Visual soil electrical parameter measurement method and system

Abstract

The invention relates to the technical field of lightning disaster early warning, provides a visual soil electrical parameter measurement method and system, and aims to solve the problems of one-sided evaluation, inaccurate early warning and low operation and maintenance efficiency caused by data isolation, model solidification, single display and lack of closed-loop control in the prior art. Obtaining lightning data, real-time meteorological data and soil resistance data, and performing unified access, analysis, abnormal value filtering and time-space alignment; carrying out parallel intelligent analysis, carrying out lightning risk prediction by adopting a CNN-LSTM-Attention mixed model, carrying out multi-dimensional feature extraction on soil resistance data, and matching with a grounding grid fault diagnosis rule base; carrying out multi-dimensional fusion visual display, and carrying out multi-dimensional visual rendering and linkage interaction on the thunder and lightning risk level and the ground screen health state in a unified interface; and performing linkage early warning and control, performing comprehensive risk assessment based on the situation view, triggering graded early warning, and generating an equipment control instruction to realize closed-loop management.

Description

Technical Field

[0001] This invention belongs to the fields of computer software technology and lightning disaster early warning technology, and specifically relates to a method and system for measuring visualized soil electrical parameters. Background Technology

[0002] Lightning disasters are one of the main threats to the safe and stable operation of power systems. The transient overvoltages and ground potential rises they cause can directly lead to equipment damage, system outages, and even casualties. Therefore, accurate early warning of lightning activity and real-time monitoring of the grounding grid are crucial. Traditionally, lightning early warning mainly relies on meteorological observation station data and atmospheric electric field strength monitoring; while grounding grid condition assessment largely depends on periodic manual inspections and measurements. These methods are generally inefficient and lack real-time performance. With the development of the Internet of Things (IoT) and software technology, systems have emerged that connect monitoring devices to the network and manage them centrally through software platforms, aiming to achieve remote monitoring and intelligent early warning.

[0003] Based on patent search and analysis, the prior art most similar to the proposal in this application includes:

[0004] 1. CN112946372A "An Intelligent Building Lightning Safety Monitoring System and Method"

[0005] The specific implementation of this technical solution is as follows: The system includes data acquisition terminals with multiple functional modules such as lightning early warning, intelligent lightning current monitoring, SPD monitoring, and intelligent grounding resistance monitoring. These terminals transmit the collected data to the data acquisition and processing module via data cables for centralized analysis and processing. The processed monitoring information is displayed on a local configuration LCD screen and simultaneously uploaded to the cloud platform via a TCP / IP network through the DTU data transmission module. Users can ultimately access the cloud platform through smart terminal devices such as computers and mobile phones to achieve remote real-time monitoring and fault alarms.

[0006] Compared to the proposed application, this technical solution differs in that its core contribution lies in constructing a multi-functional, modular, cloud-based remote monitoring architecture, enabling centralized data display. However, the various functional modules of this system (such as lightning warning and grounding resistance monitoring) are relatively independent, essentially representing a "parallel presentation" of data, lacking in-depth, cross-module data fusion analysis and collaborative early warning mechanisms. Its early warning function remains based on preset fixed thresholds, failing to reflect a design that leverages intelligent algorithms to uncover deep correlations between data to improve early warning accuracy.

[0007] 2. CN115877091A "Intelligent Lightning Monitoring System and Method for Photovoltaic Power Plants"

[0008] The specific implementation of this technical solution is as follows: The system consists of a lightning proximity early warning device, a lightning protection online monitoring device, a lightning monitoring and recording device, and a monitoring and management platform. The lightning proximity early warning device is installed on the roof of the photovoltaic system. It monitors the dynamic state of thunderclouds using a solid-state atmospheric electric field detector and separates the dynamic and static electric fields based on lightning detection methods to generate lightning warning information. The lightning protection online monitoring device is located on the AC side of the inverter and is used to monitor in real time the leakage current, temperature, and other status parameters of the surge protection device (SPD) and the tripping status of its dedicated backup protection device. The lightning monitoring and recording device records parameters such as the peak value and polarity of the lightning current that have occurred. The monitoring and management platform collects the above data via TCP / IP or a wireless network, summarizes and displays it, and issues alarms for abnormalities in the status of the lightning protection equipment.

[0009] Compared with the proposal in this application, the technical solution differs in that: its technical solution is closely designed around the lightning protection needs of the specific scenario of photovoltaic power plants, and its core function is to monitor and alarm the status of the lightning protection equipment itself and the local atmospheric electric field; its early warning logic mainly relies on a single threshold judgment based on electric field strength; the solution does not involve continuous monitoring and analysis of soil electrical parameters, nor does it diagnose and visualize the macroscopic health status of the grounding grid (such as corrosion and fracture), nor does it introduce a deep learning model based on multi-source data fusion for forward-looking prediction.

[0010] 3. CN115982243A "An Adaptive Lightning Protection Method and System Based on Meteorological Monitoring"

[0011] The specific implementation of this technical solution is as follows: the system focuses on achieving early warning by monitoring the core parameter of atmospheric electric field strength. In terms of software flow, it first monitors the electric field strength in real time, then calculates its absolute value and rate of change; subsequently, these two calculated values ​​are compared with preset multi-level thresholds, and based on different combinations of thresholds, a level one, two, or three alarm level is mechanically determined. The system also integrates monitoring functions for SPDs and grounding devices, aiming to construct a comprehensive monitoring system.

[0012] Compared to the proposal in this application, the technical solution differs in that its early warning model is essentially a rule-driven model based on fixed thresholds and simple logical judgments. The entire early warning process does not incorporate any machine learning or deep learning models that require training based on historical data and can automatically learn from the data to discover complex nonlinear relationships. Its "adaptability" is more reflected in the application of multi-level thresholds than in the model itself being able to dynamically adjust and optimize according to data characteristics. Furthermore, this solution also does not involve the continuous monitoring and intelligent diagnosis of deep geological grid state parameters such as soil resistivity.

[0013] Compared with the aforementioned closest existing technical solutions, they all share the following technical drawbacks, which are precisely the technical problems that this application aims to solve:

[0014] 1. Regarding the technical problems existing in CN112946372A: Although this solution constructs a cloud platform architecture integrating multiple modules, the data of each monitoring module is relatively independent, lacking a deep data fusion and analysis mechanism across modules. This results in severe data isolation and an inability to achieve comprehensive risk assessment. This application proposes to solve the problems of "information silos" and one-sided risk assessment in existing technologies by establishing a deep fusion and correlation analysis model of soil electrical parameters, lightning monitoring data, and meteorological data.

[0015] 2. Regarding the technical issues in CN115877091A: The technical focus of this solution is on monitoring lightning protection equipment in photovoltaic scenarios and threshold-based early warning based on a single electric field strength, which suffers from limitations in functionality and low intelligence of the early warning model. This application proposes to address the problems of insufficient grounding grid status perception and inadequate early warning accuracy in existing technologies by introducing continuous monitoring of soil electrical parameters and grounding grid status diagnosis functions, combined with a CNN-LSTM-Attention deep learning model for multi-source feature mining.

[0016] 3. Regarding the technical problems existing in CN115982243A: The early warning model of that solution is based on preset fixed thresholds and static logical judgments, lacking the ability to self-learn and optimize from data, and suffers from the technical shortcomings of simple and rigid early warning models and insufficient adaptive capabilities. This application proposes to construct an intelligent early warning system that can automatically learn complex patterns and dynamically optimize from multi-source time-series data by employing a deep learning model trained on historical data, thereby solving the problems of rigid early warning models and the inability to continuously improve accuracy in existing technologies. Summary of the Invention

[0017] The purpose of this invention is to solve the problems of isolated data, simple and rigid analysis models, limited visualization, and lack of intelligent closed-loop control mechanisms in existing soil electrical parameter monitoring and lightning early warning systems, which lead to one-sided risk assessment, insufficient early warning accuracy, and low operation and maintenance efficiency.

[0018] To achieve the above objectives, the present invention employs the following technical means:

[0019] This invention provides a method for visually measuring soil electrical parameters, comprising the following steps:

[0020] S1. Multi-source data acquisition and standardization processing: acquire lightning data collected by lightning monitoring equipment, real-time meteorological data provided by meteorological API, and soil resistance data collected by soil resistance monitoring equipment; perform unified access, parsing, outlier filtering, and spatiotemporal alignment on the multi-source heterogeneous data to generate a standardized dataset;

[0021] S2. Parallel Intelligent Analysis: A CNN-LSTM-Attention hybrid model is used to predict lightning risk from the real-time meteorological data in the standardized dataset to obtain the lightning risk level; at the same time, multi-dimensional feature extraction is performed on the soil resistance data in the standardized dataset, and it is matched with a pre-set grounding grid fault diagnosis rule base to obtain the grounding grid health status.

[0022] S3. Multi-dimensional integrated visualization display: The lightning risk level and ground grid health status are visualized and interacted with in a unified software front-end interface to generate a situational view that comprehensively reflects the external lightning risk and the internal ground grid status.

[0023] S4. Linked Early Warning and Control: Based on the situation view, a comprehensive risk assessment is performed. When any risk exceeds a preset threshold or the combination of two risks reaches a preset condition, a graded early warning is triggered, and equipment control commands are generated to adjust the operating parameters of relevant equipment, thereby realizing closed-loop management of monitoring-early warning-control.

[0024] In the above scheme, step 1 includes:

[0025] Step 1.1: The soil resistance monitoring equipment and lightning current sensor distributed on site are received in real time through the built-in TCP / IP communication server in the software to obtain the raw data stream containing the device ID and timestamp;

[0026] Step 1.2: Parse the raw data stream according to the preset communication protocol and filter out outliers that exceed the physical range to obtain the effective payload data;

[0027] Step 1.3: Perform format conversion and spatiotemporal alignment on the payload data and the meteorological data obtained from the Open-Meteo meteorological API to finally obtain a unified standard JSON structure data and store it in the database.

[0028] In the above scheme, step 2 specifically includes the following steps:

[0029] Step 2.1: Normalize and segment the real-time meteorological data in the standardized dataset using a sliding window, and call the pre-trained CNN-LSTM-Attention hybrid model for inference. The CNN module extracts the spatial coupling features of multiple meteorological parameters, the LSTM module learns the temporal evolution of the features, and the Attention module calculates the attention weights at different time steps, dynamically focusing on the key meteorological mutation features before the occurrence of lightning, and outputting the probability and risk level of lightning occurrence in the next 30 minutes.

[0030] Step 2.2: Perform time-domain and frequency-domain multi-dimensional feature calculations on the soil resistance time-series data in the standardized dataset, and perform matching analysis with the obtained feature set and the pre-set grounding grid fault diagnosis rule base to obtain the grounding grid diagnosis results with geographical location and health status identifiers.

[0031] In the above scheme, step 3 specifically includes the following steps:

[0032] Step 3.1: Use the ECharts library to plot the probability trend curve of the lightning risk level results, and use the Leaflet.js map library to display the risk distribution in the form of a heat map to obtain a visualization layer of lightning risk. Use the Three.js engine to build a three-dimensional 3D model and simulate the lightning strike effect in real time.

[0033] Step 3.2: Use canvas to build a two-dimensional topology model based on the grounding grid diagnosis results, and dynamically color it according to the health status to obtain a two-dimensional visualization layer of the grounding grid status;

[0034] Step 3.3: Integrate and link the lightning risk visualization layer, the lightning strike 3D model, and the ground grid status 2D visualization layer in the unified software front-end interface to finally form a comprehensive situation view.

[0035] In the above scheme, step 4 specifically includes the following steps:

[0036] Step 4.1: Perform a risk assessment on the lightning risk and grounding grid status information presented in the comprehensive situation view. If any risk exceeds the threshold or the combination of the two risks reaches the preset conditions, an alarm will be triggered, and the alarm level and target device list will be obtained.

[0037] Step 4.2: The alarm level and target device list are pushed in real time through the software interface, and remote control commands are generated to adjust the operating parameters of the relevant devices, ultimately realizing an intelligent closed loop from monitoring to control.

[0038] The present invention also provides a visualization soil electrical parameter measurement system, wherein the processor executes the program to implement the method described above.

[0039] Because the present invention employs the above-mentioned technical means, it has the following beneficial effects:

[0040] 1. This invention solves the "information silo" problem in existing technologies, where lightning data, meteorological data, and soil resistivity data operate independently, through a standardized processing method encompassing "protocol parsing, anomaly filtering, and spatiotemporal alignment" in step S1 (steps 1.1 to 1.3). It achieves consistent integration of multi-source heterogeneous data, eliminates time reference discrepancies between different devices, provides high-quality "net data" for subsequent parallel intelligent analysis, and elevates the risk assessment dimension from single external monitoring to a dual-dimensional collaborative assessment of "external lightning threats and internal grounding network hazards," fundamentally solving the problem of one-sided assessment.

[0041] 2. This invention, through the intelligent reasoning method of integrating a "CNN-LSTM-Attention hybrid model" in step S2 (step 2.1), solves the problems of existing technologies relying on fixed threshold warnings and lacking judgment on future trends. The system is no longer a passive data recording dashboard, but has the ability to mine complex nonlinear correlations between meteorological parameters. Compared with traditional methods, this invention can effectively identify early meteorological precursors of lightning, significantly reducing the false alarm and missed alarm rates, realizing a leap from "post-event alarm" to "pre-event defense," and reserving a valuable response window for lightning protection of tall buildings.

[0042] 3. This invention solves the technical problems of traditional grounding grid status relying on manual, blind excavation and the invisibility of hidden works by using a concrete diagnostic method in step S2 (step 2.2) of "matching topology network resistance mapping with fault diagnosis rule base". It transforms abstract soil resistance data into intuitive health status indicators of "normal / increased / corroded / disconnected" and accurately maps them to specific geographical topology branches. This achieves "transparent" management of hidden works, enabling maintenance personnel to pinpoint fault areas within minutes, avoiding the lag and blindness of manual inspections.

[0043] 4. This invention, through the "multi-dimensional integrated display method" in step S3 (steps 3.1 to 3.3), solves the problem of the inability to correlate and present isolated monitoring data in the prior art. At the visualization level, the system achieves spatial mapping and overlay of "dynamic lightning strike risk" and "static grounding network status." Although lightning warning and grounding network diagnosis operate independently at the algorithm level, at the decision display level, the system can overlay "high lightning probability areas" onto "high corrosion risk branches," helping maintenance personnel quickly locate composite risk points of "external high attack (lightning strike) + internal weak defense (corrosion)," which is impossible for a single monitoring system.

[0044] 5. This invention solves the semi-automation problem of "disconnect between monitoring and control" in existing technologies by using a closed-loop management approach of "multi-factor risk threshold judgment + hierarchical linkage control" in step S4 (steps 4.1 to 4.2). It constructs a fully automated closed loop of "monitoring-analysis-early warning-control". When the overall risk exceeds the standard, compared with the traditional manual intervention mode, it significantly improves the emergency response speed and effectively avoids the risk of equipment damage caused by delayed human decision-making.

[0045] 6. (Synergistic Effect) This invention deeply couples "data standardization (S1), time-series early warning, ground network topology diagnosis (S2), multi-dimensional display methods (S3), and closed-loop linkage control (S4)" at the system level, generating a significant synergistic effect. Specifically, the invention employs a synergistic strategy of "simultaneous association of spatiotemporal information and comprehensive multi-dimensional situational analysis." The intelligent early warning module is responsible for judging the probability of external lightning strikes, while the ground network diagnosis module is responsible for providing the health status of internal facilities. At the display and decision-making level, the system spatially maps and overlays the real-time lightning warning status with the topological health status of the ground network, constructing a comprehensive operation and maintenance view that addresses both "external threats" and "internal concerns." This overcomes the limitations of single-dimensional monitoring in operation and maintenance decision-making. Specifically, lightning warnings alone cannot clearly define defense priorities, and ground network data alone cannot determine the urgency of maintenance. This invention provides maintenance personnel with a complete decision-making context by presenting the two in relation to each other, enabling them to intuitively identify the overlapping areas of risks and make accurate decisions such as "prioritizing the weakest link during the window of opportunity before thunderstorms arrive." The overall protection effectiveness is significantly better than the isolated display of data from each module. Attached Figure Description

[0046] Figure 1 A simplified flowchart of the system process;

[0047] Figure 2 This is a system framework diagram;

[0048] Figure 3 Screenshot of the software's homepage;

[0049] Figure 4 For taking screenshots of large-screen visualizations;

[0050] Figure 5 Screenshot of the lightning warning module software interface;

[0051] Figure 6 Screenshot of the software user management interface;

[0052] Figure 7 Screenshot of the software interface role management;

[0053] Figure 8 Screenshot of the software interface for device management;

[0054] Figure 9 Take a screenshot of the software interface in real time;

[0055] Figure 10 Screenshot of data statistics;

[0056] Figure 11 Screenshot of historical data;

[0057] Figure 12 Screenshot of ground network diagnosis. Detailed Implementation

[0058] The embodiments of the present invention will be described in detail below. Although the present invention will be described and illustrated in conjunction with some specific embodiments, it should be noted that the present invention is not limited to these embodiments. On the contrary, any modifications or equivalent substitutions made to the present invention should be covered within the scope of the claims of the present invention.

[0059] Furthermore, to better illustrate the present invention, numerous specific details are set forth in the following detailed embodiments. Those skilled in the art will understand that the present invention can be practiced without these specific details.

[0060] This application revolves around the core logic of "multi-source data closed loop + intelligent analysis + multi-dimensional visualization," constructing a full-process technical solution encompassing "data acquisition - preprocessing - intelligent analysis - visualization display - hierarchical early warning."

[0061] A software system and method for visualizing soil electrical parameters measurement, the specific steps of which are as follows:

[0062] Step 1: Unify the access, parsing and standardization of multi-source heterogeneous data from lightning monitoring equipment, meteorological API and soil resistance monitoring equipment to obtain a spatiotemporally aligned standardized dataset;

[0063] Step 2: Perform lightning probability prediction on real-time meteorological data in the standardized structured data based on a CNN-LSTM-Attention hybrid model. At the same time, perform multi-dimensional feature extraction and ground grid status diagnosis on soil resistivity data in the standardized dataset to obtain two types of analysis results: lightning risk level and ground grid health status.

[0064] Step 3: Perform multi-dimensional fusion visualization rendering on the two types of analysis results, namely lightning risk level and ground grid health status, and display them collaboratively in a unified portal interface to obtain a situational view that comprehensively reflects the external lightning risk and the internal ground grid status.

[0065] Step 4: Develop intelligent early warning decisions and generate equipment control commands based on the comprehensive risk assessment results in the visualization interface, ultimately completing closed-loop management from risk perception to early warning response.

[0066] Step 1 above specifically includes the following steps:

[0067] Step 1.1: The soil resistance monitoring equipment and lightning current sensor distributed on site are received in real time through the built-in TCP / IP communication server in the software to obtain the raw data stream containing the device ID and timestamp;

[0068] Step 1.2: Parse the raw data stream according to the preset communication protocol and filter out outliers that exceed the physical range to obtain the effective payload data;

[0069] Step 1.3: Perform format conversion and spatiotemporal alignment on the payload data and meteorological data obtained from meteorological APIs such as Open-Meteo, and finally obtain JSON structure data with unified standard and store it in the database.

[0070] Step 2 above specifically includes the following steps:

[0071] Step 2.1: Normalize and segment the real-time meteorological data in the standardized dataset using a sliding window, and call the pre-trained CNN-LSTM-Attention hybrid model for inference. The CNN module extracts the spatial coupling features of multiple meteorological parameters, the LSTM module learns the temporal evolution of the features, and the Attention module calculates the attention weights at different time steps, dynamically focusing on the key meteorological mutation features before the occurrence of lightning, and outputting the probability and risk level of lightning occurrence in the next 30 minutes.

[0072] Step 2.2: Perform topology-based branch mapping on the standardized resistance monitoring data, and use the preset four-level resistance threshold rules (i.e., resistance deviation less than 10% is normal, 10%-20% is increased, 20%-50% is corrosion, and exceeding 50% or open circuit is disconnection) to perform logical matching analysis to obtain the grounding network topology diagnosis results, which include specific location, health status indicators (normal / increased / corrosion / disconnection) and corresponding color rendering instructions (green / yellow / brown / red).

[0073] Step 3 above specifically includes the following steps:

[0074] Step 3.1: Use the ECharts library to plot the probability trend curve of the lightning risk level results, and use the Leaflet.js map library to display the risk distribution in the form of a heat map to obtain a visualization layer of lightning risk. Use the Three.js engine to build a three-dimensional 3D model and simulate the lightning strike effect in real time.

[0075] Step 3.2: Use canvas to build a two-dimensional topology model based on the grounding grid diagnosis results, and dynamically color it according to the health status to obtain a two-dimensional visualization layer of the grounding grid status;

[0076] Step 3.3: Integrate and link the lightning risk visualization layer, the lightning strike 3D model, and the ground grid status 2D visualization layer in the unified software front-end interface to finally form a comprehensive situation view.

[0077] Step 4 above specifically includes the following steps:

[0078] Step 4.1: Conduct a comprehensive risk assessment by combining the future occurrence probability output by the lightning warning model in the comprehensive situation view with the branch health status output by the grounding grid topology diagnosis. Use the combined judgment logic of "single factor high risk trigger" (if the lightning probability exceeds 80% or the grounding grid branch is disconnected) and "double factor superposition upgrade" (i.e., the lightning probability exceeds 50% and the grounding grid branch is corroded) to perform matching analysis, and obtain a comprehensive control instruction that includes the exact alarm level (red / orange / yellow), risk type and list of affected target equipment.

[0079] Step 4.2: The alarm level and target device list are pushed in real time through the software interface, and remote control commands are generated to adjust the operating parameters of the relevant devices, ultimately realizing an intelligent closed loop from monitoring to control.

[0080] The innovative effects and features of this invention are as follows:

[0081] 1. A robust data "firewall" is built through unified TCP / IP access devices, standardized protocol parsing, and anomaly data removal. This solves the stability issues of existing technologies, which are susceptible to interference from complex environmental noise and equipment malfunctions, leading to distortion of analytical model input and blockage of system data flow processing logic. This standardized data processing architecture ensures high robustness of the system under conditions of local data anomalies and network fluctuations. Compared with traditional monitoring systems, it effectively avoids system paralysis caused by data errors, achieving system availability of over 99.5% and improving mean time between failures (MTBF) by approximately 40%.

[0082] 2. Through unified data access and standardized processing, lightning uplink current data, multi-level meteorological data, soil resistivity and other data are uniformly processed into a standard dataset, providing a high-quality and consistent data foundation for subsequent professional analyses of different types, and laying the cornerstone for system integration. Due to the adoption of a multi-source data fusion method, compared with existing systems with relatively independent functional modules or single data sources, a more comprehensive environmental status perception system can be established. Cross-validation of multi-dimensional data improves the reliability and accuracy of status assessment, lightning warning, and grounding grid fault identification.

[0083] 3. Through specialized and in-depth analysis of different data within the same dataset, relatively independent lightning warning and ground grid status monitoring processes are set up within a single software system. These processes are then coordinated through a comprehensive display, enabling the system to efficiently and simultaneously perform intelligent diagnosis of both the external meteorological and lightning environment and the ground grid, balancing efficiency and professionalism. Specifically, the lightning warning process employs a software implementation method based on a CNN-LSTM-Attention hybrid model. CNN is used for spatial feature extraction of meteorological data, and LSTM is used for temporal modeling. Compared to existing methods based on threshold judgment or simple statistical models, this approach can uncover deeper and more complex fault and risk characteristics, effectively reducing the false negative rate. Simultaneously, multi-dimensional feature analysis of ground grid resistance data further enhances diagnostic capabilities.

[0084] 4. Through multi-dimensional fusion visualization, lightning risk information and grounding network status information from different processing flows are integrated, rendered, and interacted with in a unified front-end interface. Using ECharts, Three.js, and Leaflet.js for collaborative visualization, the abstract lightning risk probability and grounding network topology status are transformed into intuitive graphical displays, providing users with a global and intuitive situational awareness and solving the problem of data fragmentation across multiple systems. Compared with the simple data tables or two-dimensional charts in existing technologies, the efficiency of situational awareness is significantly improved, and the fault location time is significantly shortened.

[0085] 5. Through an intelligent decision-making and linkage control mechanism based on comprehensive situational awareness, cross-functional integrated alarms are triggered based on a unified comprehensive situational view, and equipment control commands are generated. Relying on the equipment management and announcement publishing modules, the transformation from complex data to collaborative operation and maintenance is realized, forming a complete software closed loop of monitoring, analysis, early warning, control, and notification. At the same time, through the optimization of the unified data access layer, the improvement of data read and write efficiency through MySQL connection pool technology, and the lightweight model inference, the response latency for a single lightning prediction is <1 second, the latency for soil resistance data analysis is <2 seconds, and minute-level data updates and real-time early warnings are supported. Compared with existing systems with long data processing chains and complex model calculations, the system's real-time performance is significantly improved.

Claims

1. A method for visualizing soil electrical parameters, characterized in that, Includes the following steps: Step S1. Multi-source data acquisition and standardization processing: acquire lightning data collected by lightning monitoring equipment, real-time meteorological data provided by meteorological API, and soil resistance data collected by soil resistance monitoring equipment. Perform unified access, parsing, outlier filtering, and spatiotemporal alignment on the multi-source heterogeneous data to generate a standardized dataset. Step S2. Parallel intelligent analysis: The real-time meteorological data in the standardized dataset is used to predict lightning risk using a CNN-LSTM-Attention hybrid model to obtain the lightning risk level; at the same time, multi-dimensional feature extraction is performed on the soil resistance data in the standardized dataset, and it is matched with a pre-set grounding grid fault diagnosis rule base to obtain the grounding grid health status. Step S3. Multi-dimensional fusion visualization display: The lightning risk level and ground grid health status are visualized and interacted with in a unified software front-end interface to generate a situational view that comprehensively reflects the external lightning risk and the internal ground grid status. Step S4. Linked Early Warning and Control: Based on the situation view, a comprehensive risk assessment is performed. When any risk exceeds a preset threshold or the combination of two risks reaches a preset condition, a graded early warning is triggered, and equipment control commands are generated to adjust the relevant equipment operating parameters, thereby realizing closed-loop management of monitoring-early warning-control.

2. The method according to claim 1, characterized in that: Step 1 includes: Step 1.1: The soil resistance monitoring equipment and lightning current sensor distributed on site are received in real time through the built-in TCP / IP communication server in the software to obtain the raw data stream containing the device ID and timestamp; Step 1.2: Parse the raw data stream according to the preset communication protocol and filter out outliers that exceed the physical range to obtain the effective payload data; Step 1.3: Perform format conversion and spatiotemporal alignment on the payload data and the meteorological data obtained from the Open-Meteo meteorological API to finally obtain a unified standard JSON structure data and store it in the database.

3. The method according to claim 1, characterized in that: Step 2 specifically includes the following steps: Step 2.1: Normalize and segment the real-time meteorological data in the standardized dataset using a sliding window, and call the pre-trained CNN-LSTM-Attention hybrid model for inference. The CNN module extracts the spatial coupling features of multiple meteorological parameters, the LSTM module learns the temporal evolution of the features, and the Attention module calculates the attention weights at different time steps, dynamically focusing on the key meteorological mutation features before the occurrence of lightning, and outputting the probability and risk level of lightning occurrence in the next 30 minutes. Step 2.2: Perform time-domain and frequency-domain multi-dimensional feature calculations on the soil resistance time-series data in the standardized dataset, and perform matching analysis with the obtained feature set and the pre-set grounding grid fault diagnosis rule base to obtain the grounding grid diagnosis results with geographical location and health status identifiers.

4. The method according to claim 1, characterized in that: Step 3 specifically includes the following steps: Step 3.1: Use the ECharts library to plot the probability trend curve of the lightning risk level results, and use the Leaflet.js map library to display the risk distribution in the form of a heat map to obtain a visualization layer of lightning risk. Use the Three.js engine to build a three-dimensional 3D model and simulate the lightning strike effect in real time. Step 3.2: Use canvas to build a two-dimensional topology model based on the grounding grid diagnosis results, and dynamically color it according to the health status to obtain a two-dimensional visualization layer of the grounding grid status; Step 3.3: Integrate and link the lightning risk visualization layer, the lightning strike 3D model, and the ground grid status 2D visualization layer in the unified software front-end interface to finally form a comprehensive situation view.

5. The method according to claim 1, characterized in that: Step 4 specifically includes the following steps: Step 4.1: Perform a risk assessment on the lightning risk and grounding grid status information presented in the comprehensive situation view. If any risk exceeds the threshold or the combination of the two risks reaches the preset conditions, an alarm will be triggered, and the alarm level and target device list will be obtained. Step 4.2: The alarm level and target device list are pushed in real time through the software interface, and remote control commands are generated to adjust the operating parameters of the relevant devices, ultimately realizing an intelligent closed loop from monitoring to control.

6. A visual soil electrical parameter measurement system, characterized in that, When the processor executes the program, it implements the method as described in claims 1-5.

Images