Real-time highway pavement monitoring system based on Internet of Things

Abstract

The invention relates to the technical field of road surface monitoring, and discloses a road surface real-time monitoring system based on Internet of Things, comprising a sensing layer used for collecting road surface data through a vehicle-mounted device and a fixed sensor node; the communication layer is used for realizing transmission and edge calculation processing of various sensor data; the data processing and storage layer is used for fusing, analyzing and storing the multi-source data; the application layer is used for generating intelligent decision support and visual display, and vehicle-mounted equipment in the sensing layer comprises a multi-mode sensor unit. High-frequency real-time monitoring is achieved by integrating vehicle-mounted equipment and a fixed sensor, intelligent data analysis is conducted through a Bayesian statistical model and a DBSCAN clustering algorithm, the dynamic modeling and prediction capacity for road surface damage is enhanced, the maintenance strategy is automatically evaluated and optimized in cooperation with an intelligent decision support module, and the maintenance efficiency is improved. And meanwhile, road surface monitoring data are visually displayed through the GIS integration module.

Description

Technical Field

[0001] The present invention relates to the technical field of road surface monitoring, and in particular to a real-time monitoring system for a highway road surface based on the Internet of Things. Background Art

[0002] With the rapid development of social economy and the acceleration of urbanization, the demand for transportation is growing. As the main transportation infrastructure, the importance of highway maintenance and management is becoming increasingly prominent. However, traditional highway pavement monitoring methods often rely on manual inspections and regular testing, which is not only inefficient but also difficult to reflect changes in pavement conditions in a timely manner.

[0003] Traditional monitoring technology mainly collects data through fixed sensors, and the data update frequency is low, which makes it impossible to monitor the road conditions in real time. In this way, road damage often continues to deteriorate without being discovered in time, causing serious traffic accidents and economic losses. In addition, manual inspection results often rely on the subjective judgment of inspectors, lack of objectivity, and easily lead to deviations in monitoring results.

[0004] In terms of data processing and analysis, existing technologies usually rely on simple statistical methods, which makes it difficult to conduct in-depth data analysis and trend prediction. This makes the identification and judgment of road damage lack of scientific basis, thus affecting subsequent maintenance decisions. At the same time, the traditional maintenance model cannot be dynamically adjusted according to actual conditions, resulting in inefficient resource allocation and increased maintenance costs.

[0005] In response to the above problems, monitoring technology based on the Internet of Things has gradually emerged, using advanced technologies such as sensors, communication networks and data analysis to achieve real-time monitoring and intelligent analysis of road conditions. However, the existing Internet of Things monitoring system is still insufficient in terms of intelligence, data processing efficiency and resource optimization, and cannot fully meet the needs of modern traffic management.

[0006] Therefore, there is an urgent need for a new type of real-time highway pavement monitoring system that uses advanced technical means to improve the accuracy and real-time nature of monitoring, so as to promptly detect pavement problems, optimize resource allocation, reduce maintenance costs, and ensure traffic safety and smoothness. Summary of the invention

[0007] In view of the shortcomings of the prior art, the present invention provides a real-time highway pavement monitoring system based on the Internet of Things, which solves the problems of insufficient real-time performance, inaccurate data analysis and inefficient resource allocation in traditional highway pavement monitoring methods.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: A real-time monitoring system for highway pavement based on the Internet of Things, comprising:

[0009] The perception layer is used to collect road surface data through vehicle-mounted equipment and fixed sensor nodes;

[0010] Communication layer, used to realize the transmission of various sensor data and edge computing processing;

[0011] Data processing and storage layer, used to integrate, analyze and store multi-source data;

[0012] The application layer is used to generate intelligent decision support and visual display.

[0013] Preferably, the vehicle-mounted equipment in the perception layer includes:

[0014] A multimodal sensor unit, which includes an accelerometer, gyroscope, camera, and GPS module for real-time monitoring of road vibrations, vehicle posture, damage images, and location data;

[0015] Multimodal data fusion unit, which is used to fuse various sensor data through Kalman filtering algorithm;

[0016] The adaptive sampling strategy unit is used to dynamically adjust the data sampling frequency based on the reinforcement learning algorithm to optimize energy consumption and monitoring effect.

[0017] Preferably, the fixed sensor nodes in the perception layer include:

[0018] Strain gauge and vibration sensor units, which detect minute changes in stress and vibration in the road, identifying early signs of damage;

[0019] An energy harvesting unit, which is used to use solar energy or road vibration energy to power fixed nodes;

[0020] The timing synchronization unit is used to make the data collection time of different sensor nodes consistent through the distributed clock synchronization mechanism.

[0021] Preferably, the communication layer includes:

[0022] Hybrid communication network module, which is used to ensure stable and efficient transmission of sensor data;

[0023] Edge computing units are used to perform preliminary data processing and intelligent reasoning at nodes close to data sources.

[0024] Preferably, the data processing and storage layer includes:

[0025] Multi-source data fusion module, which is used to conduct comprehensive analysis and fusion processing of data from vehicle-mounted equipment and fixed nodes. By establishing a Bayesian statistical model, it combines spatiotemporal data to dynamically model and predict road damage, and uses the DBSCAN algorithm to cluster road data and identify abnormal road areas;

[0026] Big data processing unit, which is used to store, manage and analyze large-scale data in real time.

[0027] Preferably, the application layer includes:

[0028] Intelligent decision support module, which is used to automatically evaluate the degree of road damage and generate optimized maintenance strategies based on a multi-layer perceptron neural network model, and set up a multi-level alarm mechanism according to the road damage situation;

[0029] The GIS integration module is used to integrate real-time pavement monitoring data, dynamically display damage locations and damage heat maps on the map, and use Kriging interpolation to generate a high-precision damage prediction model.

[0030] Preferably, the fixed sensor nodes in the perception layer optimize deployment positions through an integer linear programming algorithm to minimize deployment costs and maximize monitoring coverage.

[0031] Preferably, the objective function and constraints of the integer linear programming algorithm are as follows:

[0032] Objective function:

[0033]

[0034] Constraints:

[0035]

[0036] Among them, C i is the deployment cost, x i Deploy variables for nodes, d ij For coverage capability.

[0037] Preferably, the communication layer uses a Markov decision process to optimize the data transmission strategy.

[0038] Preferably, the strategy function formula of the Markov decision process is:

[0039]

[0040] Among them, π is the policy function, s t is the system state, R is the reward function, and γ is the discount factor.

[0041] Working principle: When the system is running, first, at the perception layer, the system uses vehicle-mounted devices and fixed sensor nodes to collect data. The vehicle-mounted devices are equipped with multimodal sensors, including accelerometers, gyroscopes, cameras, and GPS modules. The combination of accelerometers and gyroscopes enables the system to monitor the vibration and posture changes of the vehicle in real time, thereby identifying potential road unevenness or damage. At the same time, the camera captures road images and the GPS module provides accurate location information.

[0042] Fixed sensor nodes are deployed at key sections of the road to continuously monitor small changes in stress and vibration, and can promptly detect early signs of road damage. Energy harvesting units ensure that these sensors can operate for a long time, enabling them to have continuous monitoring capabilities. In addition, the distributed clock synchronization mechanism ensures that the data collection time of different sensor nodes is consistent, providing conditions for subsequent data fusion.

[0043] At the communication layer, the collected data is transmitted through a hybrid communication network. Onboard devices and fixed sensor nodes use low-power wide area networks for long-distance data transmission, while 4G / 5G networks are used for high-priority data. At the edge computing unit, the system performs preliminary processing on the received data, performs data noise reduction and compression, and uses lightweight deep learning models to analyze images in real time to identify road damage.

[0044] Subsequently, at the data processing and storage layer, the system integrates and analyzes the collected multi-source data. By establishing a Bayesian spatiotemporal statistical model, the system is able to model and predict the dynamic development of road damage. In addition, the DBSCAN algorithm is used for anomaly detection to identify possible road damage areas. All collected data is stored in a distributed storage system, supporting real-time query and analysis.

[0045] Finally, at the application level, the system provides intelligent decision-making support for traffic management departments. Through a neural network model based on a multi-layer perceptron, the system automatically evaluates the degree of road damage and generates an optimized maintenance strategy. At the same time, the system sets up a multi-level alarm mechanism to promptly notify relevant departments to carry out maintenance based on the monitoring results. In addition, by integrating a geographic information system, the road monitoring data is displayed in real time, allowing traffic management personnel to intuitively understand the location of damage and its distribution, and assist in long-term maintenance planning.

[0046] The present invention provides a real-time monitoring system for highway pavement based on the Internet of Things. It has the following beneficial effects:

[0047] 1. The present invention realizes high-frequency real-time monitoring of highway pavement conditions by integrating vehicle-mounted equipment and fixed sensors. The vehicle-mounted equipment is equipped with multimodal sensors, which enables the system to capture road vibration and damage images in real time during vehicle driving. The real-time monitoring capability enables the system to quickly identify potential road damage and unevenness problems, thereby significantly improving traffic safety, reducing accident risks, and ensuring the safety and smoothness of vehicle driving.

[0048] 2. The present invention utilizes Bayesian statistical models and DBSCAN clustering algorithms to intelligently analyze multi-source data, and can dynamically model and predict road damage. Compared with traditional methods, it not only improves the accuracy of data processing, but also can timely predict the changing trend of road conditions, help traffic management departments to pre-identify and deal with potential problems, and enhance the system's responsiveness, enabling it to more effectively respond to emergencies.

[0049] 3. The intelligent decision support module of the present invention can automatically evaluate the degree of road damage and generate an optimized maintenance strategy. The system formulates a maintenance plan based on real-time monitoring data, taking into account the resources, time and priority required for repair, thereby achieving efficient use of maintenance resources. By giving priority to repairing high-risk sections and severely damaged areas, the system not only reduces repair time, but also reduces maintenance costs, ensuring road safety and smooth traffic.

[0050] 4. The present invention uses a geographic information system integration module to dynamically display road monitoring data, including heat maps of damage locations and their severity, so that traffic management personnel can easily identify problem areas and quickly develop targeted maintenance plans, thereby improving management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a framework diagram of the system of the present invention;

[0052] Figure 2 This is a framework diagram of the perception layer of the present invention;

[0053] Figure 3 It is a framework diagram of the communication layer of the present invention;

[0054] Figure 4 It is a framework diagram of the data processing and storage layer of the present invention;

[0055] Figure 5 It is a framework diagram of the application layer of the present invention. DETAILED DESCRIPTION

[0056] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] Please refer to the attached Figure 1 The embodiment of the present invention provides a highway pavement real-time monitoring system based on the Internet of Things, comprising:

[0058] The perception layer is used to collect road surface data through vehicle-mounted equipment and fixed sensor nodes;

[0059] Communication layer, used to realize the transmission of various sensor data and edge computing processing;

[0060] Data processing and storage layer, used to integrate, analyze and store multi-source data;

[0061] The application layer is used to generate intelligent decision support and visual display.

[0062] Please refer to the attached Figure 2 ,The on-board equipment in the perception layer includes:

[0063] A multimodal sensor unit, which includes an accelerometer, gyroscope, camera, and GPS module for real-time monitoring of road vibrations, vehicle posture, damage images, and location data;

[0064] Multimodal data fusion unit, which is used to fuse various sensor data through Kalman filtering algorithm;

[0065] The adaptive sampling strategy unit is used to dynamically adjust the data sampling frequency based on the reinforcement learning algorithm to optimize energy consumption and monitoring effect.

[0066] The fixed sensor nodes in the perception layer include:

[0067] Strain gauge and vibration sensor units, which detect minute changes in stress and vibration in the road, identifying early signs of damage;

[0068] An energy harvesting unit, which is used to use solar energy or road vibration energy to power fixed nodes;

[0069] The timing synchronization unit is used to make the data collection time of different sensor nodes consistent through the distributed clock synchronization mechanism.

[0070] The fixed sensor nodes in the perception layer are deployed at optimized locations using an integer linear programming algorithm to minimize deployment costs and maximize monitoring coverage.

[0071] The objective function and constraints of the integer linear programming algorithm are as follows:

[0072] Objective function:

[0073]

[0074] Constraints:

[0075]

[0076] Among them, C i is the deployment cost, x i Deploy variables for nodes, d ij For coverage capability.

[0077] Specifically, the real-time highway pavement monitoring system based on the Internet of Things of the present invention realizes comprehensive monitoring and timely response to the highway pavement status through the vehicle-mounted equipment and fixed sensor nodes in the perception layer. The design of the perception layer integrates a variety of sensors, which can collect road vibration, vehicle posture, damage images and their location information in real time, and perform data processing and analysis through advanced algorithms. Through effective sensor data fusion and adaptive sampling strategies, the system can optimize energy consumption in a dynamic environment while ensuring the accuracy and reliability of the monitoring effect. In addition, the fixed sensor nodes maximize the monitoring coverage and realize accurate identification of early damage through optimized deployment. The following is a detailed description of this embodiment.

[0078] In this embodiment, the vehicle-mounted device in the perception layer includes a multimodal sensor unit. The unit consists of an accelerometer, a gyroscope, a camera, and a GPS module, and monitors road vibration, vehicle posture, damage images, and location data in real time. Specifically, the accelerometer is responsible for capturing vibration information during vehicle travel, while the gyroscope detects the angular velocity of the vehicle to help assess the unevenness of the road. The camera is used to obtain image information of road damage, and the GPS module provides real-time location information so that specific geographic locations can be associated in data analysis.

[0079] In terms of data fusion, this embodiment uses a multimodal data fusion unit to comprehensively process the data from different sensors through the Kalman filter algorithm. The core of the Kalman filter algorithm is to improve the accuracy of the data through the following prediction and update steps:

[0080] Prediction steps:

[0081]

[0082] Update steps:

[0083] K k =P k|k-1 HT (HP k|k-1 H T +R) -1

[0084]

[0085] P k =(IK k H)P k|k-1

[0086] In these steps, is the state estimate, K k is the Kalman gain, P k is the error covariance matrix. By fusing sensor data, the system can reflect the road status more accurately.

[0087] To improve energy efficiency and monitoring effect, the adaptive sampling strategy unit in this embodiment dynamically adjusts the data sampling frequency based on the reinforcement learning algorithm. Using the Q-Learning algorithm, the system can self-learn and update the sampling strategy according to the road conditions and sampling results. The update formula is:

[0088]

[0089] Among them, s is the current state, a is the action taken, r is the immediate reward, α is the learning rate, and γ is the discount factor. Through adaptive sampling, the system can effectively reduce energy consumption while ensuring data quality.

[0090] In terms of fixed sensor nodes, this embodiment includes strain gauges and vibration sensor units, which are mainly used to detect small stress and vibration changes in the road and identify early signs of damage in a timely manner. The energy collection unit uses solar energy or road vibration energy to power the fixed nodes to ensure that they can operate stably for a long time. The timing synchronization unit ensures that the data collection time of different sensor nodes is consistent through a distributed clock synchronization mechanism, providing a reliable basis for data fusion and analysis.

[0091] In order to maximize the monitoring coverage and minimize the deployment cost, the deployment location of the fixed sensor nodes is optimized using the integer linear programming (ILP) algorithm. In this embodiment, the objective function and constraints of the integer linear programming algorithm are as follows:

[0092] Objective function:

[0093]

[0094] Constraints:

[0095]

[0096] Among them, Ci is the deployment cost, x i Deploy variables for nodes, d ij Through this optimization strategy, the system can effectively plan the location layout of sensors and improve the overall monitoring coverage.

[0097] Please refer to the attached Figure 3 , the communication layer includes:

[0098] Hybrid communication network module, which is used to ensure stable and efficient transmission of sensor data;

[0099] Edge computing units are used to perform preliminary data processing and intelligent reasoning at nodes close to data sources.

[0100] The communication layer uses Markov decision process to optimize data transmission strategy.

[0101] The policy function formula of the Markov decision process is:

[0102]

[0103] Among them, π is the policy function, s t is the system state, R is the reward function, and γ is the discount factor.

[0104] Specifically, in this embodiment, the hybrid communication network module of the communication layer undertakes the main task of sensor data transmission. This module combines multiple communication technologies, including low-power wide area network, 4G and 5G, to ensure stable and fast data transmission under different environmental conditions. LPWAN technology is used for long-distance, low-power data transmission and is suitable for continuous monitoring of fixed sensors. High-priority data collected in vehicle-mounted equipment is transmitted in real time through 4G or 5G networks to ensure the timeliness of data.

[0105] In terms of the edge computing unit, the unit allows preliminary data processing near the data source. The implementation of this function includes noise reduction, compression and preliminary analysis of sensor data to reduce the amount of data transmission, reduce bandwidth usage, and improve the real-time performance of data processing. The edge computing unit applies a lightweight deep learning model to the real-time collected image data to quickly identify road damage and provide timely feedback to the system.

[0106] In order to further optimize the data transmission strategy, the Markov decision process is used in this embodiment. This process dynamically adjusts the data transmission decision according to the system status through an intelligent algorithm. The strategy function formula is defined as:

[0107]

[0108] In the formula, π represents the policy function, s tRepresents the current state of the system, R is the reward function, and γ is the discount factor. The application of this formula enables the system to automatically select the optimal communication path and protocol under different network conditions and states, thereby maximizing data transmission efficiency.

[0109] Please refer to the attached Figure 4 , the data processing and storage layer includes:

[0110] Multi-source data fusion module, which is used to conduct comprehensive analysis and fusion processing of data from vehicle-mounted equipment and fixed nodes. By establishing a Bayesian statistical model, it combines spatiotemporal data to dynamically model and predict road damage, and uses the DBSCAN algorithm to cluster road data and identify abnormal road areas;

[0111] Big data processing unit, which is used to store, manage and analyze large-scale data in real time.

[0112] Specifically, the data processing and storage layer of the real-time monitoring system for highway pavement based on the Internet of Things of the present invention is responsible for comprehensive analysis, fusion processing and storage management of the collected multi-source data to support dynamic modeling, prediction and abnormal identification of pavement damage. This layer is mainly composed of a multi-source data fusion module and a big data processing unit. The functions and implementation methods of these two modules will be described in detail below.

[0113] The main function of the multi-source data fusion module is to integrate the data from vehicle-mounted devices and fixed sensor nodes, and use statistical models and algorithms to accurately analyze and predict road conditions. This module specifically includes the following:

[0114] Data integration and preprocessing:

[0115] The data from vehicle-mounted devices (including multimodal sensors) and fixed sensor nodes are first preprocessed to remove noise and fill missing values ​​to ensure the accuracy of subsequent analysis.

[0116] The preprocessed data include road vibration, strain information, location data and damage images, which will be used in the subsequent fusion process.

[0117] Bayesian statistical models:

[0118] The Bayesian statistical model is used to dynamically model and predict road damage. The model is able to update beliefs about the road surface state based on historical data and real-time data.

[0119] The formula for calculating the posterior probability is as follows:

[0120]

[0121] Among them, θ represents the model parameters, D is the observed data, P(D|θ) is the likelihood function, and P(θ) is the prior distribution. By continuously updating the posterior probability, the system can accurately predict the development trend of road damage.

[0122] DBSCAN algorithm:

[0123] The DBSCAN algorithm is used to perform cluster analysis on the processed road surface data to identify abnormal road surface areas.

[0124] The density calculation formula of DBSCAN is:

[0125] density(x)=|{y∈D∣dist(x,y)≤∈}|

[0126] Through density clustering, normal data and potential road damage areas can be effectively distinguished, and automatic anomaly detection can be achieved.

[0127] The main task of the big data processing unit is to store, manage and analyze large-scale data in real time to support the efficient operation of the system. The functions of this unit include:

[0128] Data Storage:

[0129] The distributed storage architecture ensures that the system can process a large amount of real-time generated data. All collected road monitoring data will be stored in the cloud to ensure data security and reliability.

[0130] The data storage system supports elastic expansion and can dynamically adjust storage resources according to changes in data volume.

[0131] Real-time data management:

[0132] Through the stream processing framework, the real-time data stream can be managed and scheduled. The system can process the data stream from each sensor to ensure that the data is processed and stored as soon as possible after collection.

[0133] The real-time analysis module supports rapid query and analysis of data, and can provide timely feedback on road conditions to provide support for decision-making.

[0134] Data analysis and mining:

[0135] In the big data processing unit, the system will regularly conduct data mining and trend analysis, using machine learning and statistical analysis techniques to extract valuable information.

[0136] Based on the analysis results, the system can generate reports and provide them to traffic management departments to support road maintenance and management decisions.

[0137] Please refer to the attached Figure 5 , the application layer includes:

[0138] Intelligent decision support module, which is used to automatically evaluate the degree of road damage and generate optimized maintenance strategies based on a multi-layer perceptron neural network model, and set up a multi-level alarm mechanism according to the road damage situation;

[0139] The GIS integration module is used to integrate real-time pavement monitoring data, dynamically display damage locations and damage heat maps on the map, and use Kriging interpolation to generate a high-precision damage prediction model.

[0140] Specifically, the application layer of the real-time highway pavement monitoring system based on the Internet of Things of the present invention provides in-depth analysis and visual display of highway pavement monitoring data through an intelligent decision support module and a geographic information system integration module. This layer is designed to convert the collected data into actionable information to assist traffic management departments in making decisions and allocating resources.

[0141] The core function of the intelligent decision support module is to automatically evaluate the degree of road damage and generate an optimized maintenance strategy based on a multi-layer perceptron neural network model. The specific implementation includes the following aspects:

[0142] Damage Assessment:

[0143] The module uses a trained neural network model to analyze road data collected in real time. Input features include vibration data, location information, image data, etc. from multimodal sensors.

[0144] The output of the neural network model is the assessment level of road damage, which is expressed as:

[0145] y=f(W·X+b)

[0146] Among them, y is the output damage level, W is the weight matrix, X is the input feature matrix, and b is the bias term. In this way, the system can achieve a fast and accurate assessment of road damage.

[0147] Maintenance strategy generation:

[0148] Based on the evaluation results, the system generates optimized repair strategies for different damage levels. These strategies take into account the resources, time and priority required for repair to ensure efficient use of maintenance resources.

[0149] During the maintenance strategy formulation process, the system will refer to historical data and real-time monitoring information to generate the best repair plan.

[0150] Multi-level alarm mechanism:

[0151] The module sets up a multi-level warning mechanism based on the road damage. When serious damage is detected, the system will immediately send an alert to the relevant departments and provide repair suggestions.

[0152] Alert levels are divided into minor, moderate and severe, ensuring maintenance personnel can identify and respond to pavement problems in a timely manner.

[0153] Set multi-level alarm threshold T alert , when the fused data D f Trigger an alert when a certain threshold is exceeded:

[0154]

[0155] Different levels of alarm trigger different maintenance response plans, ensuring maintenance personnel can identify and respond to road problems in a timely manner

[0156] The GIS integration module is designed to integrate real-time pavement monitoring data and dynamically display the location and distribution of pavement damage. The specific implementation is as follows:

[0157] Real-time data integration:

[0158] The module integrates the data from various sensors with the GPS location data to form a complete road monitoring data set.

[0159] These data are displayed on a map through a visualization platform, allowing traffic managers to intuitively view the distribution of road damage.

[0160] Damage heat map display:

[0161] Based on the collected pavement monitoring data, the module generates a damage heat map, which indicates the degree of damage in different areas through the depth of color.

[0162] Heat maps can help managers quickly identify concentrations of damaged areas for targeted maintenance and resource allocation.

[0163] Kriging interpolation application:

[0164] The module uses Kriging interpolation to generate a high-precision damage prediction model. This method uses existing observation data to predict unknown areas. The formula is as follows:

[0165]

[0166] Among them, Z(s0) is the damage value of the position to be estimated, Z(s i ) is the observed value at the known position, λ i is the interpolation weight. Through this method, the system can accurately predict the future road conditions.

[0167] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A real-time monitoring system for highway pavement based on the Internet of Things, characterized in that: include: The perception layer is used to collect road surface data through vehicle-mounted equipment and fixed sensor nodes; Communication layer, used to realize the transmission of various sensor data and edge computing processing; Data processing and storage layer, used to integrate, analyze and store multi-source data; The application layer is used to generate intelligent decision support and visual display.

2. The real-time monitoring system for highway pavement based on the Internet of Things according to claim 1 is characterized in that: The vehicle-mounted devices in the perception layer include: A multimodal sensor unit, which includes an accelerometer, gyroscope, camera, and GPS module for real-time monitoring of road vibrations, vehicle posture, damage images, and location data; Multimodal data fusion unit, which is used to fuse various sensor data through Kalman filtering algorithm; The adaptive sampling strategy unit is used to dynamically adjust the data sampling frequency based on the reinforcement learning algorithm to optimize energy consumption and monitoring effect.

3. The real-time monitoring system for highway pavement based on the Internet of Things according to claim 1 is characterized in that: The fixed sensor nodes in the perception layer include: Strain gauge and vibration sensor units, which detect minute changes in stress and vibration in the road, identifying early signs of damage; An energy harvesting unit, which is used to use solar energy or road vibration energy to power fixed nodes; The timing synchronization unit is used to make the data collection time of different sensor nodes consistent through the distributed clock synchronization mechanism.

4. The real-time monitoring system for highway pavement based on the Internet of Things according to claim 1 is characterized in that: The communication layer includes: Hybrid communication network module, which is used to ensure stable and efficient transmission of sensor data; Edge computing units are used to perform preliminary data processing and intelligent reasoning at nodes close to data sources.

5. The real-time road surface monitoring system based on the Internet of Things according to claim 1 is characterized in that: The data processing and storage layer includes: Multi-source data fusion module, which is used to conduct comprehensive analysis and fusion processing of data from vehicle-mounted equipment and fixed nodes. By establishing a Bayesian statistical model, it combines spatiotemporal data to dynamically model and predict road damage, and uses the DBSCAN algorithm to cluster road data and identify abnormal road areas; Big data processing unit, which is used to store, manage and analyze large-scale data in real time.

6. The real-time road surface monitoring system based on the Internet of Things according to claim 1 is characterized in that: The application layer includes: Intelligent decision support module, which is used to automatically evaluate the degree of road damage and generate optimized maintenance strategies based on a multi-layer perceptron neural network model, and set up a multi-level alarm mechanism according to the road damage situation; The GIS integration module is used to integrate real-time pavement monitoring data, dynamically display damage locations and damage heat maps on the map, and use Kriging interpolation to generate a high-precision damage prediction model.

7. The real-time monitoring system for highway pavement based on the Internet of Things according to claim 1 is characterized in that: The fixed sensor nodes in the perception layer are deployed at optimized locations through an integer linear programming algorithm to minimize deployment costs and maximize monitoring coverage.

8. The real-time monitoring system for highway pavement based on the Internet of Things according to claim 7 is characterized in that: The objective function and constraints of the integer linear programming algorithm are as follows: Objective function: Constraints: Among them, C i is the deployment cost, x i Deploy variables for nodes, d ij For coverage capability.

9. The real-time road surface monitoring system based on the Internet of Things according to claim 1 is characterized in that: The communication layer uses a Markov decision process to optimize the data transmission strategy.

10. The real-time road surface monitoring system based on the Internet of Things according to claim 9 is characterized in that: The strategy function formula of the Markov decision process is: Among them, π is the policy function, s t is the system state, R is the reward function, and γ is the discount factor.

Images