Ground wire full life cycle management method

Abstract

The invention discloses a ground wire full life cycle management method, relates to the technical field of power equipment asset management, and aims to solve the problems that manual recording is easy to tamper, cloud diagnosis delay is high and an alarm mode is single and easy to neglect in traditional ground wire management. According to the method, operation fingerprints and spatio-temporal information during grounding wire hooking are recorded through the block chain technology, and it is ensured that data cannot be tampered; a built-in lightweight AI model of the edge computing unit is used for analyzing sensor data in real time to carry out localized fault diagnosis; and triggering a multi-mode sound-light alarm controlled by the PWM signal according to the diagnosis result grade. The system realizes full-life-cycle credible traceability, sub-second fault response and high-recognition-rate alarm of the state of the grounding wire, and is suitable for intelligent operation and maintenance of the grounding wire in the field of a transformer substation and the like.

Description

Technical Field

[0001] This invention belongs to the field of power equipment asset management technology, and in particular relates to a method for the full life cycle management of grounding wires. Background Technology

[0002] In the field of power equipment asset management, grounding wires are critical equipment for ensuring operational safety, and the reliability and real-time performance of their lifecycle management are paramount. Currently, traditional grounding wire management methods have significant shortcomings in practical applications. First, usage records rely on manual registration; operators must manually fill in information such as connection time and location. This method is not only inefficient but also prone to omissions or alterations due to negligence or intentional acts, making it difficult to accurately trace historical data and hindering safe operation and maintenance and accident accountability. Second, in the fault diagnosis stage, existing methods mainly rely on cloud processing and analysis. Grounding wire monitoring data must be transmitted over the network to a remote server for judgment. Due to bandwidth limitations and long data processing links, response delays are high, failing to meet the real-time fault response requirements of substations and other on-site locations, potentially exacerbating accident risks in emergency situations. Furthermore, traditional alarms often use a single-mode audible and visual alert. In noisy, highly interference-prone industrial environments, these alarms are easily ignored by operators, resulting in abnormal conditions not being detected and addressed in a timely manner. Previous attempts to improve these issues have often faced the challenge of balancing data reliability, real-time diagnostics, and alarm effectiveness, and have lacked the technical means to integrate high-reliability diagnostics and reliable evidence storage locally. Therefore, there is an urgent need for a full lifecycle management system that can ensure the immutability of grounding wire operation data, rapid fault diagnosis response, and multi-mode reliable alarm capabilities. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention proposes a method for managing the entire lifecycle of grounding wires, thereby resolving the issues present in the prior art.

[0004] Firstly, to achieve the above objectives, the present invention provides a method for managing the entire lifecycle of a grounding wire, comprising the following steps:

[0005] Its status data is collected by sensors deployed at the grounding wire site;

[0006] Localized fault diagnosis is performed on the status data based on edge computing units;

[0007] When a fault is diagnosed, the fault characteristic data is generated into a hash value and uploaded to the blockchain network for evidence storage.

[0008] Based on the results of the fault diagnosis, the corresponding multi-mode alarm signal is triggered.

[0009] Optionally, the process of collecting status data through sensors deployed at the grounding wire site includes:

[0010] The pressure change generated when the grounding wire is connected is detected by a pressure sensor;

[0011] Alternatively, a current sensor can be used to detect changes in current generated when the grounding wire is connected to the circuit;

[0012] Using the pressure or current change as a trigger signal, a blockchain node containing time, location, and operator information is generated.

[0013] Optionally, the process of performing localized fault diagnosis on the status data based on the edge computing unit includes:

[0014] The collected data is preprocessed by filtering and normalization within the edge computing unit;

[0015] The preprocessed data is compared with a preset safety threshold to achieve initial fault screening.

[0016] For data that is initially screened as abnormal, it is input into a lightweight neural network model for deep inference to determine the fault type.

[0017] Optionally, the process of triggering the corresponding multi-mode alarm signal includes:

[0018] Based on the level of the fault diagnosis result, a PWM control signal containing a specific flashing frequency and a buzzer frequency is generated;

[0019] The PWM control signal is used to drive the LED light and buzzer to emit an audible and visual alarm signal corresponding to the fault level.

[0020] Optionally, the process further includes:

[0021] Remote configuration commands are issued to edge devices through the management platform;

[0022] The edge device parses and stores the instructions to dynamically update the mapping logic between alarm modes and vibration modes.

[0023] Optionally, after generating a hash value from the fault characteristic data and uploading it to the blockchain network for evidence storage, the process further includes:

[0024] Store failure case data in a blockchain case library;

[0025] The parameters of the lightweight neural network model are updated based on historical failure case data.

[0026] Secondly, the present invention also provides a grounding wire lifecycle management system for implementing a grounding wire lifecycle management method, the system comprising:

[0027] The status awareness module is used to collect status data of the grounding wire through sensors deployed at the grounding wire site;

[0028] An edge intelligent diagnostic module is used to perform localized fault diagnosis on the status data based on the edge computing unit.

[0029] The blockchain evidence storage module is used to generate hash values ​​from fault characteristic data and upload them to the blockchain network for evidence storage when a fault is diagnosed.

[0030] The multi-mode alarm module is used to trigger corresponding multi-mode alarm signals based on the results of the fault diagnosis.

[0031] Thirdly, the present invention also provides a computer terminal device, comprising:

[0032] One or more processors;

[0033] A memory, coupled to the processor, for storing one or more programs;

[0034] When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the grounding wire lifecycle management method in the first aspect above.

[0035] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of the grounding wire lifecycle management method described in the first aspect above.

[0036] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the grounding wire lifecycle management method described in the first aspect above.

[0037] Compared with the prior art, the present invention has the following advantages and technical effects:

[0038] This invention provides a method for full lifecycle management of grounding wires. It utilizes blockchain technology to ensure the immutability and traceability of grounding wire operation records and fault data, achieving trusted management throughout the entire lifecycle. Localized real-time fault diagnosis is performed using a lightweight artificial intelligence model built into the edge computing unit, significantly improving response speed. By defining multi-mode audio-visual coding rules and supporting remote dynamic configuration, high-recognition-rate alarms are achieved, adapting to complex field environments. The system combines fault case feedback and an online model update mechanism to continuously improve diagnostic accuracy. Simultaneously, a layered architecture and localized consensus ensure the system's basic operational capabilities under abnormal conditions such as network outages. Attached Figure Description

[0039] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0040] Figure 1 This is a flowchart illustrating the localized initial fault screening process according to an embodiment of the present invention.

[0041] Figure 2 This is a flowchart of the model deep reasoning process according to an embodiment of the present invention;

[0042] Figure 3 This is a flowchart illustrating the fault determination process according to an embodiment of the present invention. Detailed Implementation

[0043] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0044] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0045] Example 1

[0046] This embodiment provides a method for managing the entire lifecycle of a grounding wire, including:

[0047] Its status data is collected by sensors deployed at the grounding wire site;

[0048] Localized fault diagnosis is performed on the status data based on edge computing units;

[0049] When a fault is diagnosed, the fault characteristic data is generated into a hash value and uploaded to the blockchain network for evidence storage.

[0050] Based on the results of the fault diagnosis, the corresponding multi-mode alarm signal is triggered.

[0051] As one implementation method in this embodiment, the process of collecting status data by sensors deployed at the grounding wire site includes:

[0052] The pressure change generated when the grounding wire is connected is detected by a pressure sensor;

[0053] Alternatively, a current sensor can be used to detect changes in current generated when the grounding wire is connected to the circuit;

[0054] Using the pressure or current change as a trigger signal, a blockchain node containing time, location, and operator information is generated.

[0055] As one implementation method in this embodiment, the process of performing localized fault diagnosis on the status data based on the edge computing unit includes:

[0056] The collected data is preprocessed by filtering and normalization within the edge computing unit;

[0057] The preprocessed data is compared with a preset safety threshold to achieve initial fault screening.

[0058] For data that is initially screened as abnormal, it is input into a lightweight neural network model for deep inference to determine the fault type.

[0059] As one implementation method in this embodiment, the process of triggering the corresponding multi-mode alarm signal includes:

[0060] Based on the level of the fault diagnosis result, a PWM control signal containing a specific flashing frequency and a buzzer frequency is generated;

[0061] The PWM control signal is used to drive the LED light and buzzer to emit an audible and visual alarm signal corresponding to the fault level.

[0062] As one implementation method in this embodiment, the process further includes:

[0063] Remote configuration commands are issued to edge devices through the management platform;

[0064] The edge device parses and stores the instructions to dynamically update the mapping logic between alarm modes and vibration modes.

[0065] As one implementation method in this embodiment, after the process of generating hash values ​​from fault feature data and uploading them to the blockchain network for evidence storage, the method further includes:

[0066] Store failure case data in a blockchain case library;

[0067] The parameters of the lightweight neural network model are updated based on historical failure case data.

[0068] The core technical solution includes: Blockchain traceability.

[0069] Each time a blockchain node is connected, it records the time, location, and fingerprint information of the operator.

[0070] A pressure sensor is installed at the grounding wire connection point. When the pressure sensor detects a pressure change caused by the connection action, it automatically triggers the generation of a blockchain node. When the operator connects the grounding wire to the designated position, the pressure sensor senses the increase in pressure and sends a signal to the system to initiate the node generation process. A current change occurs the instant the grounding wire is connected to the circuit. This current change is detected by a current sensor. When a current signal matching the connection characteristics is detected, blockchain node generation is triggered. Normally, there is no current; after connection, a small current flows, and the current sensor recognizes this change and triggers the corresponding operation.

[0071] In addition to automatic triggering, the system also supports manual triggering. In special cases, such as sensor malfunction or when manual intervention is required, operators can manually input the connection command through a dedicated terminal, such as a smart bracelet or mobile app. Upon receiving the command, the system will generate a blockchain node.

[0072] Data encryption uses the SHA-256 algorithm and supports cross-platform auditing.

[0073] Edge intelligent diagnostics:

[0074] The terminal has a built-in AI chip (RISC-V architecture) that runs lightweight neural network models (such as MobileNet);

[0075] 1. Data normalization (feature scaling)

[0076] Pressure data: Min-max normalization is used to linearly map the original pressure values ​​(e.g., 0–100 MPa) to the [0,1] interval.

[0077]

[0078] To avoid model bias caused by differences in dimensions, and to suppress the influence of outliers.

[0079] Location data (e.g., GPS coordinates):

[0080] Convert latitude and longitude to relative position offset (with the installation base point as the origin);

[0081] Z-score normalization of the offset distance:

[0082]

[0083] Ensure the scale consistency between spatial characteristics and pressure data.

[0084] 2. Time series slices

[0085] Window partitioning: Divide the continuous data stream into segments of fixed duration (e.g., 5 minutes / segment), with each segment containing 300 sampling points (assuming a sampling rate of 1Hz);

[0086] Overlap sampling: Set a 50% overlap rate (adjacent segments overlap by 150 points) to prevent key features from being truncated;

[0087] Abnormal segment labeling: Add labels to segments with sudden pressure drops (suspected loosening) or positioning drift (displacement exceeding the threshold) for supervised training.

[0088] II. MobileNet Input Format Adaptation Method

[0089] 1. Data reconstruction into image format

[0090] The pressure-time series within a single window is converted into a grayscale image matrix (300×1 pixel), and the pressure values ​​are mapped to pixel brightness (0~255).

[0091] The location data is used to independently generate a two-dimensional heat map (50×50 pixels), and the coordinate offset is converted into pixel intensity distribution;

[0092] Dual-channel merging: forms a 300×50×2 input tensor (height×width×channel), which is adapted to the 2D convolutional structure of MobileNet.

[0093] 2. Lightweight model structure adjustment

[0094] Modify the first-layer convolution kernel: set the number of input channels to 2 (originally 3) to avoid redundant calculations;

[0095] Compressed classification layer: The number of output neurons is reduced from 1000 (ImageNet) to the number of fault types (such as loosening, corrosion, and displacement).

[0096] III. Optimized Deployment of Edge Terminals

[0097] 1. Preprocessing calculation of load allocation

[0098] Normalization and slicing are performed on the terminal MCU, with only the feature matrix being uploaded to the edge nodes;

[0099] Reduces data transmission volume by more than 60% (compared to the original streaming data).

[0100] 2. Model Quantization Compression

[0101] The trained FP32 model was converted to INT8 precision, reducing the model size to 1 / 4 and adapting to the memory limitations of RISC-V chips.

[0102] IV. Collaborative Verification with Blockchain Systems

[0103] 1. Preprocessing parameters are stored on the blockchain.

[0104] Normalized P min / P max Parameters such as positioning reference points are written into the smart contract to ensure consistency across all nodes;

[0105] 2. Edge diagnostic results are uploaded to the blockchain.

[0106] The failure probability and confidence level output by MobileNet are used to generate a hash value, which is stored in IPFS and anchored to the blockchain for tamper-proof and traceable protection.

[0107] Summary of technical advantages:

[0108] Through the above preprocessing steps, MobileNet achieves the following at the edge:

[0109] ①Low-latency diagnosis: Slicing + image processing meets real-time requirements (<500ms / inference);

[0110] ② High robustness: Normalization eliminates individual differences among sensors and improves generalization ability;

[0111] ③ Resource efficiency: The quantized model is less than 500KB, and the power consumption of the RISC-V chip is controlled within 1W.

[0112] Real-time analysis of pressure and location data allows for localized determination of fault types (such as loosening or displacement).

[0113] Fault determination logic flow:

[0114] Step 1: Real-time acquisition of multi-source sensor data (edge ​​layer);

[0115] Data collected: pressure sensor (monitors grounding wire clamping force), displacement sensor (detects horizontal / vertical offset), environmental sensor (temperature, humidity, wind speed);

[0116] Edge preprocessing: Perform data filtering and normalization at edge computing nodes to eliminate environmental noise interference;

[0117] Step 2: Localized initial fault screening (threshold judgment), such as... Figure 1 As shown;

[0118] Judgment criteria (example threshold):

[0119] Loosening fault: Pressure remains below the safe threshold (e.g., 5kN);

[0120] Offset fault: Position data deviates from the reference coordinates by more than 10cm;

[0121] Sensor anomaly: Data fluctuations exceeding the reasonable range (e.g., ±20%) trigger self-test;

[0122] Step 3: Deep inference of the model (edge ​​computing layer), such as Figure 2 As shown;

[0123] Core model:

[0124] 1. Pressure residual analysis:

[0125] ARIMA model trained on historical data triggers a fault when residuals exceed a threshold.

[0126] 2. Location trajectory clustering:

[0127] Identify anomalous displacement patterns (such as persistent drift) using K-means clustering;

[0128] 3. Multi-sensor fusion:

[0129] By combining the pressure-displacement correlation matrix, we can distinguish between real faults and environmental disturbances;

[0130] Step 4: Fault Classification and Blockchain Evidence Storage

[0131] The fault type output is shown in Table 1:

[0132] Table 1

[0133] Fault Codes type Judgment basis F001 Mechanical loosening Pressure residual error continues to exceed the standard + low frequency vibration F002 External force offset Location clustering anomalies + low correlation with wind speed F003 Sensor failure Conflicting data from multiple sources + self-check failure

[0134] Blockchain Evidence Preservation:

[0135] Edge nodes upload the fault characteristic hash value to the private chain;

[0136] The cloud platform verifies the authenticity of data through a public blockchain;

[0137] Step 5: Dynamic Optimization Mechanism

[0138] Feedback loop: Failure cases are automatically stored in the blockchain case library, driving model parameter updates;

[0139] Predictive maintenance: Based on historical failure chain analysis, a risk warning map is generated;

[0140] Fault diagnosis flowchart as follows Figure 3 As shown.

[0141] Technical advantages:

[0142] Edge-cloud collaboration efficiency

[0143] The localized fault diagnosis response time is less than 100ms (edge ​​computing), and blockchain is used to ensure that the diagnostic results are tamper-proof.

[0144] Adaptive diagnostic capability:

[0145] By using a dual-engine approach of residual analysis and pattern matching, it adapts to different grounding wire operating conditions, reducing the false alarm rate by more than 40%.

[0146] Full lifecycle traceability:

[0147] Fault records are written to the blockchain, supporting full traceability from production to disposal, which complies with asset management standards.

[0148] Multi-mode alarm system:

[0149] Define the audio-visual coding rules: emergency fault (red light flashing rapidly + 5kHz buzzer), general alarm (yellow light flashing slowly + 3kHz buzzer);

[0150] Audio-visual coding rule definition: Based on system requirements, the following two main alarm modes are defined with corresponding audio-visual coding rules: Emergency Fault: Light signal: Red LED flashes rapidly (e.g., on for 0.2 seconds, off for 0.2 seconds, period 0.4 seconds). Sound signal: Emits a continuous 5kHz buzzer. Meaning: Indicates a serious fault or safety hazard in the grounding wire, requiring immediate attention. General Alarm: Light signal: Yellow LED flashes slowly (e.g., on for 0.5 seconds, off for 0.5 seconds, period 1.0 second). Sound signal: Emits an intermittent 3kHz buzzer (e.g., on for 0.3 seconds, off for 0.7 seconds). Meaning: Indicates an abnormal grounding wire condition or the need for maintenance and inspection, but not an emergency. I. Implementation Carrier of Coding Rules: Based on the PWM signal generation of a microcontroller, the core implementation carrier of the audio-visual coding rules of this invention is a microcontroller unit (MCU) deployed on an edge computing node, which accurately generates the signals required to control the LED flashing frequency and the buzzer tone frequency through pulse width modulation (PWM) technology.

[0151] The core function of a microcontroller (MCU) is to run alarm logic processing programs and receive alarm level instructions (emergency faults or general alarms) from the system's intelligent diagnostic module (which may run at the edge layer or in the cloud). It configures the parameters of its built-in PWM generator module according to predefined encoding rules (such as the aforementioned flashing frequency, duty cycle, and tone frequency). It generates precise PWM waveform signals, which are output to the GPIO ports controlling the LED and the GPIO ports controlling the buzzer (or a dedicated PWM output port). Applications of Pulse Width Modulation (PWM) technology include: LED flashing control: The MCU controls the on / off state and brightness of the LED (if dimming is required) via PWM. For example, to achieve "fast red light flashing": The MCU generates a PWM signal with a period of 0.4 seconds (frequency 2.5Hz) and a duty cycle of 50% (0.2 seconds high, 0.2 seconds low) and outputs it to the red LED control terminal. To achieve a "slow yellow light flashing": The MCU generates a PWM signal with a period of 1.0 second (frequency 1Hz) and a duty cycle of 50% (0.5 seconds high level, 0.5 seconds low level) and outputs it to the yellow LED control terminal. The frequency of the PWM signal determines the flashing speed, and the duty cycle determines the on / off duration. Buzzer tone control: The MCU controls the buzzer's sound frequency (pitch) and sound mode (continuous or intermittent) via PWM. To achieve a "5kHz continuous buzzer": The MCU generates a 5kHz PWM signal (period 0.0002 seconds) (usually with a 50% duty cycle) and outputs it to the buzzer control terminal. To achieve a "3kHz intermittent buzzer": The MCU needs to be combined with a timer interrupt function. First, configure the PWM output frequency to 3kHz, then use another timer to control the enabling and disabling of the PWM signal to achieve an intermittent mode of 0.3 seconds of sound (PWM output enabled) and 0.7 seconds of silence (PWM output disabled). The frequency of the PWM signal determines the pitch of the buzzer. Accuracy and performance considerations (optional but recommended): For PWM signals requiring extremely high precision (such as certain special frequencies or extremely small duty cycles), the system clock can be phase-shifted using an FPGA chip or its internal phase-locked loop (PLL) to generate multiple auxiliary clocks at the same frequency, effectively increasing the clock frequency and thus producing a finer PWM waveform. This is an effective solution when it is necessary to overcome the limitations of the MCU's main frequency or pursue instrument-grade precision.

[0152] The customizable control logic capabilities of the edge computing layer enable the adjustment and updating of alarm rules (frequency, mode) to be configured remotely via software without modifying the hardware, thus improving the system's flexibility and maintainability.

[0153] Supports vibration motor-assisted prompts (mode is configurable).

[0154] Remote command configuration (management platform): Maintenance personnel issue JSON commands to blockchain nodes via the web management platform. After consensus verification, the edge device firmware is updated, dynamically switching vibration modes (e.g., strong vibration during device failure + weak vibration during high temperature warning). Configuration permissions are tiered by role: on-site personnel only have access to basic mode switching, while administrators can customize the mapping logic between diagnostic alarms and vibration.

[0155] Key technical implementation points: Bluetooth protocol: The APP uses the low-power GATT protocol to write configuration data, and the edge device parses the instructions through characteristics and stores them in the local EEPROM2. Mode flexibility: Supports scenario-based templates (e.g., "Maintenance Mode" = high-frequency vibration, "Standby Mode" = feedback off), reducing the risk of accidental operation. This design achieves a balance between user-friendliness and management efficiency through multi-level interaction, while relying on blockchain to ensure the integrity and traceability of configuration instructions.

[0156] Hardware design:

[0157] The AI ​​chip is integrated into the main control board, with a computing power of ≥1 TOPS;

[0158] Blockchain nodes are deployed on edge servers and support lightweight consensus algorithms (such as PoA).

[0159] PoA consensus algorithm adaptation and optimization strategies under edge server resources: PoA adaptation mechanism for edge resources: Node number limit: PoA only authorizes trusted nodes (such as edge gateways) to participate in consensus, reducing computational overhead. For example, the number of edge server nodes can be set to a finite set (such as 10-20) to avoid large-scale network communication pressure and adapt to low-bandwidth environments. Simplified verification mechanism: Only authoritative nodes verify transaction signatures, reducing computing power requirements (AI chips ≥1 TOPS can handle this efficiently). Resource constraints: Lightweight PoA abandons high-energy-consuming computation (such as PoW mining), utilizing local edge server resources to prioritize verification tasks, reducing blockchain latency and improving system response efficiency. Edge optimization strategies: Pre-authorized node list: System administrators maintain a dynamic authorization list, based on node identity authentication (such as digital certificates) or resource status (such as remaining CPU threshold ≥30%), only allowing high-reputation nodes to join the consensus. Dynamic load balancing: Combined with resource awareness mechanisms (such as monitoring edge node load), a leader node is dynamically elected to avoid single-point bottlenecks, and the resource reputation strategy of RAFT-RRLB consensus is referenced to improve utilization. Lightweight verification steps: Optimize transaction signature algorithms (such as ECDSA Simplified Version), reduce data bandwidth usage, and ensure low-latency operation in edge environments.

[0160] Algorithm Design:

[0161] The fault feature library contains 20 typical abnormal data patterns;

[0162] The construction method and process of the feature library for the grounding wire lifecycle management system based on blockchain and edge computing are as follows: I. Feature Library Construction Method: Multi-source data fusion modeling: Combining real-time monitoring data collected by edge computing nodes (such as current waveforms, temperature, and vibration spectra) with historical operation and maintenance data stored on the blockchain, a hybrid modeling approach is adopted: Expert rules define 12 types of physical characteristic anomalies (such as insulation layer damage characteristic waveforms and connection point overheating thresholds); Supervised learning generates 8 types of implicit correlation patterns (based on LSTM networks to mine anomaly correlations in time-series data); Layered implementation of the annotation process: Edge-side initial screening: Pre-annotating the original signal using a lightweight CNN model to extract suspected anomaly fragments; Blockchain verification: Uploading the hash values ​​of the annotated data to the blockchain, and cross-confirming by verification nodes composed of the operation and maintenance party and equipment manufacturers; Expert knowledge embedding: Based on smart contracts to solidify industry standards (such as DL / T 879-2024 Grounding Device Maintenance Specification), correcting AI misjudgment samples. II. Key Steps in the Generation Process (Supervised Learning Framework)

[0163] Data Acquisition Phase: Edge intelligent terminals are deployed to continuously collect grounding wire operating data at a 5kHz sampling rate. This data is then stored on the blockchain via IPFS sharding. Feature Engineering Phase: Wavelet packet decomposition is used to extract time-frequency domain features. Combined with Gram angle field analysis, the time-series data is transformed into a two-dimensional feature map. Model Training Phase: Based on a federated learning framework, a ResNet-18 model is trained distributed across multiple edge nodes. Model parameters are synchronized via the blockchain. Continuous Optimization Mechanism: An online learning module is deployed to automatically expand the training set after new anomaly cases are verified by consensus nodes. This solution ensures data traceability and label credibility through blockchain and achieves real-time feature extraction through edge computing. The feature library classification accuracy reaches 93.7% (actual measured data). Typical anomaly patterns cover three major categories: electrical performance degradation, mechanical structure damage, and environmental corrosion. Each category contains 4-7 sub-feature dimensions.

[0164] The audio-visual encoding drive circuit supports PWM frequency modulation and dimming.

[0165] The PWM parameters are mapped to alarm modes as follows: Duty cycle → Alarm intensity. The duty cycle directly controls the brightness / volume intensity of the audio-visual signal. For example: 80% duty cycle: high-brightness red light + high-frequency buzzer (emergency fault); 30% duty cycle: medium-brightness yellow light + intermittent buzzer (normal warning). Principle: A high duty cycle extends the power-on time, increasing energy output intensity. Frequency → Alarm rhythm. The frequency determines the flashing / buzzing frequency. The mapping logic is: 5Hz: rapid flashing (>3 times / second), used for emergency faults; 1Hz: slow flashing (1 time / second), used for low battery reminders. Principle: The higher the frequency, the shorter the cycle, and the more urgent the visual / auditory warning. This system collects grounding wire status data (such as current and temperature) in real time through edge computing nodes. After analysis by the intelligent diagnostic module 1, the corresponding alarm level is triggered. The audio-visual drive circuit uses a PWM programmable chip to dynamically adjust the duty cycle and frequency based on diagnostic results: Emergency fault (80% duty cycle + 5Hz): high-energy red light flashes rapidly to ensure alarm penetration; Low battery reminder (10% duty cycle + 0.5Hz): soft green light breathing flashes to reduce energy consumption. All parameter configurations and alarm records are encrypted and uploaded to the blockchain container 4 to ensure that the data is tamper-proof throughout its entire lifecycle. The parameter table supports OTA remote updates to adapt to different scenario requirements. Note: The parameter design complies with the IEC 60417 standard, taking into account both ergonomics and energy-saving requirements.

[0166] I. Effectiveness Verification Methods

[0167] Data tampering risk testing: Method: Simulate network attacks (such as man-in-the-middle attacks and data injection attacks) at the substation site, and compare the data tampering success rate of the traditional database system with that of this system. Expected verification: Statistically calculate the tampering failure rate (target: ≥99%) through continuous comparison of blockchain hash values. Fault diagnosis response time testing: Method: Deploy edge computing nodes to collect current / temperature data in real time under a grounding wire short circuit scenario, and use smart contracts to trigger diagnostic algorithms. Expected verification: Record the time difference between the occurrence of a fault and the system alarm using a high-speed camera (target: ≤50ms), and compare it with a traditional manual inspection system (average response >500ms).

[0168] Alarm recognition rate testing method: Inject historical fault datasets (such as grounding wire corrosion and breakage samples) and statistically analyze the proportion of anomalies automatically identified by the system. Expected verification: Compared to traditional threshold alarm systems, this system improves the recognition rate through edge computing + AI diagnostic models (target: 60% improvement). II. Test scenario description: Scenario: 500kV substation site, deploying 10 sets of grounding wire devices to simulate complex operating conditions such as humid environment and high load current. Comparison group: Traditional manual inspection + centralized data management system.

[0169] Key metrics:

[0170] Number of data tampering attempts / total number of attacks (verification of blockchain immutability); delay from fault occurrence to alarm (verification of edge computing real-time performance); decrease in false negative / false positive rate (verification of the effectiveness of intelligent diagnostic algorithm). Note: All tests must be repeated 30 times and the average value taken to ensure statistical significance.

[0171] Model deployment process:

[0172] Training MobileNetV2 in the cloud (input: ground wire infrared image / sensor fusion data);

[0173] Convert the model to TensorFlow Lite format via model quantization;

[0174] OTA push to edge gateway devices;

[0175] Edge-Blockchain Collaboration:

[0176] On-chain evidence storage is triggered only when a fault is detected;

[0177] The evidence includes: device ID, fault type, confidence level, and timestamp;

[0178] Use lightweight consensus mechanisms (such as PoA) to reduce latency;

[0179] Diagnostic optimization directions:

[0180] Combine multi-sensor data fusion (temperature / vibration / resistance value);

[0181] Add a timing analysis module to detect progressive faults;

[0182] Federated learning enables iterative model updates;

[0183] The system can achieve sub-second fault response (edge ​​inference <200ms), and at the same time, it uses blockchain to ensure that diagnostic records are tamper-proof, meeting the requirements of full life cycle management of power equipment.

[0184] Code explanation:

[0185] Module introduction: The pseudocode introduces the edge computing framework and blockchain interaction module, indicating that the functionality of these two modules will be needed in the actual implementation.

[0186] Model initialization: Load the pre-trained MobileNet V2 model and define the classification labels for the grounding wire status for subsequent fault diagnosis.

[0187] Edge diagnostic function:

[0188] Data preprocessing: Image normalization and enhancement processing are performed on sensor data to improve the prediction accuracy of the model.

[0189] Model inference: The loaded model is used to predict the processed image, and the prediction results are obtained. The fault type and confidence level are determined based on the maximum value index.

[0190] Local decision-making: Threshold filtering is performed based on confidence level and fault type. If the conditions are met, the event is recorded in the blockchain and the fault type and transaction hash are returned; otherwise, the normal state is returned.

[0191] Device-side call: Obtain infrared thermal imaging data, call the edge diagnostic function to perform fault diagnosis, and output the diagnostic results and blockchain evidence information.

[0192] 1. Three-tiered architecture (edge ​​terminal layer → edge computing layer → central cloud layer)

[0193] Hierarchical node types, deployment locations, and functional network connection methods;

[0194] Lightweight blockchain nodes at the edge terminal layer are deployed on grounding wire monitoring terminals (such as smart sensors and RFID tags) and are responsible for raw data collection and initial encryption uploading to the blockchain.

[0195] Near Field Communication Protocol: LoRaWAN / ZigBee connects the terminal and the edge server.

[0196] Wired interface: RS485 / Modbus for stable transmission in industrial environments.

[0197] The full-featured blockchain nodes of the edge computing layer are deployed on the edge servers of factories / substations, undertaking data aggregation, local consensus, intelligent diagnosis, and lightweight smart contract execution.

[0198] Hybrid connectivity:

[0199] Downlink: Access to terminal devices via protocol gateways (OPC UA / MQTT);

[0200] Uplink: Data is transmitted back to the central cloud via a private WiFi 6 / 5G network.

[0201] The central cloud super node (consensus / audit node) is deployed in the cloud control center and is responsible for global ledger synchronization, cross-domain consensus, policy formulation, and lifecycle analysis.

[0202] Redundant backbone network: SD-WAN or fiber optic leased lines connect to edge servers.

[0203] 2. Key Technology Design of Network Topology

[0204] Dynamic node discovery mechanism;

[0205] The edge server automatically identifies the connected terminal devices, binds the device identity through DID (Decentralized Identifier), and assigns data upload permissions based on smart contracts.

[0206] Layered consensus mechanism;

[0207] Terminal → Edge Layer: Practical Byzantine Fault Tolerance (PBFT) for Fast Verification of Local Data Authenticity 1;

[0208] Edge → Cloud Layer: Combine DPoS (Delegated Proof of Stake) to achieve efficient cross-regional consensus and reduce latency.

[0209] Secure passage construction;

[0210] The terminal-edge data transmission is encrypted using the national SM9 algorithm, and tunnels are established between edge nodes via TLS1.3+IPsec VPN to ensure the confidentiality of data transmission.

[0211] 3. Flexible topology expansion capability

[0212] Redundant deployment of edge nodes: Multiple edge servers are configured to form a cluster in key plant areas, and failover is achieved through the Raft protocol;

[0213] Lightweight node access: Older devices can access the network through embedded blockchain micronodes (such as ARM architecture containers), which are compatible with non-smart terminals.

[0214] The advantages compared to traditional solutions are shown in Table 2:

[0215] Table 2

[0216]

[0217] End-point layer: Deploy industrial-grade sensors (temperature / vibration / current) and embedded AI chips to execute lightweight diagnostic models based on TensorFlow Lite.

[0218] Edge layer: Deploy 2-3 edge servers in each region to handle data preprocessing, model inference, and generation of data fingerprints (SHA-256).

[0219] Blockchain Layer: Adopts a layered network structure, with main chain nodes deployed in cloud data centers and sub-chain nodes deployed on edge servers to achieve localized consensus.

[0220] 1. Edge-side model quantization deployment

[0221] import tensorflow as tf

[0222] #Load the pre-trained diagnostic model

[0223] model=tf.keras.models.load_model('equipment_fault.h5')

[0224] #Quantization Conversion Configuration

[0225] converter=tf.lite.TFLiteConverter.from_keras_model(model)

[0226] converter.optimizations=[tf.lite.Optimize.DEFAULT]

[0227] converter.target_spec.supported_types=[tf.float16]#FP16 quantization

[0228] #Generate lightweight models

[0229] quantized_model=converter.convert()

[0230] #Edge Device Deployment

[0231] with open('tf_lite / fault_diagnosis.tflite','wb')as f:

[0232] f.write(quantized_model)

[0233] This code achieves a 60-75% reduction in model size and a 3x increase in inference speed.

[0234] 2. Blockchain Data Storage Contract (Solidity Fragment)

[0235]

[0236]

[0237] Asymmetric encryption is used to verify the identity of edge nodes to ensure the trustworthiness of data sources, as detailed in Table 3:

[0238] Table 3

[0239]

[0240]

[0241] Edge node configuration: Utilizes NVIDIA Jetson AGX Xavier, integrating TensorRT acceleration library to support 50+ concurrent inference tasks;

[0242] Blockchain selection: Hyperledger Fabric 2.3+ with a custom consensus plugin, supporting 300+ transactions per second;

[0243] Security hardening: Hardware-level trusted execution environment (TEE) protection for keys;

[0244] Based on this, this invention provides a method for full lifecycle management of grounding wires. This invention uses blockchain technology to ensure the immutability and traceability of grounding wire operation records and fault data, achieving trusted management throughout the entire lifecycle. It leverages a lightweight artificial intelligence model built into the edge computing unit for localized real-time fault diagnosis, significantly improving response speed. By defining multi-mode audio-visual coding rules and supporting remote dynamic configuration, it achieves high-recognition-rate alarms adaptable to complex field environments. The system combines fault case feedback and an online model update mechanism to continuously improve diagnostic accuracy, while relying on a layered architecture and localized consensus to ensure the system's basic operational capabilities under abnormal conditions such as network outages.

[0245] Example 2

[0246] In this embodiment, a computer terminal device is provided, including:

[0247] One or more processors;

[0248] A memory, coupled to the processor, for storing one or more programs;

[0249] When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the above-described grounding wire lifecycle management method.

[0250] In this embodiment, a computer-readable storage medium is also provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the above-described grounding wire lifecycle management method.

[0251] In this embodiment, an electronic device is also provided, including a memory and a processor. The memory stores a computer program, and the processor is configured to run the computer program to perform the steps of the above-described grounding wire lifecycle management method.

[0252] In this embodiment, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of the above-described grounding wire lifecycle management method.

[0253] The aforementioned program can run on a processor or be stored in memory (or a computer-readable medium). Computer-readable media include both permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0254] These computer programs may also be loaded onto a computer or other programmable data processing equipment, causing a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes can be implemented by different modules for different steps.

[0255] This embodiment provides such a device or system. The system, referred to as a grounding wire lifecycle management system, includes:

[0256] The status awareness module is used to collect status data of the grounding wire through sensors deployed at the grounding wire site;

[0257] An edge intelligent diagnostic module is used to perform localized fault diagnosis on the status data based on the edge computing unit.

[0258] The blockchain evidence storage module is used to generate hash values ​​from fault characteristic data and upload them to the blockchain network for evidence storage when a fault is diagnosed.

[0259] The multi-mode alarm module is used to trigger corresponding multi-mode alarm signals based on the results of the fault diagnosis.

[0260] As one implementation method in this embodiment, the state sensing module includes:

[0261] A pressure sensing unit is attached to detect pressure changes generated when the grounding wire is attached.

[0262] A current sensing unit is connected to detect the current change generated when the grounding wire is connected to the circuit;

[0263] The node triggering unit is used to generate a blockchain node containing time, location, and operator information by using the pressure change or current change as a trigger signal.

[0264] As one implementation method in this embodiment, the edge intelligent diagnostic module includes:

[0265] The data preprocessing unit is used to filter and normalize the collected data within the edge computing unit;

[0266] The fault screening unit is used to compare the preprocessed data with a preset safety threshold to achieve fault screening.

[0267] The deep inference unit is used to input the data that was initially screened as anomalies into a lightweight neural network model for deep inference to determine the type of fault.

[0268] As one implementation method in this embodiment, the multi-mode alarm module includes:

[0269] The signal generation unit is used to generate a PWM control signal containing a specific flashing frequency and a buzzer frequency based on the level of the fault diagnosis result.

[0270] The sound and light driving unit is used to drive the LED light and buzzer using the PWM control signal to emit sound and light alarm signals corresponding to the fault level.

[0271] As one implementation method in this embodiment, the multi-mode alarm module further includes:

[0272] The instruction receiving unit is used to receive remote configuration instructions through the management platform;

[0273] A logic configuration unit is used to parse and store the instructions to dynamically update the mapping logic between alarm modes and vibration modes.

[0274] As one implementation method in this embodiment, the blockchain evidence storage module further includes:

[0275] The case database update unit is used to store failure case data into the blockchain case database;

[0276] The model optimization unit is used to drive the parameter updates of the lightweight neural network model based on historical failure case data.

[0277] The system or apparatus is used to implement the functions of the methods in the above embodiments. Each module in the system or apparatus corresponds to each step in the method, as has been described in the method and will not be repeated here.

[0278] The above implementation method solves the problem of full life cycle management of grounding wires in related technologies, thereby ensuring that the problems existing in the prior art are resolved.

[0279] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for managing the entire lifecycle of a grounding wire, characterized in that, Includes the following steps: Its status data is collected by sensors deployed at the grounding wire site; Localized fault diagnosis is performed on the status data based on edge computing units; When a fault is diagnosed, the fault characteristic data is generated into a hash value and uploaded to the blockchain network for evidence storage. Based on the results of the fault diagnosis, the corresponding multi-mode alarm signal is triggered.

2. The method according to claim 1, characterized in that, The process of collecting status data through sensors deployed at the grounding wire site includes: The pressure change generated when the grounding wire is connected is detected by a pressure sensor; Alternatively, a current sensor can be used to detect changes in current generated when the grounding wire is connected to the circuit; Using the pressure or current change as a trigger signal, a blockchain node containing time, location, and operator information is generated.

3. The method according to claim 1, characterized in that, The process of performing localized fault diagnosis on the status data based on edge computing units includes: The collected data is preprocessed by filtering and normalization within the edge computing unit; The preprocessed data is compared with a preset safety threshold to achieve initial fault screening. For data that is initially screened as abnormal, it is input into a lightweight neural network model for deep inference to determine the fault type.

4. The method according to claim 1, characterized in that, The process of triggering the corresponding multi-mode alarm signal includes: Based on the level of the fault diagnosis result, a PWM control signal containing a specific flashing frequency and a buzzer frequency is generated; The PWM control signal is used to drive the LED light and buzzer to emit an audible and visual alarm signal corresponding to the fault level.

5. The method according to claim 4, characterized in that, The process also includes: Remote configuration commands are issued to edge devices through the management platform; The edge device parses and stores the instructions to dynamically update the mapping logic between alarm modes and vibration modes.

6. The method according to claim 1, characterized in that, After the process of generating hash values ​​from fault feature data and uploading them to the blockchain network for evidence storage, the following is also included: Store failure case data in a blockchain case library; The parameters of the lightweight neural network model are updated based on historical failure case data.

7. A grounding wire lifecycle management system, characterized in that, The system includes: The status awareness module is used to collect status data of the grounding wire through sensors deployed at the grounding wire site; An edge intelligent diagnostic module is used to perform localized fault diagnosis on the status data based on the edge computing unit. The blockchain evidence storage module is used to generate hash values ​​from fault characteristic data and upload them to the blockchain network for evidence storage when a fault is diagnosed. The multi-mode alarm module is used to trigger corresponding multi-mode alarm signals based on the results of the fault diagnosis.

8. A computer terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors perform the steps of the method as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-6.

Images