Rock slope intelligent monitoring and early warning system and method based on Internet of Things technology

Abstract

The invention discloses a rock-soil slope intelligent monitoring and early warning system based on the Internet of Things technology, which comprises a sensor module, a data acquisition module, a data analysis module, a cloud computing module, a data visualization module, a three-dimensional model display module, an intelligent inspection module and an early warning module, and is characterized in that the data acquisition module is used for collecting and preprocessing monitoring data; the data analysis module is used for analyzing the preprocessed real-time monitoring data; the data visualization module is used for generating a self-defined report and historical simulation data; the three-dimensional model display module is used for real scene fusion display; and the early warning module is used for automatically triggering an early warning mechanism when abnormal data is detected, and evaluating the risk level of the rock-soil slope in real time. The system monitors various deformation key parameters of the rock-soil slope in real time, predicts potential landslide and settlement problems through an advanced data analysis algorithm, and provides a visual user interface and an early warning system at the same time to improve the safety and economic benefits of slope engineering.

Description

Technical Field

[0001] The present invention relates to the technical field of geotechnical engineering monitoring, and particularly relates to an intelligent monitoring and early warning system and method for geotechnical slopes based on Internet of Things technology. Background Art

[0002] In the field of geotechnical engineering, monitoring the stability of geotechnical slopes is a complex and crucial task, which is directly related to engineering safety and the protection of people's lives and property. Traditional geotechnical engineering monitoring methods mainly rely on manual regular inspections and simple instrument measurements, and these methods have many limitations. First of all, manual monitoring is inefficient and cannot achieve continuous and real-time data collection, resulting in the omission and delay of key data. Secondly, traditional methods lack effective data processing and analysis capabilities, and it is difficult to extract valuable information from a large amount of monitoring data, let alone accurately predict potential geological disasters. In addition, due to the dispersion of monitoring data and the lack of centralized management and analysis, it is extremely difficult to establish a remote monitoring and early warning system.

[0003] With the development of Internet of Things technology, real-time monitoring and remote monitoring have become possible. Internet of Things technology collects data through sensor modules and network communication technology, and uses cloud computing and big data analysis technology to process and analyze the data, so as to achieve continuous monitoring of the stability of geotechnical slopes. However, most of the existing Internet of Things monitoring systems focus on data collection, and there are still deficiencies in data analysis, early warning mechanisms, and user interaction. These systems often lack in-depth mining and intelligent processing of data, and cannot provide accurate early warning information, resulting in the inability to take timely and effective preventive measures before disasters occur.

[0004] The existing monitoring systems also have deficiencies in terms of user experience. The interfaces are not friendly, the operations are complex, and there are no intuitive display and interaction functions, making it difficult for non-professionals to understand and use these systems. These problems limit the effective application and development of Internet of Things technology in the field of geotechnical engineering monitoring.

[0005] Therefore, it is necessary to develop a new type of intelligent monitoring and early warning system for geotechnical slopes based on Internet of Things technology. This system can not only achieve real-time monitoring of the stability of geotechnical slopes, but also predict potential geological disasters through advanced data analysis algorithms, and provide timely feedback to users through an intuitive user interface and early warning system. Such a system will greatly improve the efficiency and accuracy of geotechnical slope monitoring, reduce the risk of geological disasters, and has important practical value and broad market application prospects. This system will bring revolutionary changes to the field of geotechnical slope monitoring by integrating the latest sensor technology, data processing algorithms, cloud computing platforms, and user interface designs. Summary of the Invention

[0006] The object of the present invention is to provide a geotechnical slope intelligent monitoring and early warning system and method based on Internet of Things technology. This system integrates a variety of sensor devices to monitor various key parameters of geotechnical slopes in real time, predicts potential landslide and settlement problems through advanced data analysis algorithms, and provides an intuitive user interface and early warning system to improve the safety and economic benefits of slope engineering.

[0007] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0008] A geotechnical slope intelligent monitoring and early warning system based on Internet of Things technology, including a sensor module, a data acquisition module, a data analysis module, a cloud computing module, a data visualization module, a three-dimensional model display module, an intelligent inspection module, an early warning module, a mobile management module, a system security module, a maintenance and diagnosis module, and a user management module, wherein:

[0009] The sensor module is used to monitor the key parameters of the geotechnical body in real time;

[0010] The data acquisition module is used to collect monitoring data from the sensor module and perform preprocessing to generate preprocessed real-time monitoring data;

[0011] The data analysis module is used to analyze the preprocessed real-time monitoring data of the data acquisition module and generate real-time monitoring data analysis results;

[0012] The cloud computing module is used to store and analyze the real-time monitoring data analysis results from the data analysis module;

[0013] The data visualization module is used to obtain the real-time monitoring data analysis results from the cloud computing module and generate custom reports and historical simulation data;

[0014] The three-dimensional model display module is used to construct a three-dimensional model of the slope, superimpose the real-time monitoring data analysis results and historical simulation data on the three-dimensional model, and accurately map the image content in the real-time video stream to the corresponding surface of the three-dimensional model, so that the three-dimensional model has a real-time dynamic visual presentation;

[0015] The intelligent inspection module is used to allow users to perform inspection and roaming in a virtual scene through virtual reality or augmented reality technology in the three-dimensional model, and view the location, status and surrounding environment of the equipment;

[0016] The early warning module is used to automatically trigger an early warning mechanism when abnormal data is detected according to the preprocessed real-time monitoring data, and evaluate the risk level of the geotechnical slope in real time;

[0017] The mobile management module is used to allow users to remotely access the system through mobile devices for real-time monitoring and management;

[0018] System security module, used to ensure the security of data collection, transmission and storage, preventing unauthorized access and data leakage;

[0019] Maintenance and diagnosis module, used for the daily maintenance and fault diagnosis of the system to ensure the stable operation of the system;

[0020] User management module, used for user permission allocation and user behavior auditing to ensure the compliant use of the system.

[0021] Preferably, in the data collection module, the preprocessing function includes filtering, denoising, data synchronization, and data format conversion, where:

[0022] Filtering uses a low-pass filter to remove high-frequency noise;

[0023] Denoising uses statistical methods to identify and process outliers, reducing random fluctuations and outliers in the data;

[0024] Data synchronization makes the data time from different sensors consistent;

[0025] Data format conversion unifies the data from different sensors into a format recognizable by the system.

[0026] Preferably, the data analysis module includes sampling and interpolation functions, machine learning algorithms, and statistical analysis tools, where:

[0027] Sampling and interpolation functions, used to convert the preprocessed real-time monitoring data into regular time-interval sequence data for in-depth analysis by the Transformer model;

[0028] Machine learning algorithms use the Transformer model for pattern recognition, anomaly detection, and short-term landslide displacement prediction;

[0029] Statistical analysis tools, used to perform statistical analysis on the collected data and calculate statistical parameters such as mean and variance.

[0030] Preferably, the data visualization module includes custom report generation and historical data trend analysis, where:

[0031] Custom report generation is used for users to customize the content and format of reports according to their needs, including graphics, tables, and text;

[0032] Historical data trend analysis is used to perform trend judgment and simulation on historical data and output historical simulation data.

[0033] Preferably, the intelligent inspection module supports voice control and gesture recognition, and users can control the inspection process through voice commands or gestures.

[0034] Preferably, the early warning module includes a multi-level early warning mechanism, which takes different response measures according to different risk levels, as well as real-time risk assessment.

[0035] A method for intelligent monitoring and early warning of geotechnical slopes based on Internet of Things technology is realized according to an intelligent monitoring and early warning system for geotechnical slopes based on Internet of Things technology, and includes the following steps:

[0036] S1, System startup and initialization;

[0037] S2, Real-time collection of monitoring data of the rock and soil mass and preprocessing to generate preprocessed real-time monitoring data;

[0038] S3, Data analysis of the preprocessed real-time monitoring data to generate real-time monitoring data analysis results;

[0039] S4, Visual display of the real-time monitoring data analysis results to generate custom reports and historical simulation data;

[0040] S5, Overlay the real-time monitoring data analysis results and historical simulation data on the 3D model, and combine with the video surveillance footage for real-scene fusion display;

[0041] S6, Intelligent inspection in the 3D model;

[0042] S7, According to the preprocessed real-time monitoring data, automatically trigger the early warning mechanism when abnormal data is detected, and real-time evaluate the risk level of the geotechnical slope.

[0043] Preferably, the specific steps of S5 are as follows:

[0044] S501, Construct a 3D slope model;

[0045] S502, Overlay the real-time monitoring data analysis results and historical simulation data on the 3D model;

[0046] S503, Accurately map the image content in the real-time video stream to the corresponding surface of the 3D model, so that the 3D model has a real-time dynamic visual presentation.

[0047] The beneficial effects of the present invention are as follows:

[0048] 1. Improve the real-time and accuracy of monitoring: By integrating a variety of high-precision sensors and data acquisition modules, this system can continuously monitor the key parameters of geotechnical slopes. The advanced algorithm based on the Transformer model ensures the accuracy and reliability of the data. This real-time and accurate data acquisition and analysis ability enables the system to detect tiny changes in the rock and soil mass in a timely manner, thus early warning potential landslide and settlement problems, and significantly improving the real-time and accuracy of monitoring.

[0049] 2. Enhance early warning and response capabilities: The built-in multi-level early warning mechanism of the system can quickly trigger an early warning when abnormal data is detected, and notify relevant personnel through multiple methods such as text messages, emails, and application push notifications. Combined with the intelligent inspection module, the system can provide more accurate risk assessment and deformation prediction, enabling management personnel to quickly take corresponding preventive measures according to the early warning level, thus effectively avoiding the occurrence of geological disasters and ensuring project safety and personnel safety.

[0050] 3. Improve user experience and system management efficiency: The data visualization module and 3D model display module of this system provide intuitive data display and simulation functions, enabling non-professional personnel to easily understand and analyze monitoring data. The use of the mobile management module allows management personnel to access the system through mobile devices at any location and at any time for real-time monitoring and management. The user management module ensures the compliant use of the system and data security. The integration of these functions greatly improves the user experience and the efficiency of system management, making the geotechnical slope monitoring work more convenient and efficient. Description of the Drawings

[0051] Figure 1 It is the overall structural block diagram of the intelligent monitoring and early warning system for geotechnical slopes of the present invention.

[0052] Figure 2 It is the flowchart of the intelligent monitoring and early warning method for geotechnical slopes of the present invention. Detailed Embodiments

[0053] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention.

[0054] As Figure 1 shown, an intelligent monitoring and early warning system for geotechnical slopes based on Internet of Things technology includes a sensor module, a data acquisition module, a data analysis module, a cloud computing module, a data visualization module, a 3D model display module, an intelligent inspection module, an early warning module, a mobile management module, a system security module, a maintenance and diagnosis module, and a user management module.

[0055] The sensor module is used to continuously monitor key parameters such as the displacement, pressure, stress, and strain of the geotechnical body.

[0056] The data acquisition module is used to collect monitoring data from the sensor module and perform preprocessing to generate preprocessed real-time monitoring data.

[0057] The data analysis module is used to analyze the preprocessed real-time monitoring data of the data acquisition module to generate real-time monitoring data analysis results.

[0058] Cloud computing module, used to store and analyze the real-time monitoring data analysis results from the data analysis module.

[0059] Data visualization module, used to obtain the real-time monitoring data analysis results from the cloud computing module, and generate customized reports and historical simulation data.

[0060] 3D model display module, used to construct a 3D slope model, overlay the real-time monitoring data analysis results and historical simulation data on the 3D model, and accurately map the image content in the real-time video stream to the corresponding surface of the 3D model, enabling the 3D model to have a real-time dynamic visual presentation.

[0061] Intelligent inspection module, allowing users to perform inspection tours in a virtual scene through virtual reality or augmented reality technology in the 3D model, and view the location, status and surrounding environment of the equipment.

[0062] Early warning module, used to automatically trigger an early warning mechanism when abnormal data is detected based on the preprocessed real-time monitoring data, and real-time evaluate the risk level of the geotechnical slope.

[0063] Mobile management module, used to allow users to remotely access the system through mobile devices for real-time monitoring and management.

[0064] System security module, used to ensure the security of data collection, transmission and storage, and prevent unauthorized access and data leakage.

[0065] Maintenance and diagnosis module, used for the daily maintenance and fault diagnosis of the system to ensure the stable operation of the system.

[0066] User management module, used for user permission allocation and user behavior auditing to ensure the compliant use of the system.

[0067] (1) Sensor module: In the intelligent monitoring and early warning system for geotechnical slopes, the sensor module plays a core role in data collection. The sensor module integrates a variety of high-precision sensors for real-time monitoring of key parameters such as geotechnical displacement, pressure, stress, and strain. High-precision sensors include but are not limited to strain gauges, vibrating wire pressure sensors, inclinometers, temperature sensors, humidity sensors, and pore water pressure sensors. The following are the models and parameters of some key sensors:

[0068] The SM-5 series vibrating wire surface strain gauges are deployed. These strain gauges are designed specifically for measuring the strain on the surface of the internal structure of rock and soil masses, featuring high sensitivity and stability, with a measuring range of 3000 με and a collection accuracy of ±0.5% F.S. The VH vibrating wire anchor cable load cell is used to measure the force of the anchor cable, with a collection accuracy of ±0.5% F.S. The PISA-M vertical fixed inclinometer is used to continuously and unattendedly measure the horizontal displacement inside soil bodies, rock bodies, and structural planes, with a measuring angle range of ±10°, a collection resolution of 0.0025°, and a repeatability of ±0.006°. The TH series temperature sensors are used to monitor the temperature of rocks and soil bodies, with a collection accuracy of 0.1°C. The SCYG318 piezometer is used to measure the pore water pressure inside rock and soil masses, with a measuring range of -100 kPa to 20 MPa and an accuracy of ±0.5% F.S.

[0069] (2) Data Acquisition Module: The data acquisition module is a key component in the intelligent monitoring and early warning system for geotechnical slopes. It is responsible for collecting monitoring data from the sensor module and performing preliminary processing. The data acquisition module can automatically eliminate and identify abnormal data, improving the accuracy of data collection. When faced with uncontrollable factors on-site, such as the monitoring points being blocked, misaligned, or damaged, this module can automatically screen and analyze abnormal data, repeatedly observe abnormal points, and eliminate error-type data. The STM32F103 single-chip microcomputer is used as the core processor of the lower computer system. The CPU frequency can reach up to 72 MHz at most. It is packaged with LQFP48, has 3 USART serial ports, 2 groups of 10-channel 12-bit synchronous ADC (Analog-to-Digital Converter), the RAM (Random Access Memory) can write 20K of information, and the FLASH memory is 64K, meeting the compilation requirements. The data acquisition module is a device for collecting data from various distributed sensors, with functions of automatic measurement, signal processing, control, and wireless communication, and can operate reliably for a long time in harsh outdoor environments. The overall device adopts a low-power design. The built-in battery plus the solar panel of the protection box can ensure operation at a frequency of once a day for six years. The number of channels: 4 or 8 channels, the range is 450 - 6000 Hz, the accuracy is ±0.002% for vibrating wire sensors, the display resolution is ±0.001% for vibrating wire sensors, the temperature resolution is 0.1 °C, and the memory can store 63,500 records in the non-volatile memory. The data acquisition module can match the parameter requirements of various types of sensors and meet the data collection needs in different monitoring scenarios. The data acquisition module converts the analog signal into a digital signal through the ADC (Analog-to-Digital Converter), which is one of the core links in data acquisition. The digital signal is then stored in the central processing unit or a dedicated storage device, providing a basis for subsequent data analysis and processing. The data can be transmitted through the wired communication USB interface or through wireless communication technology. Once the data is successfully collected and preprocessed, they can be displayed on the screen or used for other purposes, such as further analysis or report generation.

[0070] Data preprocessing functions, including filtering, denoising, data synchronization, and data format conversion functions. Among them: Filtering uses a low-pass filter to remove high-frequency noise; Denoising uses statistical methods to identify and process outliers, reducing random fluctuations and outliers in the data; Data synchronization makes the data time from different sensors consistent; Data format conversion unifies the data from different sensors into a format recognizable by the system, facilitating analysis by the data analysis module.

[0071] (3) The data analysis module includes sampling and interpolation functions, machine learning algorithms, and statistical analysis tools. The sampling and interpolation functions are used to convert the preprocessed real-time monitoring data into regular time-interval sequence data for in-depth analysis by the Transformer model. The machine learning algorithm uses the Transformer model for pattern recognition, anomaly detection, and short-term landslide displacement prediction. The statistical analysis tool is used to perform statistical analysis on the collected data and calculate statistical parameters such as mean and variance.

[0072] The data analysis module is responsible for receiving the real-time monitoring data preprocessed by the data collection module and conducting in-depth analysis using advanced machine learning algorithms. In terms of technical implementation, the constructed Transformer model combines the Temporal Convolutional Network (TCN) and the Transformer decoder. The inputs include displacement and rainfall sequences, and the output is the displacement prediction results for the next three days. The self-attention mechanism of the model is implemented through the calculation formula where d k is the number of dimensions of the input sample, Q is the query vector, K is the key vector, V is the value vector, and T is the matrix transpose. This normalization process helps to make the attention scores more approximate to the normal distribution, thus optimizing the training effect of the model. The data analysis module also integrates a statistical analysis tool for calculating statistical parameters such as mean and variance. The mean calculation formula is where x i is the random variable and n is the number of samples. The variance calculation formula is where x i is the random variable, n is the number of samples, and μ is the mean. These statistical parameters not only help to understand the central tendency and dispersion degree of the data but also provide a solid foundation for further data analysis. Through the application of these technologies, the data analysis module can provide accurate deformation prediction and risk assessment for the geotechnical slope, thereby improving the construction safety and efficiency. The combination of these technologies not only improves the timeliness and accuracy of early warning but also realizes an effective response to different risk levels, providing a strong guarantee for the safety of the geotechnical slope.

[0073] (4) Cloud computing module: The cloud computing module has the ability to automatically expand computing resources, dynamically adjust computing resources according to data processing requirements, and has high availability and fault tolerance mechanisms to ensure that the system can still operate stably when some hardware or software fails.

[0074] The cloud computing module adopts MySQL database technology and runs on the MySQL database version 8.0, which provides advanced features such as descending indexes and enhanced JSON functions, making it more efficient to handle complex queries and data analysis. To ensure that the system can process and store a large amount of monitoring data, the database server is configured with 32GB of memory and 1TB of storage space, and the InnoDB buffer pool size is set to 24GB to optimize random read and write performance. On the server side, the Apache HTTP server, version 2.4.41, is used. It is an open-source and high-performance HTTP server that provides intrusion detection and defense functions for web applications through the mod_security module, enhancing the security of the system. For backend development, PHP 7.4 and Python 3.8 are adopted. Both languages have good compatibility with MySQL and have rich database operation libraries and frameworks, such as PDO for PHP and PyMySQL for Python, making database access and operation more flexible and efficient. To adapt to different scales of data processing requirements, the cloud computing module is deployed on the AWS cloud platform and utilizes its auto-scaling service. This service can automatically increase or decrease the number of EC2 instances according to CPU and memory usage rates, and the CPU usage rate threshold is set to 75%. In addition, the database adopts a master-slave replication architecture, where the master database processes write operations and the slave database processes read operations, ensuring high data availability. An RDS multi-availability zone deployment is also set up to prevent single points of failure and ensure the continuous availability of the service. To ensure data security, a data backup strategy is implemented, including full backups every week and incremental backups every day. The full backups are retained for 30 days, and the incremental backups are retained for 7 days. All backups are stored on Amazon S3, which has high durability and high reliability. The recovery time objective (RTO) is less than 1 hour, ensuring rapid recovery in case of data loss or corruption.

[0075] (5) Data visualization module: The data visualization module supports custom report generation and historical data trend analysis. Custom report generation is used for users to customize the content and format of reports according to their needs, including graphs, tables, and text; historical data trend analysis is used to make trend judgments and simulations on historical data and output historical simulation data.

[0076] The data visualization module converts complex data analysis results into intuitive charts and graphs, greatly enhancing the readability and comprehensibility of the data. This module supports multiple chart types, including line charts, bar charts, pie charts, and scatter plots, to meet the needs of different types of data display. For example, for displacement data, a line chart is usually used to show the trend over time. The horizontal axis represents time (in days), and the vertical axis represents the displacement amount (in millimeters). Users can customize the content and format of the report according to their needs, including selecting specific data fields, time ranges, and chart types. The time range for report generation can be customized. Users can choose data from the past 1 day, 1 week, or 1 month to generate reports. Such flexibility enables users to conduct in-depth analysis for specific time periods. In addition, the module has built-in historical data trend analysis tools that can statistically analyze the data over a period of time to identify potential trends and patterns. For example, by analyzing the monthly average displacement data for the past year, the change rate of the monthly average displacement can be calculated. This change rate is in millimeters per month and helps predict future displacement trends. At the technical level, the data visualization module uses the D3.js library, which is a powerful chart-drawing tool based on Web standards and can generate complex interactive charts and graphs. The module supports a screen resolution of 1920x1080 pixels to ensure the clarity of the charts on high-resolution screens. At the same time, the module provides user interaction functions, such as tooltips. When the user hovers the mouse over a data point on the chart, the specific value of that point and other relevant information will be displayed. In addition, the module also supports click events on data points. After the user clicks on a data point, the detailed data record of that point can be viewed, enhancing the user experience.

[0077] (6) 3D model display module: The 3D model display module can achieve dynamic simulation and real-time data overlay, directly overlaying sensor data on the 3D model in the form of graphics or colors to intuitively display the real-time state of the geotechnical slope.

[0078] The following are the specific implementation details of this module, including technical parameters and numerical parameters:

[0079] Using BIM technology, combined with Lizheng software, Autodesk Civil 3D, and Revit software tools, a 3D geological model of the geotechnical slope is established. These models not only include the display of stratigraphic information but also enable the 3D display of the exploration report, improving the technical level of exploration and design.

[0080] The Triangulated Irregular Network (TIN) model is a commonly used 3D surface modeling technique that constructs the 3D shape of the surface by connecting points scattered in space. A texture mapping technique with adjustable scale is designed to achieve the realistic visualization of geological bodies and the visualization information query of strata and soil samples. This technique can adjust the scale of the texture according to the user's needs to achieve a more realistic visual effect.

[0081] Overlay the analysis results of real-time monitoring data and historical simulation data on the 3D model. Through advanced algorithms and software, accurately map the image content in the real-time video stream to the corresponding surface of the 3D model, enabling the 3D model to have a real-time dynamic visual presentation. For example, the video from a surveillance camera can be fitted to the walls and passageways of the 3D building model, providing users with a more realistic and intuitive situation awareness. In 3D geological modeling, the TIN model is adopted, which not only ensures the accuracy of the calculation results but also improves the work efficiency.

[0082] (7) Intelligent inspection module: The intelligent inspection module supports voice control and gesture recognition. Users can control the inspection process through voice commands or gestures, improving the convenience and efficiency of inspection.

[0083] The intelligent inspection module integrates virtual reality (VR) or augmented reality (AR) technology, enabling users to conduct immersive inspection tours in virtual scenarios, greatly enhancing the convenience and efficiency of inspection work. The module supports advanced voice control and gesture recognition functions. The comprehensive recognition rate of the voice recognition technology exceeds 95%. Even in a noisy environment, high-accuracy voice recognition can be maintained through iFlytek's core noise reduction algorithm to achieve precise control of the inspection process. In addition, the module adopts edge computing technology and supports pure offline application scenarios, enabling fast and stable voice recognition without network support, suitable for various complex terrains and outdoor high-altitude operations. This hardware-software integrated solution not only improves the safety of inspection but also ensures high recognition accuracy for keywords in vertical fields through keyword customization and continuous optimization, meeting the specific needs of intelligent inspection in different industries.

[0084] (8) Early warning module: The early warning module includes a multi-level early warning mechanism that takes different response measures according to different risk levels, as well as real-time risk assessment. Based on the preprocessed real-time monitoring data and historical data (data monitored during landslides and other disasters in previous other projects), the risk level of the rock and soil slope is evaluated in real time.

[0085] When the early warning module detects abnormal data, it automatically triggers the early warning mechanism, including multiple notification methods such as text messages, emails, and application push notifications, and includes a multi-level early warning mechanism that takes different response measures according to different risk levels. The system should clarify the early warning management mechanism corresponding to each early warning level, including the early warning information feedback mechanism and the early warning information disposal mechanism. The early warning information feedback can be informed to relevant departments and personnel by means of system information release, mobile phone text messages, emails, and audible and visual alarms. The system should automatically and real-time feedback the early warning information to managers at all levels according to the early warning information feedback mechanism, and the managers should promptly handle the early warning information according to the early warning information disposal mechanism.

[0086] (9) Mobile management module: The mobile management module allows users to remotely access the system through mobile devices for real-time monitoring and management, and supports multi-platform operations, including iOS, Android, and Windows systems.

[0087] (10) System security module: The system hardware should meet the lightning protection requirements, and the on-site equipment and hardware should adapt to the temperature, humidity, waterproof, and dustproof requirements of the construction site environment. Uninterruptible power supplies should be configured for important equipment such as data collectors and data center servers. Information transmission equipment should meet the network access standards of the installation site. The system software should meet the requirements of information security and stable operation.

[0088] (11) Maintenance and diagnosis module: Used for the daily maintenance and fault diagnosis of the system. The system self-check cycle is set to once a week, and the fault response time is set within 2 hours. The system runs stably and responds to faults in a timely manner, ensuring the continuity and reliability of the monitoring data.

[0089] (12) User management module: The user management module is used for user permission allocation and user behavior auditing. Users are divided into three levels, with different permissions for each level, and the audit log retention time is set to 1 year. The compliant use of the system is guaranteed, and user behavior is effectively monitored and managed.

[0090] As Figure 2 shown, a geotechnical slope intelligent monitoring and early warning method based on Internet of Things technology includes the following steps:

[0091] S1, System startup and initialization.

[0092] After the system starts, it first conducts a self-check of the sensor network module. This process includes hardware diagnostic tests, communication protocol tests, and power supply detection of sensors to ensure that all sensors and modules are in the correct working state. Subsequently, the system loads the preset configuration parameters, which include sensor calibration data and communication protocols, laying a foundation for subsequent data collection and analysis. This step is the starting point for the operation of the entire system, providing the necessary initialization conditions for subsequent steps.

[0093] S101, after the system starts, the self-check of the sensor network module is carried out first.

[0094] The self-check of the sensor network module includes hardware diagnostic tests, communication protocol tests, and power supply detection of sensors to ensure that the physical connections, power supplies, and communication links of all sensors are in normal states.

[0095] S102, the system loads the preset configuration parameters.

[0096] The configuration parameters include calibration data of sensors, sampling frequencies, measurement ranges, and communication protocols to ensure the consistency and accuracy of data acquisition.

[0097] S103, the system performs network connection tests and user interface initialization.

[0098] The system performs network connection tests to check the network connection with the central monitoring system to ensure that data can be transmitted in real time. This includes ping tests, port scans, and packet transmission tests to confirm the stability of the network connection and whether the data transmission latency is within an acceptable range. At the same time, the user interface is also initialized, including the login page, real-time data display page, and system status monitoring page, supporting multiple languages and adapting to different user needs. The system log records every step during the startup process, including timestamps, operator IDs, and operation results, which provides a detailed historical record for fault troubleshooting and performance monitoring. Finally, when all initialization steps are completed, the system enters the ready state, the indicator lights of all sensors and modules show green lights, and the "System Ready" prompt is displayed on the user interface, indicating that the system is fully operational and ready to start data acquisition and analysis tasks.

[0099] S2, real-time collect the monitoring data of the rock and soil mass and perform preprocessing to generate the preprocessed real-time monitoring data.

[0100] S201, the data acquisition module real-time collects the key parameters of the rock and soil mass.

[0101] After the system initialization is completed, the sensor network module is activated, the sensor module starts to work, and real-time collects the key parameters such as displacement, pressure, stress, and strain of the rock and soil mass. The data acquisition cycle is set to 1 minute, and continuous acquisition is carried out for 72 hours to ensure the continuity and real-time nature of the data. These raw data will serve as the basis for downstream data analysis and processing.

[0102] Data acquisition involves the collaborative work of multiple sensors, including strain gauges, vibrating wire pressure sensors, inclinometers, temperature sensors, humidity sensors, acoustic emission sensors, and pore water pressure sensors. These sensors are carefully deployed at various key positions within the rock slope to ensure comprehensive monitoring of the geotechnical mass. Taking strain gauges as an example, they are usually installed on the surface or inside the structural plane to measure the tiny deformations caused by stress. Vibrating wire pressure sensors use the frequency change of the vibrating wire to measure pressure changes and are widely used in geotechnical engineering due to their high precision and stability. Inclinometers are used to monitor the inclination of the structure, while temperature and humidity sensors monitor the ambient temperature and humidity respectively, and these environmental factors have a direct impact on the stability of the geotechnical structure. Acoustic emission sensors can detect the acoustic wave signals generated during the stress release inside the geotechnical mass, which is crucial for predicting the stability of the rock slope. Pore water pressure sensors are used to monitor the pore water pressure in the soil, which is crucial for evaluating soil saturation and permeability. During the data acquisition process, each sensor will work according to the preset sampling frequency (e.g., 1Hz) to ensure the continuity and real-time nature of the data. The collected data will be transmitted to the central processing unit via wired or wireless means.

[0103] S202, preprocess the key parameters of the collected geotechnical mass.

[0104] Data preprocessing is an important step in the intelligent monitoring and warning system of rock slopes to ensure data accuracy and consistency: preprocess the collected raw data, and the preprocessing includes filtering, denoising, data synchronization, and data format conversion. The preprocessed data will be used in the downstream data analysis module to provide a clean and accurate data basis for more complex analysis.

[0105] In the filtering process, a low-pass filter is used to remove high-frequency noise. According to the technical standards of geotechnical engineering monitoring, the cut-off frequency of the filter is set to 10Hz to ensure effective removal of high-frequency interference while retaining useful low-frequency signals.

[0106] In the denoising process, statistical methods are used to identify and process outliers. According to the standard, the 3σ criterion is usually adopted for outlier processing, that is, any reading exceeding the mean ± 3 times the standard deviation is regarded as an outlier and processed. This step helps to reduce the random fluctuations and outliers in the data, which may be caused by instantaneous errors of the sensors or changes in environmental factors. The denoising process can improve the accuracy and consistency of the data.

[0107] Data synchronization ensures that the data from different sensors is consistent in time. According to the technical specifications, the timestamp accuracy of all sensors should reach the millisecond level to ensure the accuracy of data synchronization.

[0108] In the data format conversion stage, the data from different sensors is uniformly converted into a format that the system can recognize and process. According to the technical specification of the dedicated WLAN communication module for sensor devices, the data format should be unified into JSON or XML format for further processing by the system. Ensure the accuracy and consistency of the data to provide high-quality input for data analysis and early warning.

[0109] S3. Conduct data analysis on the preprocessed real-time monitoring data to generate the analysis results of the real-time monitoring data.

[0110] The data analysis module receives the preprocessed real-time monitoring data from the data acquisition module. This data includes key parameters such as the displacement, pressure, stress, and strain of the rock and soil mass. Data analysis is to conduct in-depth analysis on the preprocessed real-time monitoring data to predict the stability of the rock and soil slope and potential landslide displacement. The data analysis results will be used in the data visualization and early warning modules to provide a scientific basis for decision-making.

[0111] The following is a specific description of the data analysis process, including specific parameters and algorithm details:

[0112] The data analysis module has functions such as sampling and interpolation, ensuring the regularity and integrity of the data, and providing high-quality input for the Transformer model. For example, through resampling and interpolation, the preprocessed real-time monitoring data is converted into regular time interval sequence data for in-depth analysis by the subsequent Transformer model.

[0113] The data analysis module uses a machine learning algorithm based on the Transformer model for in-depth analysis. The historical data set used for model training is the preprocessed data samples in S2. The number of self-attention layers of the model is set to 6 layers to predict the short-term displacement of the landslide. The Transformer model calculates the input and output representations through the self-attention mechanism and has a stronger expressive ability for long-term dependencies, especially suitable for the long-distance dependency problem brought by the high time resolution in the short-term landslide displacement prediction task. The parameter settings of the model include but are not limited to: the number of model layers, the number of heads, and the model dimension. These parameters are adjusted according to the specific task and data set. For example, the model may contain 6 Transformer encoder layers, each layer has 8 attention heads, and the model dimension is 512. This is a common configuration, but the specific parameters will be adjusted according to the actual application. The self-attention mechanism of the Transformer model can capture the information of all positions in the input sequence and calculate the importance of each position for the current position, thus obtaining a more accurate representation. This mechanism is particularly suitable for predicting the short-term displacement of the landslide because it can focus on the key information near the displacement peak and heavy rainfall, improving the prediction accuracy and reliability of the model.

[0114] In addition to the Transformer-based model, the data analysis module also uses statistical analysis tools to calculate statistical parameters such as mean and variance. These parameters help to understand the central tendency and dispersion of the data, providing a basis for further analysis. These statistical parameters can be used for feature extraction and model training, improving the model's sensitivity to data changes and the accuracy of predictions.

[0115] Through data analysis, the system can accurately predict the stability of geotechnical slopes and potential landslide displacements, providing strong data support for engineering safety.

[0116] The analyzed data is stored in a database by the cloud computing module. The stored data can be used for historical trend analysis and future data mining.

[0117] The data storage uses MySQL database technology, running on a MySQL database version 8.0 to ensure data security and accessibility. Utilize its new features such as descending indexes, common table expressions (CTEs), and JSON functions, as well as the transaction processing and row-level locking capabilities of the InnoDB storage engine to handle high-concurrency and large-data-volume scenarios. The database server is configured with 32GB of memory and at least 1TB of storage space to meet the storage requirements of a large amount of monitoring data. At the same time, use the Apache server as the backend, combined with the PHP and Python development languages, utilize their database operation libraries and frameworks, PDO in PHP and PyMySQL in Python, as well as RESTful API design and JSON data exchange format to ensure the flexibility and scalability of data interaction. For data security, implement SSL encrypted connections, strict network security measures, permission control, and real-time monitoring and logging. Any database access and operation will be recorded and audited to facilitate tracking and analysis of potential security issues, thus ensuring data security and accessibility.

[0118] S4, visually display the results of real-time monitoring data analysis, generating custom reports and historical simulation data.

[0119] The results of data analysis are visually displayed in the form of charts and graphs through the data visualization module. The data visualization module converts complex data analysis results into intuitive charts and graphs, facilitating user understanding and operation, and quickly grasping key information.

[0120] The following is a detailed description of the data visualization process, including specific parameters and algorithm details:

[0121] First, the data visualization module uses a variety of chart display techniques, such as scatter plots, heat area charts, and treemaps, to visually display the monitoring data of rock and soil masses. For example, scatter plots are used to show the relationship between two continuous variables. By observing the positions of data points on a plane, one can intuitively observe their distribution, trends, and potential outliers. In terms of technical implementation, this module supports custom report generation, and users can customize the content and format of reports according to their needs. This includes selecting different chart types, color schemes, and data display methods. For example, users can choose to use the Seaborn library, which provides built-in themes, color palettes, and function tools based on Matplotlib, making it easier to create graphs.

[0122] In addition, the data visualization module also provides historical data trend analysis functions. By analyzing historical data, users can identify correlations or other patterns between variables, such as positive correlations and negative correlations. In terms of technical parameters, the data visualization module supports the visualization of large-scale numerical data and may use distributed computing and parallel algorithms for visualization; for spatio-temporal data, it may be necessary to use geographic information systems (GIS) or other map-based visualization methods; for text data, word cloud and topic model visualization techniques may be used.

[0123] Finally, the data visualization module also takes into account the user experience and provides interactive functions, such as clicking or hovering over a chart to display hidden information and detailed information about data points.

[0124] S5. Overlay the analysis results of real-time monitoring data and historical simulation data onto the 3D model, and combine with the video surveillance footage for real-scene fusion display.

[0125] The 3D model display module constructs a 3D model of the slope and combines it with the video surveillance footage for real-scene fusion display. The model resolution is set to 1080p, and the fusion accuracy between the video surveillance footage and the model reaches the pixel level. This step combines the data analysis results with the 3D model to provide an intuitive real-time status display.

[0126] 3D model display: The 3D model display module constructs a 3D model related to the project and combines it with the video surveillance footage for real-scene fusion display, realizing dynamic simulation and real-time data overlay, and intuitively displaying the real-time status of the rock and soil slope.

[0127] S501. Construct a 3D model of the slope.

[0128] This module first conducts 3D modeling of the geotechnical slope through the method of constructing a real-scene 3D model. This process involves two major categories of 3D models, namely surface 3D models and solid 3D models, including point clouds of terrain and ground surface, surface triangulation models, and 3D surface models of ground features. For example, through data acquisition integrated with multiple aerial sensors and combined with the construction of high-fidelity terrain and ground surface 3D models, accurate modeling of the geotechnical slope is achieved.

[0129] When constructing the 3D model, the parametric modeling method is adopted. This method can automatically establish 3D models of complex facilities and interactively edit and modify them according to design parameters. The construction of the model involves the organic integration and synchronous update of 3D geometric models and their parameter information.

[0130] S502, superimpose the analysis results of real-time monitoring data and historical simulation data onto the 3D model.

[0131] In terms of real-time data superimposition, the 3D model display module supports accessing data from temperature, humidity, water quality, and monitoring sensors to achieve real-time monitoring.

[0132] S503, accurately map the image content in the real-time video stream onto the corresponding surface of the 3D model through advanced algorithms and software, enabling the 3D model to have a real-time dynamic visual presentation.

[0133] S6, conduct intelligent inspection in the 3D model.

[0134] The intelligent inspection module allows users to conduct inspection roaming in the virtual scene through virtual reality (VR) or augmented reality (AR) technology, greatly improving the convenience and efficiency of inspection. The refresh rate of the virtual reality device is set to 90Hz, and the gesture recognition accuracy rate exceeds 99%. This step improves the efficiency and safety of inspection through advanced technical means.

[0135] The intelligent inspection module adopts the Simultaneous Localization and Mapping (SLAM) technology. The SLAM technology enables the inspection robot to perform autonomous localization and mapping in an unknown environment. The implementation of this module is based on the Extended Kalman Filter (EKF) and the Rao-Blackwellised Particle Filter (RBPF), which are the mainstream methods for solving the localization (SLAM) problem at present. In the EKF SLAM algorithm, optimizations are made for problems such as high estimation noise, large data association errors, and low computational efficiency. In addition, in the prediction stage of the RBPF SLAM algorithm, the Extended Kalman Filter is used to fuse the data of the gyroscope and the compass to obtain a more accurate pose angle, thereby improving the modeling accuracy and real-time performance of the algorithm. The intelligent inspection module also adopts the HPSO-ACO algorithm, which is an algorithm combining Hybrid Particle Swarm Optimization (HPSO) and Ant Colony Optimization (ACO), and is used to optimize the path of the inspection robot. The HPSO-ACO algorithm saves more computational resources in optimizing the closed curve detection path, has a faster calculation speed, improves the detection efficiency of the robot, and reduces the detection cost. This algorithm realizes the path optimization of the inspection robot under multi-objective conditions by continuously optimizing the path until the optimal path is found. The intelligent inspection module also integrates computer vision algorithms to process the images and videos in the monitoring data through object detection, face recognition, and image segmentation to discover the objects and patterns therein. For example, in intelligent inspection, computer vision algorithms can be used to identify the faulty parts in the equipment.

[0136] S7 automatically triggers the early warning mechanism when abnormal data is detected based on the preprocessed real-time monitoring data, and evaluates the risk level of the geotechnical slope in real time.

[0137] The early warning module automatically triggers the early warning mechanism when abnormal data is detected, including multiple notification methods such as text messages, emails, and application push notifications. The early warning threshold is set to a displacement change exceeding 5 mm / 24 hours. This step ensures that relevant personnel can be notified and take actions in time before potential risks occur.

[0138] The early warning module adopts a threshold-based triggering mechanism. When the monitored data exceeds the preset safety threshold, the system will automatically start the early warning program. These thresholds are determined based on historical data and statistical analysis and can reflect the safety status of the geotechnical slope. For example, if the displacement monitoring data exceeds the set threshold (5mm / 24 hours), the system will consider this an abnormal situation and immediately notify the relevant personnel. The system implements a multi-level early warning mechanism and takes different response measures according to different risk levels. This mechanism refers to the "meteorological early warning + response measures" intelligent early warning linkage mechanism. By automatically matching meteorological early warnings with the early warning response measures of industry departments, townships (sub-districts), it forms a "meteorological early warning + response measures" work instruction. In geotechnical engineering, this means that when the system detects a certain risk level, it will automatically trigger the corresponding preset response measures, such as evacuating personnel, increasing support or emergency reinforcement. The system is configured with a short message gateway and sends short messages through the URL interface, supporting instant message push. The prefix of the short message content is

Early Warning Notice

Claims

1. An intelligent monitoring and early warning system for rock and soil slopes based on Internet of Things technology, characterized in that: It includes sensor module, data acquisition module, data analysis module, cloud computing module, data visualization module, 3D model display module, intelligent inspection module, early warning module, mobile management module, system security module, maintenance and diagnosis module, and user management module, among which: Sensor modules for real-time monitoring of key parameters of rock and soil masses; The data acquisition module is used to collect monitoring data from the sensor module, perform preprocessing, and generate preprocessed real-time monitoring data; A data analysis module is used to analyze the real-time monitoring data preprocessed by the data acquisition module and generate real-time monitoring data analysis results; A cloud computing module, used for storing and analyzing the real-time monitoring data analysis results from the data analysis module; Data visualization module, used to obtain real-time monitoring data analysis results from the cloud computing module and generate customized reports and historical simulation data; The 3D model display module is used to construct the 3D model of the slope, superimpose the real-time monitoring data analysis results and historical simulation data on the 3D model, and accurately map the image content in the real-time video stream to the corresponding surface of the 3D model, so that the 3D model has real-time dynamic visual presentation; The intelligent inspection module is used to allow users to inspect and roam in the virtual scene through virtual reality or augmented reality technology in the 3D model to view the location, status and surrounding environment of the equipment; The early warning module is used to automatically trigger the early warning mechanism when abnormal data is detected based on the pre-processed real-time monitoring data, and to evaluate the risk level of the geotechnical slope in real time; Mobile management module, used to allow users to remotely access the system through mobile devices for real-time monitoring and management; System security module, used to ensure the security of data collection, transmission and storage, and prevent unauthorized access and data leakage; Maintenance and diagnosis module, used for daily maintenance and fault diagnosis of the system to ensure stable operation of the system; The user management module is used for user authority allocation and user behavior auditing to ensure the compliance of the system.

2. The intelligent monitoring and early warning system for rock and soil slopes based on Internet of Things technology according to claim 1 is characterized in that: In the data acquisition module, the preprocessing functions include filtering, denoising, data synchronization, and data format conversion, among which: Filtering uses a low-pass filter to remove high-frequency noise; Denoising uses statistical methods to identify and process outliers, reducing random fluctuations and outliers in the data; Data synchronization makes the data from different sensors consistent in time; Data format conversion unifies the data from different sensors into a format that the system can recognize.

3. The intelligent monitoring and early warning system for rock and soil slopes based on Internet of Things technology according to claim 1 is characterized in that: The data analysis module includes sampling and interpolation functions, machine learning algorithms, and statistical analysis tools, among which: Sampling and interpolation functions are used to convert pre-processed real-time monitoring data into regular time series data for in-depth analysis by the Transformer model; The machine learning algorithm uses the Transformer model for pattern recognition, anomaly detection, and short-term landslide displacement prediction; Statistical analysis tools are used to perform statistical analysis on the collected data and calculate the mean and variance statistical parameters.

4. The intelligent monitoring and early warning system for rock and soil slopes based on Internet of Things technology according to claim 1 is characterized in that: The data visualization module includes custom report generation and historical data trend analysis, where: Custom report generation is used for users to customize the content and format of reports according to their needs, including graphics, tables and texts; Historical data trend analysis is used to judge and simulate the trend of historical data and output historical simulation data.

5. The intelligent monitoring and early warning system for rock and soil slopes based on Internet of Things technology according to claim 1 is characterized in that: The intelligent inspection module supports voice control and gesture recognition, and users can control the inspection process through voice commands or gestures.

6. The intelligent monitoring and early warning system for rock and soil slopes based on Internet of Things technology according to claim 1 is characterized in that: The early warning module includes a multi-level early warning mechanism, taking different response measures according to different risk levels, and real-time risk assessment.

7. An intelligent monitoring and early warning method for rock and soil slopes based on Internet of Things technology, characterized in that: The implementation of the intelligent monitoring and early warning system for rock and soil slopes based on the Internet of Things technology according to any one of claims 1 to 6 comprises the following steps: S1, system startup and initialization; S2, real-time acquisition of monitoring data of the rock and soil body and preprocessing, generating preprocessed real-time monitoring data; S3, performing data analysis on the pre-processed real-time monitoring data to generate real-time monitoring data analysis results; S4, visualizes the analysis results of real-time monitoring data and generates customized reports and historical simulation data; S5, superimpose the real-time monitoring data analysis results and historical simulation data on the three-dimensional model, and combine them with the video surveillance screen for real-scene fusion display; S6, intelligent inspection in 3D model; S7, based on the pre-processed real-time monitoring data, automatically triggers the early warning mechanism when abnormal data is detected, and evaluates the risk level of the geotechnical slope in real time.

8. The method for intelligent monitoring and early warning of rock and soil slopes based on Internet of Things technology according to claim 7 is characterized in that: The specific steps of S5 are: S501, constructing a three-dimensional slope model; S502, superimposing the real-time monitoring data analysis results and the historical simulation data onto the three-dimensional model; S503, accurately mapping the image content in the real-time video stream to the corresponding surface of the three-dimensional model, so that the three-dimensional model has real-time dynamic visual presentation.

Images