band into continuous subcarriers and collect the current channel state parameters of each subcarrier in real time; wherein, the current channel state parameters include background noise power and instantaneous signal-to-noise ratio;

[0076] Specifically, this step first involves dividing the pre-defined communication frequency band from 0.7MHz to 12MHz into subcarriers, discretizing the entire available spectrum into multiple equally spaced or unequally spaced frequency units, forming a set of continuously distributed subcarriers. This division is not a static configuration, but rather achieved by the high-performance DSP chip in the power line carrier communication module through its built-in FFT (Fast Fourier Transform) engine, which performs frequency domain mapping on the analog front-end ADC sampling data to obtain the energy distribution map on each subcarrier.

[0077] Based on this, the system collects the current channel state parameters of each subcarrier in real time, mainly including background noise power (in dBm) and instantaneous signal-to-noise ratio (SNR). Background noise power reflects the average interference intensity at that frequency point during periods without effective signal transmission, directly determining whether the subcarrier has basic communication feasibility. Instantaneous SNR, on the other hand, is the ratio of useful signal power to noise power calculated during periods of signal activity, and is a key indicator for measuring demodulation success rate. These two parameters together constitute the basic dataset for channel quality assessment. The measurement accuracy relies on a high-resolution ADC, a low-phase-noise local oscillator, and a precise time synchronization mechanism to ensure reliable channel sensing results even under weak signal conditions.

[0078] Step S202: Select a set of candidate subcarriers that meet a preset noise threshold based on the background noise power;

[0079] The system initially filters all subcarriers based on a pre-set noise threshold, eliminating those with excessively high background noise power that cannot support stable communication, thus forming a candidate subcarrier set. For example, if the noise threshold is set to -90dBm, any subcarrier with background noise higher than this value will be excluded, because excessively high noise floor means that the frequency band is occupied by strong interference sources, and even if it is used forcibly, it will lead to an extremely high bit error rate. The subcarriers retained after filtering constitute a pool of potentially usable frequency resources. However, these subcarriers may be fragmented and discontinuous, therefore further integration is needed to form an effective bandwidth range that can be used for actual communication.

[0080] Step S203: Calculate the signal-to-noise ratio gradient change rate of adjacent subcarriers in the candidate subcarrier set, and merge adjacent subcarriers with continuous signal-to-noise ratio gradient change rates to form candidate bandwidth intervals;

[0081] To construct a practically valuable communication channel, the system further analyzes the spatial variation trend of the signal-to-noise ratio (SNR) within the candidate subcarrier set. Specifically, it calculates the SNR gradient rate of change between adjacent subcarriers. The gradient rate of change reflects the smoothness of the channel quality along the frequency axis: when the SNR difference between two adjacent subcarriers is small and the variation trend is consistent, it indicates that the channel response within this local frequency band is relatively flat and suitable for use as a whole bandwidth; conversely, if the gradient changes drastically, it indicates the presence of significant frequency-selective fading or narrowband interference peaks and valleys, making merging unsuitable.

[0082] Based on this principle, the system groups consecutive subcarriers with a signal-to-noise ratio gradient change rate below a certain threshold (e.g., 3dB / subcarrier) into a candidate bandwidth interval. For example, if subcarriers 128-150 are detected at a certain moment and not only meet the noise level requirements but also have an SNR fluctuation of less than 2dB, they can be merged into a continuous bandwidth block with a width of approximately 2.2MHz. This method avoids the "sawtooth" bandwidth fragmentation problem caused by traditional methods that only divide the bandwidth according to an absolute SNR threshold, improves spectrum utilization efficiency, and also facilitates subsequent filter design and equalization algorithm convergence.

[0083] Step S204: Based on the total bandwidth and average signal-to-noise ratio of the candidate bandwidth range, dynamically select the target communication bandwidth and target modulation scheme by referring to the preset bandwidth-modulation scheme mapping table;

[0084] The preset bandwidth-modulation mapping table is essentially an experience-driven rule base that defines the optimal combination of communication parameters to be used under different channel conditions. For example, if the total bandwidth of a candidate interval reaches more than 8MHz and the average SNR exceeds 25dB, a large bandwidth of 10MHz combined with 16-QAM modulation is preferred to maximize data throughput; if the bandwidth is between 3 and 5MHz and the SNR is in the range of 15 to 25dB, a bandwidth of 3.2MHz combined with QPSK modulation is selected to achieve a balance between rate and reliability; and for cases where the bandwidth is less than 2.5MHz or the SNR is less than 15dB, the bandwidth is downgraded to 2.2MHz and BPSK modulation is enabled, sacrificing rate for the strongest anti-interference capability.

[0085] Understandably, this hierarchical decision-making mechanism draws heavily on the fundamental idea of ​​Shannon's channel capacity theorem, which states that there exists a maximum achievable information rate given a bandwidth and signal-to-noise ratio. The system optimizes communication performance by dynamically approximating this theoretical limit. More importantly, the entire selection process is autonomously completed by the power line carrier communication module, without manual intervention or remote command issuance, greatly enhancing the system's real-time response capability and field adaptability.

[0086] Step S205: Configure the carrier signal modulation parameters according to the selected target communication bandwidth and target modulation method.

[0087] Once the target communication bandwidth and target modulation scheme are determined, the system immediately reconfigures the modulation parameters of the carrier signal according to the selected parameters, including but not limited to the DAC output frequency range, transmit filter cutoff frequency, modulation order, symbol mapping scheme, and coding redundancy. For example, when deciding to use 16-QAM modulation, the corresponding constellation mapping logic needs to be enabled, and the forward error correction code rate needs to be increased to cope with the higher risk of bit errors; while when switching to BPSK mode, the coding overhead can be appropriately reduced to reduce latency.

[0088] Furthermore, the receiver's power gain amplifier (PGA) and adaptive equalizer also update their parameter settings synchronously, ensuring a complete parameter loop is formed throughout the entire link. The entire configuration process is completed within milliseconds, guaranteeing that the characteristic signal can be optimally transmitted based on the latest channel state before each transmission, thereby significantly improving demodulation success rate and system stability.

[0089] The above implementation achieves refined management and intelligent utilization of the non-ideal channel of the return cable. From frequency domain subcarrier partitioning to background noise screening, from signal-to-noise ratio gradient analysis to continuous bandwidth clustering, and then to dynamic parameter matching and modulation configuration based on multi-dimensional indicators, this mechanism can not only effectively avoid strong interference frequency bands, but also tap into the maximum communication potential in limited high-quality spectrum resources, enabling stable and reliable characteristic signal transmission even in harsh electromagnetic environments. This not only ensures the master's accurate perception of the slave's status, but also provides a solid data foundation for the precise location of subsequent fault sections, improving the practicality, robustness, and intelligence of the return cable status monitoring system.

[0090] Reference Figure 3 As one implementation of step S106, the step of generating the first anomaly flag bit based on the matching result of the device ID and the current timestamp with the device topology relationship table includes:

[0091] Step S301: Obtain the set of slave device IDs of the direct slave list bound in the device topology relationship table;

[0092] The slave device ID set is a statically configured data structure after master-slave pairing is completed during system initialization. It records the unique identifiers of all slave devices whose signals the current master should normally receive, typically represented by MAC addresses, serial numbers, or custom codes. These IDs constitute a "whitelist" of expected communication targets for the master; any signal not on this list will be considered illegal or external interference and will not be processed. This whitelist-based access control mechanism not only enhances system security and prevents false responses caused by external noise or crosstalk from nearby lines, but also provides a basis for subsequent accurate positioning.

[0093] Step S302: Parse the device ID and current timestamp in the carrier characteristic signal;

[0094] After the host successfully demodulates the carrier characteristic signal received by the disseminated and amplified signal, the system immediately initiates the information extraction process to separate two key fields: device ID and current timestamp. The device ID identifies the source of the signal, while the timestamp reflects the precise time information of the signal transmission. Together, they constitute the basic metadata of a valid communication event.

[0095] It should be noted that due to the characteristics of power line channels, such as strong noise, multipath effects, and frequency-selective fading, the original bit stream may be flipped or lost during transmission. Therefore, this step relies on highly robust forward error correction coding (such as convolutional codes or LDPC), frame synchronization header detection, and CRC check mechanisms to ensure the authenticity and integrity of the extracted data. Once the data packet passes the check, the system considers its content reliable and proceeds to the next stage of logical analysis.

[0096] Step S303: Match the parsed device ID with the slave device ID set;

[0097] Specifically, this process employs a precise string comparison algorithm, requiring that the ID to be checked must be completely identical to any member in the set at the character level to be considered a successful match. For example, if a slave ID is "SL-2024-HD01", but the receiver parses it as "SL-2024-HD02" or, due to a bit error, the last character becomes "SL-2024-HD0L", even if the difference is only one character, it is considered a mismatch. This strict consistency requirement eliminates the risk of misjudgment that may arise from fuzzy matching, especially in large-scale deployment scenarios, avoiding false alarms caused by confusion between devices with similar IDs.

[0098] Step S304: Determine whether the match is successful based on the matching result; if not, proceed to step S305; if yes, proceed to step S306.

[0099] Step S305: Output the first exception flag as true;

[0100] Step S306: Calculate the absolute time difference between the parsed current timestamp and the host system time;

[0101] When the system detects that the ID of a certain response signal cannot be found in the direct list, it immediately outputs the first abnormal flag as true, indicating that the source of the signal is not within the management scope of this monitoring unit, and is most likely caused by a broken signal path, power failure of the slave device, replacement of equipment, or communication link disorder.

[0102] However, even if the device ID matches successfully, it cannot be simply assumed that communication is normal. Considering that the return cable is a non-ideal transmission medium, the signal may experience significant delays during propagation, especially when the line is long, has many connectors, or contains local impedance discontinuities. Reflection and multiple refractions can cause the signal arrival time to deviate significantly from expectations. Furthermore, each slave device's internal clock source may have varying degrees of crystal oscillator drift, and the accumulated time deviation after long-term operation may exceed reasonable limits. Therefore, ID matching alone is insufficient to comprehensively assess communication quality; a time dimension verification must be introduced.

[0103] To this end, the system further calculates the absolute time difference between the parsed current timestamp and the host's local system time. This difference reflects the total delay between the slave's sending time and the host's receiving time, including the physical transmission time of the signal, processing delay, and clock asynchrony components.

[0104] Step S307: Determine whether the absolute time difference exceeds the preset time tolerance threshold. If yes, proceed to step S305; otherwise, proceed to step S308.

[0105] Step S308: Output the first exception flag as false.

[0106] If the difference exceeds the preset time tolerance threshold, it is also judged as abnormal, and the first abnormal flag is output as true.

[0107] It should be noted that this time tolerance threshold is not a fixed constant, but rather a dynamically adjusted result based on historical communication behavior. Specifically, the system continuously calculates the average transmission delay of all normal communication events over a past period, multiplies this by a safety factor greater than 1 (e.g., 1.5), and then adds a certain clock drift compensation (e.g., ±10 seconds) to form the final tolerance window. For example, in a typical 3-kilometer return cable line, the measured average transmission delay is approximately 80 milliseconds. After safety amplification, the tolerance is set to 150 milliseconds. If a received timestamp shows a delay of 600 milliseconds, it can be suspected that the signal has undergone abnormal detours or multiple hops for retransmission, which is an abnormal path and should trigger an anomaly flag. This dynamic threshold mechanism avoids frequent false alarms caused by overly strict fixed threshold settings, while also preventing missed detections due to overly lenient thresholds.

[0108] Furthermore, after the entire anomaly detection process is completed, the system can introduce an anomaly verification mechanism to further confirm the validity of the first anomaly flag. That is, when a slave device is initially determined to be disconnected or out of clock synchronization, the master device will not immediately report the final conclusion, but will proactively send a retransmission command to the slave device to attempt to re-establish communication. If no valid response is received within the preset retransmission timeout period, the anomaly is officially confirmed, the first anomaly flag is set to valid, and the system is prepared to upload the data. This mechanism effectively filters out temporary communication failures caused by momentary interference, brief disconnections, or wake-up asynchrony, ensuring that only persistent and reproducible faults are recorded and alarmed, greatly improving the system's stability and reliability.

[0109] In the above implementation, a dual-verification logic framework is constructed through precise matching of device IDs and dynamic evaluation of timestamp differences, realizing a composite anomaly identification mechanism that ranges from "whether a signal was received" to "whether a correct and timely signal was received." This technical solution can not only identify complete disconnection faults (such as cable cuts or equipment damage), but also capture latent anomalies (such as signal detours or clock drift), and eliminate interference from accidental factors through a retransmission verification mechanism. Thus, while ensuring sensitivity, accuracy is also taken into account, improving the reliability of the system's criteria and its ability to resist false alarms in complex electromagnetic environments.

[0110] Reference Figure 4 As one implementation of step S107, the step of calculating the signal strength attenuation rate of the continuously received carrier characteristic signal and generating a second abnormal flag bit based on the comparison result of the signal strength attenuation rate and a preset attenuation threshold includes:

[0111] Step S401: Obtain the received signal strength values ​​of the carrier characteristic signals of the same slave device received N consecutive times, and generate a signal strength sequence;

[0112] Specifically, the Received Signal Strength Indication (RSSI) values ​​from N consecutive carrier characteristic signals received from the same slave device are first obtained from the host side, forming a time-ordered signal strength sequence. Here, N is a preset statistical window length, typically 5-10 sampling periods. The specific value can be configured according to the on-site communication frequency (e.g., once every 5 minutes) to balance response speed and judgment stability. Each RSSI value represents the energy level of the slave signal after transmission via the return cable to the host coupling module at the current moment, measured in dBm. Its magnitude is affected by various factors, including the conductor resistance of the return cable itself, connector contact impedance, external electromagnetic interference, changes in ambient temperature and humidity, and the degree of equipment aging. Arranging these discrete data points in chronological order forms the basic dataset reflecting the evolution trend of the communication link quality.

[0113] It is worth noting that due to the inherent volatility of power line channels, a single RSSI measurement is easily affected by instantaneous noise and may fluctuate. Therefore, a fault cannot be determined solely based on a single low reading. It is necessary to extract the essential change patterns by modeling the overall trend of multiple historical samples and eliminate misjudgments caused by random disturbances.

[0114] Step S402: Using the sampling time point as the independent variable and the signal strength value as the dependent variable, perform linear fitting on the signal strength sequence and calculate the signal strength attenuation rate per unit time.

[0115] This process essentially involves abstracting a complex, non-stationary signal into an optimally approximating trend line, thereby quantifying the average change in signal strength per unit time, i.e., the signal strength attenuation rate.

[0116] In this embodiment, to achieve this goal, the system employs the Least Squares Method as the fitting algorithm. This method calculates the slope k that best represents the overall trend by minimizing the sum of squares of the vertical distances between all actual sampling points and the fitted line. This slope k is the attenuation rate, usually expressed in dB / h (decibels per hour). It intuitively reflects the rate of performance degradation of the communication link: a positive value indicates signal enhancement (generally rare), while a larger absolute value of a negative value indicates more severe signal attenuation. For example, on a well-functioning return cable, if the RSSI of seven consecutive samples are -68, -69, -70, -71, -72, -73, and -74 dBm, corresponding to times of 0, 5, 10, 15, 20, 25, and 30 minutes, the fitted attenuation rate is approximately -2.4 dB / h, indicating that the signal is decaying at a relatively rapid rate, far exceeding the normal natural fluctuation range, which should raise concerns. This mathematical modeling method effectively overcomes the problem of the traditional threshold comparison method being sensitive to occasional low values, and realizes the leap from transient judgment to trend early warning.

[0117] Step S403: Compare the signal strength attenuation rate with the preset attenuation threshold. If the signal strength attenuation rate is greater than the preset attenuation threshold, generate a line attenuation anomaly flag.

[0118] The calculated attenuation rate is compared with a preset attenuation threshold to preliminarily determine whether there is an abnormal trend. The preset attenuation threshold is not a fixed constant, but is dynamically adjusted based on the maximum attenuation rate sample of similar cable segments under historical normal conditions.

[0119] Specifically, the system continuously collects and records attenuation rate data during the initial stable phase of equipment operation, calculates its moving average, and then superimposes a compensation coefficient that considers the material aging effect during long-term service, forming a dynamic threshold that changes slowly over time. For example, the typical attenuation rate of a newly installed return cable connector is -0.3 dB / h, and the initial threshold is set to -1.0 dB / h. As the years of operation increase, slight oxidation of the contact surface causes the background attenuation to rise to -0.6 dB / h, at which point the system automatically relaxes the threshold to -1.3 dB / h to avoid false alarms triggered by normal aging. This adaptive threshold mechanism significantly improves the robustness and applicability of the system, enabling it to both accurately detect sudden degradation and tolerate performance drift within a reasonable range.

[0120] Step S404: Count the number of times the line attenuation anomaly flag is triggered in N consecutive detections. When the number of triggers reaches the preset continuous trigger threshold, output the second anomaly flag as true.

[0121] When the attenuation rate exceeds a preset threshold, the system immediately generates a line attenuation anomaly flag as a temporary alarm for the current detection cycle. However, considering the possibility of false trends caused by brief strong interference in complex electromagnetic environments (such as common-mode noise bursts caused by large load start-ups and shutdowns), a single exceedance is insufficient to confirm a genuine fault. Therefore, the system further introduces a continuous triggering mechanism: it counts the number of times the line attenuation anomaly flag has been activated in the most recent N detections and compares this count with a preset continuous triggering threshold. Only when the cumulative number of triggers reaches or exceeds this threshold is the second anomaly flag ultimately output as true.

[0122] Furthermore, the trigger threshold itself is dynamically adjustable, adaptively configured based on the current channel stability. For example, when the background noise fluctuation variance is low (indicating a quiet environment and stable channel), the threshold requirement can be appropriately lowered (e.g., set to N / 2) to improve sensitivity; while when noise fluctuations are severe (e.g., during thunderstorms or when nearby high-frequency equipment is operating), the threshold is increased to 2N / 3 to enhance anti-interference capabilities and prevent false triggering. This dual dynamic mechanism (dynamic threshold + dynamic threshold) makes the entire detection process both flexible and rigorous.

[0123] Furthermore, after generating the second anomaly flag, the system can initiate an anti-interference verification process to further confirm its effectiveness. For example, it can actively switch to a backup communication frequency band to re-collect several sets of signal strength data and observe whether the attenuation rate exceeds the limit. Since different frequency bands are affected by interference sources differently, if a consistent deterioration trend is observed in multiple independent frequency bands, the possibility of local spectrum pollution can be largely ruled out, thus highly confirming that the attenuation originates from the physical degradation of the return cable or connection point, enhancing the technical credibility of the second anomaly flag.

[0124] The above implementation achieves early identification and reliable warning of progressive degradation of the return cable line. It not only overcomes the limitations of traditional binary "on / off" judgments but also detects potential problems such as aging, loosening, and corrosion at connection points before the signal is completely interrupted, providing maintenance personnel with a valuable window of opportunity for intervention. Especially in unattended field scenarios, this mechanism effectively prevents the risk of minor defects gradually evolving into major accidents, significantly improving the predictive maintenance capabilities and overall safety of the return cable condition monitoring system.

[0125] Reference Figure 5 As one implementation of step S108, the step of generating a comprehensive anomaly decision code by fusing the first anomaly flag bit and the second anomaly flag bit includes:

[0126] Step S501: Obtain the current state value combination of the first abnormal flag bit and the second abnormal flag bit;

[0127] The first anomaly flag originates from the dual verification logic of device identity and communication timing, reflecting whether the slave device has lost connection or has a serious time deviation. Its essence is a manifestation of communication failure, such as complete signal interruption, device power failure, ID disorder, etc. The second anomaly flag is based on the trend analysis of the received signal strength (RSSI) of N consecutive times. It calculates the attenuation rate per unit time through linear fitting and combines dynamic threshold and continuous triggering mechanism to determine whether there is gradual performance degradation. Therefore, it is more of a warning signal of potential physical degradation risk, such as slow-developing defects such as loose connectors, oxidation of contact surfaces, and local corrosion.

[0128] Step S502: Match the current state value combination with the pre-stored anomaly decision rule table to extract the corresponding decision code; wherein, the anomaly decision rule table defines the mapping relationship between different flag bit combinations and decision codes;

[0129] The system combines these two Boolean state values ​​into a two-dimensional logical input pair (first exception flag, second exception flag), and uses this as an index key to match the exception decision rule table pre-stored in the host's local memory. This rule table is not static, fixed code, but rather a mapping logic system refined from engineering experience, clearly defining the decision meanings corresponding to four basic combinations, specifically including:

[0130] When both the first and second anomaly flags are true, it maps to the highest risk level code; when both the first and second anomaly flags are false, it maps to the device communication anomaly code; when both the first and second anomaly flags are false, it maps to the line attenuation anomaly code; when both the first and second anomaly flags are false, it maps to the normal operating status code.

[0131] Specifically, when both flags are false, it indicates that all slave devices are communicating normally with no obvious signal attenuation trend, and the system is in a stable operating state, which should be mapped to the "normal operation status code". If only the first abnormal flag is true, it indicates that a slave device is completely disconnected or has serious time synchronization issues, which is a typical equipment-level communication fault and should be classified as the "equipment communication abnormal code". If only the second abnormal flag is true, it indicates that although a signal can be received, its strength shows a continuous downward trend, which is very likely caused by line aging or poor contact, corresponding to the "line attenuation abnormal code". The most dangerous situation is that both are true at the same time, which often means that the fault has evolved from early deterioration to complete disconnection, and the system has entered a high-risk stage, which must be responded to immediately, and is therefore assigned the "highest risk level code".

[0132] Step S503: Encapsulate the decision code into a fixed-width binary field to generate a comprehensive anomaly decision code.

[0133] The system further encapsulates the extracted decision code into a fixed-width binary field, ultimately generating a unified comprehensive anomaly decision code. This design fully considers the resource constraints and communication efficiency requirements of embedded systems, using 4 bits as the standard length. Each bit independently represents a specific state category, and only one bit in any valid codeword is "1," with the rest being "0," forming the so-called "One-Hot Encoding" format. Specifically, "1000" represents the highest risk level, "0100" indicates device communication anomaly, "0010" indicates line attenuation anomaly, and "0001" indicates normal operation.

[0134] Understandably, the advantages of this encoding strategy are twofold: firstly, it avoids the ambiguous interpretations that may arise from traditional multi-base encoding (e.g., "1100" might be misinterpreted as a compound fault); secondly, it greatly simplifies the parsing logic at the receiving end, requiring only the detection of which position is "1" to quickly identify the current state type, making it suitable for low-power MCUs and narrowband long-distance transmission scenarios. More importantly, this fixed format facilitates subsequent packaging and uploading with other telemetry data to the monitoring platform, supporting automated alarm classification, event archiving, and linkage strategy invocation.

[0135] Furthermore, to further ensure the reliability of the decision code, a crucial decision code verification mechanism is executed after encapsulation. The system actively calculates the Hamming weight (i.e., the number of "1"s) of the binary field. If the result is not equal to 1, an error is determined to have occurred in the encoding process. This situation may stem from underlying anomalies such as memory overflow, pointer out-of-bounds errors, rule table loading failures, or concurrent access conflicts. Once such a problem is detected, the system will trigger an error handling process, reacquire the original flag bit state, and restart the entire decision code generation process until a valid codeword is output. This closed-loop self-checking mechanism significantly enhances the system's fault tolerance and data integrity assurance level, preventing false alarms or missed alarms due to internal software failures.

[0136] The above implementation achieves a leap from dispersed and heterogeneous monitoring signals to standardized state summaries. This not only effectively integrates multi-dimensional diagnostic information from the communication and physical layers but also ensures the uniqueness, interpretability, and high reliability of the output results through carefully designed coding rules and verification mechanisms. Particularly noteworthy is that this mechanism endows the host with a certain degree of intelligent judgment capability, transforming it from a mere data forwarding node into an edge computing unit with basic inference capabilities. This not only reduces reliance on the real-time processing capabilities of the central platform but also improves the overall system's response speed and robustness, providing support for the safe and stable operation of high-voltage cable lines.

[0137] Reference Figure 6 As one implementation of step S109, the step of determining the fault segment identifier by querying the logical association in the device topology table based on the comprehensive anomaly decision code and the device ID of the slave device includes:

[0138] Step S601: Obtain the device ID and corresponding comprehensive anomaly decision code of the slave device that failed to match;

[0139] The comprehensive anomaly decision code is a structured state summary generated locally by the host, encapsulating the joint judgment result of the first anomaly flag (identity / timing anomaly) and the second anomaly flag (signal attenuation trend anomaly). It reflects the severity and type of the current fault, such as whether it is the highest risk level or whether it is merely a communication interruption. The slave device ID that failed to match is the direct source triggering this location process, representing the specific object whose signal was lost. These two parameters together constitute the basic entry conditions for all subsequent analyses, ensuring that the location behavior has a clear target and contextual relevance.

[0140] Step S602: Query the direct host ID corresponding to the device ID according to the device topology table;

[0141] The device topology table is a logical mapping data structure generated after master-slave pairing is completed during system initialization. It records the affiliation between each slave device and its controlling master, i.e., "who is responsible for monitoring whom." For example, if slave device BS3 reports a loss of connection, the system first identifies its direct master as BM (B-area master), thus narrowing the analysis scope to the monitoring units under BM's jurisdiction. This step is essentially a spatial coordinate mapping operation, transforming physical device failures into logical node problems within the information system, providing anchor points for subsequent topology reasoning.

[0142] Step S603: Based on the logical association, retrieve the set of neighbor device IDs of the adjacent segment bound to the direct host ID;

[0143] Unlike traditional one-way monitoring, this system requires each host to not only manage multiple slave devices under its own control, but also to explicitly bind to the nearest device in an adjacent segment during the installation and configuration phase (such as the last slave device in the neighboring segment or the host edge interface), forming a logical link across segments. For example, in addition to managing BS1~BS4, host BM also binds to the AS-Edge at the end of area A and the CS-Edge at the beginning of area C, forming a front-to-back interconnected topology. This "front-to-back" networking method breaks down information silos between monitoring units, enabling a single host not only to grasp the status of its own segment, but also to indirectly perceive connectivity changes in neighboring areas, providing a necessary prerequisite for boundary fault diagnosis.

[0144] Step S604: Send a status verification request to each neighboring device in the neighboring device ID set;

[0145] The status verification request is transmitted via a power line carrier channel. Upon receiving it, the target device must return a response data packet containing its own operational status identifier. It is worth noting that in actual communication, factors such as electromagnetic interference, signal attenuation, or device sleep / wake-up delays may cause response delays or even packet loss. Therefore, before formal parsing, the system implements a strict response verification mechanism: the received data packet undergoes a CRC-16 check to ensure its integrity; if the check fails, it is determined to be a transmission error, and the verification request is immediately resent until the preset maximum number of retries (e.g., 3 times) is reached. If no valid response is ultimately obtained, the neighboring device is marked as "timeout," and its status is considered untrustworthy. This mechanism effectively eliminates misjudgments caused by instantaneous channel fluctuations, improving the reliability of the verification results.

[0146] Step S605: Receive response data packets from each neighboring device, parse the running status identifiers in them, and generate a neighbor verification status vector.

[0147] The neighbor verification status vector is a binary sequence, with each bit corresponding to the status of a neighboring device in the set: if the returned running status identifier indicates normal operation (such as "OK" or "RUNNING"), the corresponding bit is assigned a value of 0; if an abnormal status is returned (such as "FAULT" or "OFFLINE") or there is still no response after multiple retries (timeout), the value is assigned a value of 1. For example, if BM binds two neighbors AS-Edge and CS-Edge, and the former responds normally while the latter does not respond, the generated vector will be [0,1]. This vector is essentially a quantitative expression of "whether the neighboring cells have lost synchronous connection", which is the core basis for distinguishing between intra-cell faults and boundary faults.

[0148] Step S606: Input the integrated anomaly decision code and the neighbor verification state vector into the fault location decision tree, and output the fault segment identifier.

[0149] The fault location decision tree is used to perform classification reasoning and output a unique fault segment identifier. This decision tree is not a simple list of rules, but a hierarchical logical judgment model that leads to different conclusions based on different combinations of paths.

[0150] The final output fault segment identifier adopts a standardized fixed-length string structure, typically consisting of five fields: fault segment type identifier (1 byte, 'S' indicates a single segment, 'C' indicates a cross-segment), host ID field (4 bytes), connection symbol: " (1 byte), and boundary device ID field (4 bytes). This unified encoding format facilitates automatic parsing, data archiving, and visualization rendering by remote platforms, supporting rapid retrieval of map locations, linkage with video surveillance, or generation of work orders.

[0151] The above implementation achieves refined identification of return cable fault locations. This technical solution not only inherits the deployment flexibility advantages of distributed master-slave architecture, but also significantly improves the system's spatial resolution and diagnostic accuracy by introducing cross-segment logical association and bidirectional state verification mechanisms. Especially when facing complex line topologies and hidden faults, this method can complete high-confidence fault interval location using existing communication resources without increasing hardware costs, greatly shortening inspection and troubleshooting time and improving the level of intelligent operation and maintenance of the power grid.

[0152] As one implementation of the fault location decision tree, the specific mapping logic includes:

[0153] When the decision code value of the comprehensive anomaly decision code represents the highest risk level and all neighbor verification status bits are 0, a fault segment identifier combining the host ID and the abnormal slave ID is generated.

[0154] The comprehensive anomaly decision code is a multi-field encoding structure used to characterize the nature of the anomaly detected by the current host, such as whether it belongs to the highest risk level, whether it is only a communication timeout, or whether there is a continuous signal attenuation trend. The neighbor verification status bit value is a set of binary identifiers generated after actively probing adjacent segment bound devices. Each bit corresponds to the online status of a neighbor device: 1 indicates anomaly or no response, and 0 indicates normal response. These two dimensions together constitute the input variables of the decision tree, enabling the system not only to identify the existence of faults but also to accurately attribute their physical location.

[0155] When the decision code value of the comprehensive anomaly decision code represents the highest risk level and the neighbor verification status bit value is not all 0, a fault segment identifier is generated by combining the host ID and the ID of the first neighbor device with a status of 1.

[0156] Specifically, when the system determines that the comprehensive anomaly decision code represents the highest risk level, it means that the host has confirmed that a slave device's signal has been completely lost and accompanied by high-confidence anomalies such as strong interference and sudden interruptions. This type of situation usually points to serious physical damage. If all neighboring devices return normal responses at this time (i.e., all neighbor verification status bits are 0), it indicates that the communication link in the adjacent segment is intact, and the fault's impact is limited to the end of this segment. Therefore, it can be reasonably inferred that the fault point is located in the cable segment between the host and its directly affiliated abnormal slave device.

[0157] For example, if the host BM cannot receive signals from the slave BS4, but its bound downstream neighbor CS-Edge can still communicate normally, it indicates that the problem has not yet affected the next segment. The fault should occur between "BM:BS4". Therefore, a fault segment identifier composed of the host ID and the abnormal slave ID is generated to achieve accurate location of the sub-segment within the segment.

[0158] Conversely, if some neighboring devices also exhibit abnormal responses (i.e., not all status bits are 0) under this high-risk condition, it indicates that the fault has spread to the boundary area, most likely due to damage to the connection terminal at the junction of the two sections. In this case, the system will select the first neighboring device ID with a status of 1 to participate in the combination, outputting an identifier such as "BM:CS-Edge", clearly indicating that the fault direction is towards the next level section, thereby guiding maintenance personnel to prioritize checking the junction of the two sections rather than blindly checking the entire line.

[0159] When the decision code value of the comprehensive anomaly decision code indicates that the device communication is abnormal and all neighbor verification status bits are 0, a fault segment identifier combining the host ID and the abnormal slave ID is generated.

[0160] When the decision code value of the comprehensive anomaly decision code indicates that the device communication is abnormal and the neighbor verification status bit value is not all 0, a fault segment identifier is generated by combining the host ID and the boundary identifier; wherein, the boundary identifier indicates that the fault occurs at the connection boundary between the current host device and the adjacent segment device.

[0161] Specifically, when the comprehensive anomaly decision code only indicates "device communication anomaly" rather than the highest risk level, it means that the signal interruption may be caused by transient interference, polling failure, or device sleep delay, and does not have strong physical damage characteristics. In this case, if the neighbor verification results are all normal (all status bits are 0), it is still tended to be considered that the problem is localized, possibly due to a fault in the slave device's own module or poor coupling. Therefore, the fault identifier is still output using a combination of the host ID and the abnormal slave ID to maintain diagnostic consistency.

[0162] However, if neighboring devices are also found to be unresponsive (status bits not all set to 0), this cannot be simply attributed to a single point of failure. Instead, issues such as loose shared interfaces or corrosion of the common grounding busbar should be considered. These potential problems typically occur at segment boundaries. To address this, the system introduces a "boundary identifier" as a special marker, output in the format "BM:Edge," explicitly indicating that the fault occurs at the connection boundary between the current host-managed segment and adjacent segments, rather than between any specific device. This design avoids misjudgments due to insufficient information and improves diagnostic accuracy in complex scenarios.

[0163] When the decision code value of the comprehensive anomaly decision code indicates an abnormal line attenuation, a fault segment identifier is generated by combining the host ID and the abnormal slave ID.

[0164] Specifically, when the comprehensive anomaly decision code indicates "line attenuation anomaly," it reflects a gradual decline in signal strength rather than a sudden interruption. This situation is common in slowly developing aging problems such as connector oxidation and increased contact resistance. Since this type of fault has not yet caused communication interruption and usually occurs on the transmission path between the master and a specific slave, regardless of the status of the neighbor, it is prioritized to be located within the master-slave link, outputting the standard format "Master ID: Abnormal Slave ID" to facilitate early warning and planned maintenance.

[0165] When the decision code value of the comprehensive anomaly decision code indicates a normal operating state, an empty fault identifier is output.

[0166] When the comprehensive anomaly decision code displays "normal operation status", it means that all slave signals can be received normally without any abnormal signs. At this time, the system outputs an empty fault identifier, which saves storage resources and avoids generating invalid alarm records, ensuring the cleanliness and reliability of platform data.

[0167] In the above embodiments, by making full use of the existing topological association relationships and two-way communication capabilities in the distributed master-slave architecture, intelligent and refined identification of the fault section of the return cable is achieved. This technical solution not only breaks through the limitations of traditional monitoring systems, but also significantly improves the spatial resolution and diagnostic confidence of the system by integrating device-level status information and network-level topological relationships without the need to add additional sensors. Especially when facing complex cable lines with long distances, multiple sections, and field installations, this method can quickly lock in potential hazard areas, greatly shorten the inspection path and disposal time, and effectively support the proactive operation and maintenance and intelligent dispatching of the power grid.

[0168] Referring to Figure 7 , as a further embodiment of the online monitoring method for the return cable status, after the step of generating a status report and uploading it to the remote monitoring platform, it further includes:

[0169] Step S701, the remote monitoring platform receives the status report and extracts the fault section identifier and the comprehensive anomaly decision code therein;

[0170] Among them, when the master / slave system detects communication anomalies and completes the generation of the fault section identifier, it will send a structured data packet including the comprehensive anomaly decision code and the fault section identifier to the monitoring platform via a 4G or Ethernet channel. After receiving the report, the platform first decodes it and extracts two key semantic fields: one is the comprehensive anomaly decision code (such as the highest risk level, device communication anomaly, etc.) reflecting the severity and type of the fault, and the other is the fault section identifier indicating the specific physical location (such as "BM:BS4" or "CM:Edge"). These two parameters jointly constitute the trigger basis and context environment for all subsequent linkage actions, ensuring that the response actions have a high degree of target orientation and context adaptability.

[0171] Step S702, query the pre-configured linkage policy mapping table according to the comprehensive anomaly decision code to determine the target linkage device type and the corresponding operation parameters;

[0172] Among them, the linkage policy mapping table is a knowledge base established based on engineering experience and operation and maintenance rules, which clearly defines the types of external devices to be activated for different anomaly types and their operation parameters. For example, when the decision code is "the highest risk level code", it means that sudden damage events such as fractures may occur. At this time, the system determines that the video monitoring system needs to be linked, and sets its zoom factor to 16x and the pan-tilt rotation angle to the direction of the fault point; if the decision code is "line attenuation anomaly", it may involve slowly developing problems of poor contact, and it is more suitable to use a fiber optic vibration measurement system to capture micro-vibration characteristics. Therefore, the sampling frequency is set to 2kHz and the vibration threshold is set to 0.5g.

[0173] Understandably, this decision code-driven strategy matching mechanism is essentially an event-driven control logic that enables different levels of anomalies to trigger detection methods that match their risk characteristics, thus avoiding resource waste or insufficient response.

[0174] Step S703: Based on the fault section identifier, search the spatial location database to obtain the coordinate set of associated monitoring devices that match the target linkage device type within the preset radius of the fault section, and generate a linkage instruction queue;

[0175] The system uses a known fault segment identifier as an index to access the platform's built-in spatial location database, retrieving associated monitoring devices within a preset radius around the fault area. This database records the geographic coordinates, coverage area, model parameters, and access methods of all auxiliary sensing devices (such as cameras and fiber optic vibration meters) deployed on-site. The default search radius is 500 meters, a value that takes into account both the density of power facility layout and the effective detection distance of sensing devices. Within this range, the system filters out a set of devices that match the current target linkage device type. For example, in the "highest risk level" scenario, only video surveillance devices are selected; in the "gradual degradation" scenario, fiber optic vibration meters with high-sensitivity vibration detection capabilities are prioritized. The final set of devices constitutes the coordinate set of the associated monitoring devices to be activated, and is encapsulated together with the target device type into a linkage instruction queue, serving as the basis for the next step of batch control.

[0176] Step S704: Send a device-specific control protocol packet to each associated monitoring device in the linkage command queue; wherein, the device-specific control protocol packet contains operating parameters and execution time window;

[0177] The device-specific control protocol packets are not general broadcast messages, but rather customized command sequences for each type of device, containing specific operating parameters and execution time windows. For example, for a dome camera, the protocol packet embeds the RTSP stream address, PTZ control commands, focus mode, and zoom ratio; for a distributed fiber optic vibration measurement system (DAS), it sets the laser pulse frequency, integration time, dynamic filtering window, and alarm threshold.

[0178] More importantly, the protocol packet also defines an execution time window (e.g., starting continuous data collection for 10 minutes within the next 30 seconds), ensuring that multiple devices respond synchronously in time and space, facilitating subsequent data fusion and analysis. This step demonstrates the system's unified scheduling capability for heterogeneous devices, supporting multiple protocol encapsulations such as MQTT, ONVIF, and private TCP, ensuring compatibility.

[0179] Step S705: Receive the real-time monitoring data stream returned by each associated monitoring device and extract the device status feature vector from it;

[0180] In some embodiments, video surveillance returns a sequence of H.264 encoded image frames, and the fiber optic vibration measurement system outputs a three-dimensional time-frequency-amplitude spectrum. After receiving this streaming data, the platform uses its built-in edge computing module or AI inference engine to extract device status feature vectors reflecting the device's operating status. For example, visual features such as image sharpness, number of moving targets, and rate of change of illumination are extracted from the video stream; dynamic features such as dominant frequency components, energy distribution entropy, and peak offset are extracted from the vibration signal. These feature vectors are highly abstract representations of the original data, compressing information volume while retaining key patterns that can be used for comparison.

[0181] Step S706: Perform similarity matching calculation between the device status feature vector and the preset benchmark feature template;

[0182] The baseline feature template is a standard model generated through long-term data collection and clustering under normal equipment conditions, representing an ideal reference state that is "undisturbed" and "anomaly-free". The matching algorithm can use methods such as cosine similarity, Mahalanobis distance, or deep learning embedding space distance to measure the degree of deviation between the current observation state and the standard state.

[0183] Step S707: When the similarity is lower than the preset similarity threshold, generate a device anomaly confidence score;

[0184] When the similarity score is below a preset similarity threshold (e.g., below 0.7), it indicates that the phenomenon perceived by the device deviates significantly from normal behavior, most likely due to genuine external disturbances rather than noise or false triggering. In this case, the system generates a device anomaly confidence score, which is typically a continuous variable between 0 and 1. A higher score indicates a greater likelihood that the currently linked device has detected abnormal behavior. For example, if the camera detects someone approaching a cable manhole cover, the confidence score can reach above 0.9; if only a slight tremor is caused by wind or grass, the confidence score is lower.

[0185] Step S708: Update the monitoring platform visualization interface and overlay the fault section topology map, the status heat map of the linked equipment, and the anomaly confidence score.

[0186] The system integrates the above analysis results into the monitoring platform's visualization interface, enabling the fusion and display of multi-dimensional information. The platform automatically loads the topology map of the faulty section, highlights abnormal link segments in red, and overlays a heat map of the participating devices around them—the darker the color, the higher the confidence level of the reported anomaly. Simultaneously, the system lists the status score, collected screenshots, or vibration spectrum curves of each device in the sidebar or pop-up window, allowing maintenance personnel to quickly assess the situation. Furthermore, the system can automatically generate event reports and push them to the mobile app or the dispatch center's large screen, supporting one-click dispatch or voice alerts.

[0187] The above implementation achieves intelligent and comprehensive response to abnormal events in the return cable. It not only overcomes the information limitations of traditional single monitoring methods but also, by introducing heterogeneous sensing modalities such as video and vibration, forms a collaborative diagnostic capability across systems and spaces. Particularly noteworthy is that this mechanism fully utilizes existing infrastructure (such as cameras along the line and fiber optic vibration measurement networks), significantly improving monitoring efficiency without requiring additional hardware. This enhances the power grid's response speed to emergencies and reduces the cost of manual inspections.

[0188] Reference Figure 8 In actual deployment, the overall system architecture of this application is divided into three logical layers: the perception layer, the transmission layer, and the application layer. The perception layer includes a coupling module 101, a power line carrier communication module 102, a host 103, and a slave 104, which are used to collect and preliminarily process return cable status information; the transmission layer consists of a communication module 105, which is responsible for remotely uploading field data to the central platform; the application layer includes a database server 106 and a monitoring platform 107, which realize data storage, visualization, alarm management, and operation and maintenance support functions.

[0189] The coupling module 101, as the core component for non-contact signal injection and reception, is essentially a high-frequency current transformer. Its magnetic core is made of a high-permeability, low-loss material and can be a closed or easily installed open structure. The primary winding is the return cable under test, which passes through the center of the magnetic core to form an effective electromagnetic coupling path. The secondary winding is formed by multiple turns of insulated enameled wire wound on the magnetic core and is connected to the power line carrier communication module 102 via a coaxial cable or shielded twisted pair. During signal transmission (slave side), the carrier characteristic signal output by the power line carrier communication module is excited by the secondary winding, generating an alternating magnetic field in the magnetic core. This induces a current signal of the same frequency in the primary return cable through electromagnetic induction, realizing the "injection" of the characteristic signal. During signal reception (host side 103), when the carrier characteristic signal transmitted on the return cable flows through the coupling module, it also generates an alternating magnetic field in the magnetic core, thereby inducing an electromotive force in the secondary winding. This electromotive force is received and demodulated by the power line carrier communication module to restore the original data. Preferably, the coupling module is designed as a split-type clip structure, consisting of two symmetrical "C"-shaped half-body magnetic cores, encapsulated in a high-strength plastic shell. During installation, it is closed and fixed to the ground wire of the grounding box through a locking mechanism to form a complete magnetic circuit. Installation can be completed without damaging the original protective structure, which significantly improves the convenience and safety of construction.

[0190] The power line carrier communication module 102 serves as a signal conversion bridge between the master / slave unit and the coupling module. It primarily consists of a master control modem chip, a transmitting channel, a receiving channel, a coupler interface circuit, and a digital interface. The master control modem chip integrates a DSP, ADC, DAC, and line driver, responsible for executing the underlying communication protocol and processing signals. The transmitting channel includes a power amplifier and a transmitting bandpass filter; the former enhances the signal output power to ensure sufficient coupling energy, while the latter suppresses out-of-band spurious signals and reduces electromagnetic interference. The receiving channel includes a receiving bandpass filter and a controllable gain amplifier; the former selects the effective frequency band signal and suppresses noise, while the latter automatically adjusts the gain according to signal strength to adapt to attenuation changes caused by long-distance transmission. The coupler interface circuit includes passive components such as an impedance matching transformer and isolation capacitors, ensuring efficient signal transmission while achieving high-voltage isolation. The digital interface adopts the UART or SPI standard to achieve data interaction with the master / slave master control MCU. This module operates in the 0.7MHz to 12MHz frequency band and features adaptive bandwidth and modulation mode switching capabilities. Depending on the channel signal-to-noise ratio (SNR), it can dynamically select between three bandwidths: 10MHz, 3.2MHz, and 2.2MHz. High bandwidth is used to improve communication speed when channel conditions are good, while switching to low bandwidth ensures communication reliability when interference is severe. It also supports automatic switching between multiple modulation modes such as BPSK, QPSK, and 16-QAM, comprehensively considering anti-interference performance and data rate to achieve optimal link quality matching. The power line carrier communication module at the host 103 wakes up at a preset period (e.g., every 5 minutes), first performing multi-subband channel listening, assessing the noise level, and then selecting the optimal channel to broadcast a polling command. Upon receiving the corresponding command, the slave 104 adaptively configures parameters based on the current channel status and sends back a characteristic signal containing its unique ID. The host 103 receives, demodulates, and decodes this signal, completing one communication cycle.

[0191] The master and slave units together form a distributed topological monitoring network. While highly homogeneous in hardware, they play different roles in the system logic. Its core components include a master control MCU, power management circuitry, power line communication module interface, remote communication module interface, RS485 interface, DI input, auxiliary power output, status indicator lights, real-time clock (RTC), and non-volatile memory. The master control MCU is responsible for overall device control, data processing, protocol operation, and power scheduling, possessing abundant I / O resources and multiple low-power modes. The power management circuit receives external power input (e.g., DC 12V), converts it to the system's required stable voltage (e.g., 3.3V, 5V), and supports both main power input (connected to a battery) and auxiliary power input (connected to a photovoltaic panel or inductive power supply). It has voltage sampling capabilities, enabling real-time monitoring of power status and remote reporting. The remote communication module interface reserves a Mini-PCIe or USB interface for connecting the communication module 105. At least one RS485 interface supports protocols such as Modbus and can be used to connect external sensors or... Other intelligent devices; at least one DI input, supporting passive dry contact signal access, which can be used to receive external status quantities such as access control switches and vibration sensors, and has the function of triggering device wake-up; auxiliary power output provides one DC12V output to power small external devices; status indicator lights are equipped with two-color lights for running status (red) and 4G communication status (green) to facilitate on-site diagnosis; RTC has a built-in high-precision clock, supports timestamp marking and timed wake-up / sleep control, and can be remotely synchronized through the base station NITZ or monitoring platform; non-volatile memory uses Nor-Flash chip to save device parameters, operation logs and temporary cache data to prevent data loss when power is off.

[0192] Slave device 104 is primarily responsible for generating and injecting characteristic signals. It spends most of its time in deep sleep to minimize power consumption. It is only awakened when it receives a polling command from master device 103 or is triggered by an internal timer. The master MCU controls the power line carrier communication module to generate a characteristic signal carrying a unique device ID, which is then injected into the return cable via the coupling module. After completion, it immediately returns to sleep mode. Master device 103 is responsible for signal monitoring, status judgment, data aggregation, and remote communication. After being awakened by a configurable period or a DI signal trigger, it first controls the power line carrier communication module to perform channel listening, and then sequentially sends polling commands to all slave devices 104 bound to it, listening for and attempting to receive response signals. The master MCU executes a status judgment algorithm to analyze whether a complete and correct characteristic signal from a slave device 104 has been successfully received within a specified time. If the reception is successful, the cable segment connection between the slave device 104 and master device 103 is determined to be normal. If no signal is received within the timeout period or decoding fails, a connection anomaly (such as a broken wire) is determined in that segment. The host 103 packages the status results of all slaves 104, its own power supply voltage, DI status and other information into a data frame, starts the communication module 105, and uploads it to the monitoring platform 107 via 4G or Ethernet. After the task is completed, it enters sleep mode.

[0193] The system supports flexible topology deployment. Each monitoring unit consists of one host 103 and up to four slaves 104, managing a 500-1000 meter return cable section. During field installation, the host 103 is configured to bind its directly subordinate slaves (up to four) and the nearest neighbor device in the adjacent section (up to two), thus achieving cross-section logical association.

[0194] Reference Figure 9 For example, the host BM configures four slave devices BS1 to BS4 within its segment B, and associates them with the edge devices AS-Edge and CS-Edge in segments A and C, forming a complete cable topology. After configuration, the host BM uploads the final device topology table to the database server 106 for persistent storage via the communication module. Based on this, the monitoring platform automatically generates or updates a visual GIS topology map, intuitively presenting the device layout and connection relationships, achieving a precise mapping from "physical devices" to "logical topology".

[0195] The fault location mechanism relies on the above topology. Under normal conditions, the master unit (BM) should be able to stably receive carrier signals from all its directly subordinate slave units (BS1~BS4) and neighboring devices (AS-Edge, CS-Edge). When an anomaly occurs:

[0196] Scenario 1 (Internal Fault): If the master BM can only receive signals from BS1 and BS2, but cannot receive signals from BS3 and BS4, but can receive CS-Edge signals normally, then the fault point can be determined to be located in the sub-segment between BS2 and BS3 of the slave.

[0197] Scenario 2 (Boundary Fault): If the master BM cannot receive signals from its direct slave BS4 or its neighbor CS-Edge, while the master in segment C reports that CS-Edge is working normally, then the fault can be determined to occur at the segment boundary between BS4 and CS-Edge.

[0198] The communication module 105 primarily uses a 4G LTE module, supporting full-band LTE FDD (Band 1 / 3 / 5 / 8) and LTE TDD (Band 34 / 38 / 39 / 40 / 41) to ensure compatibility and coverage. It also includes an Ethernet controller as a backup communication solution, providing an RJ45 interface and supporting 10 / 100M adaptive connectivity. The SIM card slot is compatible with standard SIM cards and features anti-drop and electrostatic discharge (ESD) protection. The antenna interface uses a standard SMA-K interface, supporting external 4G antennas to enhance communication quality in weak signal areas. The main processor interface connects to the host MCU via Mini-PCIe or USB 2.0 for high-speed data exchange. The firmware layer is pre-configured or supports configuration of various commonly used IoT and power system communication protocols such as MQTT, HTTP, IEC 104, and DLT698.45. Users can select the appropriate protocol based on their monitoring platform requirements, and custom development and programming of special protocols are also supported.

[0199] Database server 106, serving as the system's data hub, is deployed at the backend of the monitoring platform, responsible for the persistent storage and efficient access services of massive amounts of real-time and historical data. An enterprise-grade high-availability architecture is recommended, with compute and storage nodes forming a cluster of at least two high-performance servers, running in master-slave or load-balanced mode. Each server should be configured with a multi-core CPU and large-capacity ECC memory. The storage system should employ RAID 10 or RAID 5 redundant arrays to prevent data loss due to single-disk failure. For large-scale deployments, it can be expanded to a NAS or SAN architecture. For networking, gigabit or higher interfaces should be provided, and a dedicated line or VPN connection should be established with the communication module to ensure secure and stable data transmission. The database software employs a hybrid architecture of relational and time-series databases: the relational database (such as MySQL or PostgreSQL) stores metadata, including device information tables (containing device ID, segment, location, topology binding relationships, etc.), system topology tables, user management tables, alarm configuration tables, etc.; the time-series database (such as InfluxDB or TDengine) is dedicated to efficiently storing time-series data with timestamps, mainly including device status data (signal reception status, power voltage, network strength, etc. reported by each host 103) and alarm event records (including alarm time, device, type, location segment, recovery time, etc.). The server runs a data receiving service, listening on a designated port. After performing security verification, format parsing, and decryption on the received data packets, it disassembles and writes them to the corresponding databases, and provides a unified API interface for application layer calls, achieving separation of business logic and data.

[0200] The monitoring platform 107 serves as the human-machine interaction center, employing a B / S architecture. Users can access it directly via a browser or mobile device without installing client software. The platform software is divided into a presentation layer, a business logic layer, and a data access layer: the presentation layer is a web front-end interface displaying real-time status and operation entry points; the business logic layer is deployed on a web server, implementing functions such as device communication management, a data analysis engine, alarm triggering logic, user authentication, and access control; the data access layer is responsible for calling the APIs provided by the database server 106 to complete data read and write operations. The platform can be deployed in a local private cloud or public cloud environment, offering excellent deployment flexibility. The core interface is a GIS-based system topology map, accurately marking the geographical locations of all master / slave devices and dynamically displaying the status of each device and cable segment using color coding (green—normal, red—abnormal, gray—offline). Clicking on any device icon displays a detailed panel showing the latest power voltage, signal strength, neighbor status, and other key parameters. The platform supports rule-based alarm policy configuration, such as triggering an alarm only when "no signal is received for two consecutive monitoring cycles," to reduce false alarms. Alarms are managed in tiers (urgent, important, warning), and relevant personnel are notified via SMS, email, and app push notifications. It provides alarm confirmation, work order dispatch, processing feedback, and archiving processes, forming a closed-loop operation and maintenance management system. Furthermore, it supports remote software upgrades, monitoring cycle adjustments, and alarm threshold modifications for the host (103), comprehensively improving operational efficiency.

[0201] In summary, this application utilizes non-contact coupling technology combined with power line carrier communication, employing existing return cables as communication channels. This avoids the need for additional dedicated lines, does not disrupt the original installation environment, and offers simple installation with high security. The distributed master-slave architecture and flexible topology configuration enable segmented monitoring of long-distance return cables and precise location of faulty sections, significantly improving operation and maintenance efficiency. The system supports multiple remote communication protocols and intelligent linkage mechanisms, achieving remote data monitoring, visualization, and automated alarms, significantly enhancing the intelligent operation and maintenance level of the power grid. The power supply scheme is flexible and diverse, supporting various modes such as battery, photovoltaic, and induction power generation, suitable for outdoor environments without mains power supply. This application can be widely used for status monitoring of return cables in high-voltage cable lines, effectively ensuring the stable operation of transmission lines and improving the reliability of power supply in the power system.

[0202] This application also discloses an online monitoring system for the status of return cables with a distributed master-slave architecture.

[0203] A distributed master-slave architecture online monitoring system for return cable status, the monitoring system includes:

[0204] The topology table acquisition module is used to receive a pre-configured device topology table; wherein, the topology table contains a list of directly subordinate slave devices bound to the master device and the logical association between the master device and neighboring devices in adjacent segments;

[0205] The slave control module is used to control the slave devices in the direct slave list to generate an initial characteristic signal carrying the device ID and the current timestamp;

[0206] The communication configuration module is used to scan the current channel status parameters of the return cable and dynamically select the communication bandwidth and modulation method based on the signal-to-noise ratio of the channel status parameters.

[0207] The signal modulation module is used to modulate the initial characteristic signal based on the selected communication bandwidth and modulation method, generate the carrier characteristic signal, and inject it into the return cable through a non-contact coupler;

[0208] The host control module is used to control the host device to receive the carrier characteristic signal on the return cable through the non-contact coupler, and to extract the device ID and current timestamp from the carrier characteristic signal;

[0209] The first anomaly detection module is used to generate a first anomaly flag based on the matching result between the device ID and the current timestamp and the device topology relationship table;

[0210] The second anomaly detection module is used to calculate the signal strength attenuation rate of continuously received carrier characteristic signals and generate a second anomaly flag bit based on the comparison result of the signal strength attenuation rate and the preset attenuation threshold.

[0211] The comprehensive anomaly decision module is used to merge the first anomaly flag bit and the second anomaly flag bit to generate a comprehensive anomaly decision code.

[0212] The fault segment determination module is used to query the logical relationship in the device topology table based on the comprehensive anomaly decision code and the device ID of the slave device to determine the fault segment identifier.

[0213] The status report upload module is used to generate a status report and upload it to the remote monitoring platform based on the comprehensive anomaly decision code and fault section identifier.

[0214] The distributed master-slave architecture online monitoring system for return cable status according to an embodiment of this application can implement any of the above methods, and the specific working process of each module in the system can refer to the corresponding process in the above method embodiments.

[0215] In the several embodiments provided in this application, it should be understood that the provided methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for example, the division of a certain module is merely a logical functional division, and in actual implementation there may be other division methods, such as multiple modules can be combined or integrated into another system, or some features can be ignored or not executed.

[0216] In this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0217] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. A single processor or other unit may implement several of the functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0218] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A method for online monitoring of the status of the return cable in a distributed master-slave architecture, characterized in that, The monitoring method includes: Receive a pre-configured device topology table; wherein the topology table includes a list of directly subordinate slave devices bound to the master device and the logical association between the master device and neighboring devices in adjacent segments; Control the slave devices in the direct slave list to generate an initial feature signal carrying the device ID and the current timestamp; Scan the current channel status parameters of the return cable, and dynamically select the communication bandwidth and modulation method based on the signal-to-noise ratio of the channel status parameters; The initial characteristic signal is modulated based on the selected communication bandwidth and modulation method to generate a carrier characteristic signal, which is then injected into the return cable through a non-contact coupler. The control host device receives the carrier characteristic signal on the return cable through a non-contact coupler and extracts the device ID and current timestamp from the carrier characteristic signal; Based on the matching result between the device ID and the current timestamp and the device topology relationship table, a first anomaly flag is generated; Calculate the signal strength attenuation rate of the continuously received carrier characteristic signal, and generate a second abnormal flag bit based on the comparison result of the signal strength attenuation rate and the preset attenuation threshold; The first and second anomaly flags are combined to generate a comprehensive anomaly decision code; Based on the comprehensive anomaly decision code and the device ID of the slave device, query the logical association in the device topology table to determine the fault segment identifier; Based on the comprehensive anomaly decision code and fault section identifier, a status report is generated and uploaded to the remote monitoring platform; The step of determining the fault segment identifier by querying the logical association relationship in the device topology table based on the comprehensive anomaly decision code and the device ID of the slave device includes: Obtain the device ID and corresponding comprehensive exception decision code of the slave device that failed to match; Query the direct host ID corresponding to the device ID according to the device topology table; Based on the logical association, retrieve the set of neighbor device IDs of the adjacent segment bound to the direct host ID; Send a status verification request to each neighbor device in the set of neighbor device IDs; Receive response data packets from each neighboring device, parse the running status identifiers within them, and generate a neighbor verification status vector; The integrated anomaly decision code and the neighbor verification state vector are input into the fault location decision tree, and the fault segment identifier is output.

2. The method for online monitoring of the return cable status in a distributed master-slave architecture according to claim 1, characterized in that, The steps of scanning the current channel state parameters of the return cable and dynamically selecting the communication bandwidth and modulation method based on the signal-to-noise ratio of the channel state parameters include: Within a preset communication frequency band, continuous subcarriers are divided, and the current channel state parameters of each subcarrier are collected in real time; wherein, the current channel state parameters include background noise power and instantaneous signal-to-noise ratio; Based on the background noise power, a set of candidate subcarriers that meet a preset noise threshold is selected; Calculate the signal-to-noise ratio gradient change rate of adjacent subcarriers in the candidate subcarrier set, and merge adjacent subcarriers with continuous signal-to-noise ratio gradient change rates to form a candidate bandwidth interval; Based on the total bandwidth and average signal-to-noise ratio of the candidate bandwidth range, the target communication bandwidth and target modulation method are dynamically selected by referring to the preset bandwidth-modulation method mapping table. Configure the carrier signal modulation parameters according to the selected target communication bandwidth and target modulation method.

3. The method for online monitoring of the return cable status in a distributed master-slave architecture according to claim 1, characterized in that, The step of generating the first anomaly flag based on the matching result between the device ID and the current timestamp and the device topology table includes: Obtain the set of slave device IDs from the list of directly subordinate slave devices bound in the device topology table; Parse the device ID and current timestamp from the carrier characteristic signal; Match the parsed device ID with the set of slave device IDs; Determine whether the match is successful based on the matching result; if the match fails, output the first exception flag as true; if the match is successful, calculate the absolute time difference between the parsed current timestamp and the host system time. Determine whether the absolute time difference exceeds a preset time tolerance threshold. If yes, output the first abnormal flag as true; otherwise, output the first abnormal flag as false.

4. The method for online monitoring of the return cable status in a distributed master-slave architecture according to claim 3, characterized in that, The step of calculating the signal strength attenuation rate of continuously received carrier characteristic signals and generating a second anomaly flag based on the comparison result of the signal strength attenuation rate and a preset attenuation threshold includes: Obtain the received signal strength values ​​of the carrier characteristic signal of the same slave device received N consecutive times, and generate a signal strength sequence; Using the sampling time point as the independent variable and the signal strength value as the dependent variable, the signal strength sequence is linearly fitted to calculate the signal strength attenuation rate per unit time. The signal strength attenuation rate is compared with a preset attenuation threshold. If the signal strength attenuation rate is greater than the preset attenuation threshold, a line attenuation anomaly flag is generated. The number of times the line attenuation anomaly flag is triggered in N consecutive detections is counted. When the number of triggers reaches a preset consecutive trigger threshold, the second anomaly flag is output as true.

5. The method for online monitoring of the return cable status in a distributed master-slave architecture according to claim 4, characterized in that, The steps for generating a comprehensive anomaly decision code by fusing the first anomaly flag bit and the second anomaly flag bit include: Obtain the current state value combination of the first anomaly flag and the second anomaly flag; The current state value combination is matched with a pre-stored anomaly decision rule table to extract the corresponding decision code; wherein, the anomaly decision rule table defines the mapping relationship between different flag bit combinations and decision codes; The decision code is encapsulated into a fixed-width binary field to generate a comprehensive anomaly decision code.

6. The method for online monitoring of the return cable status in a distributed master-slave architecture according to claim 5, characterized in that, The anomaly decision rule table includes the following mapping relationships: When both the first and second anomaly flags are true, the code is mapped to the highest risk level. When the first exception flag is true and the second exception flag is false, it is mapped to a device communication exception code; When the first abnormal flag is false and the second abnormal flag is true, it is mapped to a line attenuation abnormal code; When the first exception flag is false and the second exception flag is false, it is mapped to the normal operating status code.

7. The method for online monitoring of the return cable status in a distributed master-slave architecture according to claim 1, characterized in that, The mapping logic of the fault location decision tree includes: When the decision code value of the comprehensive anomaly decision code represents the highest risk level and all neighbor verification status bits are 0, a fault segment identifier combining the host ID and the abnormal slave ID is generated. When the decision code value of the comprehensive anomaly decision code represents the highest risk level and the neighbor verification status bit value is not all 0, a fault segment identifier is generated by combining the host ID and the ID of the first neighbor device with a status of 1. When the decision code value of the comprehensive anomaly decision code indicates that the device communication is abnormal and all neighbor verification status bits are 0, a fault segment identifier combining the host ID and the abnormal slave ID is generated. When the decision code value of the comprehensive anomaly decision code indicates that the device communication is abnormal and the neighbor verification status bit value is not all 0, a fault segment identifier combining the host ID and the boundary identifier is generated. When the decision code value of the comprehensive anomaly decision code indicates an abnormal line attenuation, a fault segment identifier combining the host ID and the abnormal slave ID is generated. When the decision code value of the comprehensive anomaly decision code indicates a normal operating state, an empty fault identifier is output.

8. A method for online monitoring of the return cable status in a distributed master-slave architecture according to any one of claims 1 to 7, characterized in that, After generating a status report and uploading it to the remote monitoring platform, the following steps are also included: The remote monitoring platform receives the status report and extracts the fault segment identifier and comprehensive anomaly decision code from it; The target linkage device type and corresponding operation parameters are determined by querying the pre-configured linkage strategy mapping table based on the comprehensive anomaly decision code. Based on the fault section identifier, the spatial location database is retrieved to obtain the coordinate set of associated monitoring devices that conform to the target linkage device type within the preset radius of the fault section, and a linkage instruction queue is generated. Send a device-specific control protocol packet to each associated monitoring device in the linkage instruction queue; wherein, the device-specific control protocol packet contains operation parameters and execution time window; Receive real-time monitoring data streams returned by each associated monitoring device and extract the device status feature vectors from them; The similarity matching calculation is performed between the device state feature vector and the preset benchmark feature template; When the similarity is lower than the preset similarity threshold, an anomaly confidence score is generated. Update the monitoring platform's visualization interface to overlay and display the fault section topology map, the status heat map of linked equipment, and the anomaly confidence score.

9. A distributed master-slave architecture online monitoring system for return cable status, characterized in that, The monitoring system includes: The topology table acquisition module is used to receive a pre-configured device topology table; wherein, the topology table includes a list of directly subordinate slave devices bound to the master device and the logical association between the master device and neighboring devices in adjacent segments; The slave control module is used to control the slave devices in the direct slave list to generate an initial feature signal carrying the device ID and the current timestamp; The communication configuration module is used to scan the current channel status parameters of the return cable and dynamically select the communication bandwidth and modulation method according to the signal-to-noise ratio of the channel status parameters. The signal modulation module is used to modulate the initial characteristic signal based on the selected communication bandwidth and modulation method, generate a carrier characteristic signal, and inject it into the return cable through a non-contact coupler; The host control module is used to control the host device to receive the carrier characteristic signal on the return cable through a non-contact coupler, and to extract the device ID and current timestamp from the carrier characteristic signal; The first anomaly detection module is used to generate a first anomaly flag based on the matching result between the device ID and the current timestamp and the device topology relationship table; The second anomaly detection module is used to calculate the signal strength attenuation rate of the continuously received carrier characteristic signal, and generate a second anomaly flag bit based on the comparison result of the signal strength attenuation rate and the preset attenuation threshold. The comprehensive anomaly decision module is used to merge the first anomaly flag bit and the second anomaly flag bit to generate a comprehensive anomaly decision code. The fault segment determination module is used to query the logical association relationship in the device topology table based on the comprehensive anomaly decision code and the device ID of the slave device to determine the fault segment identifier. The status report upload module is used to generate a status report and upload it to the remote monitoring platform based on the comprehensive anomaly decision code and the fault section identifier. The fault section determination module is configured as follows: Obtain the device ID and corresponding comprehensive exception decision code of the slave device that failed to match; Query the direct host ID corresponding to the device ID according to the device topology table; Based on the logical association, retrieve the set of neighbor device IDs of the adjacent segment bound to the direct host ID; Send a status verification request to each neighbor device in the set of neighbor device IDs; Receive response data packets from each neighboring device, parse the running status identifiers within them, and generate a neighbor verification status vector; The integrated anomaly decision code and the neighbor verification state vector are input into the fault location decision tree, and the fault segment identifier is output.

Images