Monitoring management method and system for smart tunnel based on Internet of Things

Abstract

The invention relates to the field of smart tunnel monitoring and management, and provides a smart tunnel monitoring and management method and system based on the Internet of Things, and the method comprises the steps: collecting multi-source heterogeneous monitoring data in a tunnel through multiple types of sensor nodes; performing standardization processing on the multi-source heterogeneous monitoring data to obtain standardized tunnel monitoring data; synchronizing and aligning the standardized tunnel monitoring data, and processing time differences, missing values and abnormal values of different data sources to obtain tunnel monitoring time synchronization data; extracting multi-dimensional features from the standardized tunnel monitoring data to obtain tunnel state feature data; constructing a deep learning early warning model, and identifying the tunnel state feature data to obtain tunnel early warning data; and through multi-channel information pushing, the tunnel state feature data and the tunnel early warning data are pushed to management personnel, so that the management personnel can monitor the tunnel state of the smart tunnel in real time. According to the invention, timely and accurate tunnel state monitoring management information is provided for managers.

Description

Technical Field

[0001] The present invention relates to the field of smart tunnel monitoring and management, and in particular to a monitoring and management method and system for a smart tunnel based on the Internet of Things. Background Art

[0002] Smart tunnels, a crucial component of smart transportation, deploy multiple sensors, communication networks, and intelligent analysis systems to enable real-time monitoring, analysis, and management of multi-dimensional data, including the tunnel environment, structure, and traffic flow, thereby ensuring the safety, stability, and efficiency of tunnel operations. The core goal of smart tunnels is to improve tunnel operational efficiency, reduce accident rates, and provide timely and accurate decision-making support for managers through intelligent means. Currently, smart tunnel monitoring and management face the following challenges: a single source of monitoring data, making it difficult to fully cover tunnel status; insufficient data processing capabilities, making it difficult to handle multi-source, heterogeneous data; a lagging early warning mechanism lacking intelligent analysis capabilities; and a single information push method, resulting in inefficient emergency response.

[0003] Existing technologies use threshold-based early warning methods for tunnel status monitoring. These methods deploy environmental sensors such as temperature and humidity, along with video surveillance equipment, within the tunnel, and set fixed thresholds to detect abnormal conditions and issue early warnings. However, this method fails to effectively capture the dynamic nature of tunnel conditions. The preset fixed thresholds are difficult to adapt to the complex and changing tunnel environment, and are prone to missed or false alarms. The data processing is simplistic, lacking the ability to deeply analyze multi-source, heterogeneous data and unable to identify correlations between different types of data. This results in inaccurate early warnings and significantly reduces the timeliness of monitoring and management. Summary of the Invention

[0004] In view of this, the present invention proposes a monitoring and management method and system for smart tunnels based on the Internet of Things, which solves the problem that the existing technology lacks the ability to deeply analyze multi-source heterogeneous data and cannot mine the correlation characteristics between different types of data, resulting in insufficient early warning accuracy and greatly reduced timeliness of monitoring and management.

[0005] The technical solution of the present invention is implemented as follows: In a first aspect, the present invention provides a monitoring and management method for a smart tunnel based on the Internet of Things, comprising the following steps:

[0006] Collect multi-source heterogeneous monitoring data in the tunnel through multi-type sensor nodes;

[0007] Standardizing the multi-source heterogeneous monitoring data to obtain standardized tunnel monitoring data;

[0008] Synchronizing and aligning the standardized tunnel monitoring data, processing time differences, missing values, and outliers from different data sources, and obtaining tunnel monitoring time-synchronized data;

[0009] Extracting multidimensional features from the standardized tunnel monitoring data to obtain tunnel status feature data;

[0010] Constructing a deep learning early warning model to identify the tunnel status feature data and obtain tunnel early warning data;

[0011] Through multi-channel information push, tunnel status feature data and tunnel warning data are pushed to management personnel, so that management personnel can monitor the tunnel status of the smart tunnel in real time.

[0012] Based on the above technical solution, preferably, the collecting of multi-source heterogeneous monitoring data in the tunnel by using multi-type sensor nodes includes:

[0013] According to the tunnel structure characteristics and monitoring requirements, a sensor deployment plan is set up to determine the layout and installation method of various types of sensors in the tunnel;

[0014] Differentiated data collection frequencies and rules are set for different types of monitoring data to comprehensively monitor the tunnel status.

[0015] Based on the above technical solution, preferably, the step of performing standardization processing on the multi-source heterogeneous monitoring data to obtain standardized tunnel monitoring data includes:

[0016] Perform protocol parsing and data format conversion on multi-source heterogeneous monitoring data, and convert data of different protocols and formats into a unified data model;

[0017] The data after protocol parsing is verified and cleaned to eliminate erroneous data, redundant data and invalid data to obtain standardized tunnel monitoring data.

[0018] Based on the above technical solution, preferably, synchronizing and aligning the standardized tunnel monitoring data, processing time differences, missing values and abnormal values of different data sources, and obtaining tunnel monitoring time synchronization data include:

[0019] Perform timestamp alignment on the standardized tunnel monitoring data to obtain a time-consistent dataset;

[0020] The time-consistent data set is filled with missing values and corrected for outliers to obtain the tunnel monitoring time-synchronized data.

[0021] Based on the above technical solution, preferably, extracting multidimensional features from the standardized tunnel monitoring data to obtain tunnel status feature data includes:

[0022] Extract time domain features and frequency domain features from tunnel monitoring time synchronization data, including:

[0023] Extracting time domain features by statistical analysis methods, wherein the time domain features include mean, variance, peak, skewness and kurtosis; extracting frequency domain features by fast Fourier transform, wherein the frequency domain features include dominant frequency, spectral energy distribution and spectral density;

[0024] Based on the correlation analysis between multiple parameters, the Pearson correlation coefficient is used to calculate the correlation characteristics between environmental data and traffic flow data;

[0025] Based on the dynamic time warping algorithm, key correlation features are extracted to obtain tunnel status feature data;

[0026] The calculation formula of the Pearson correlation coefficient is:

[0027]

[0028] Where r is the Pearson correlation coefficient, x i and y i are the values of the environmental data variable and the traffic data variable in the i-th sample, N is the total number of samples, is the mean of the environmental data variable, is the mean of the traffic flow data variable.

[0029] Based on the above technical solution, preferably, the deep learning warning model is constructed to identify the tunnel status feature data to obtain tunnel warning data, including:

[0030] A multi-layer neural network was used to construct a pre-trained model, which extracted features from historical tunnel status data. Time series modeling was then performed on the feature data using a long-short-term memory network. The pre-trained model was trained using labeled warning event data through supervised learning, and model parameters were optimized to obtain a deep learning model for tunnel status prediction.

[0031] The loss function of the tunnel state prediction deep learning model is calculated as:

[0032]

[0033] Among them, L is the loss function, M is the total number of samples, z j is the true label of the j-th sample, is the probability value predicted by the model, λ is the regularization coefficient, W is the model weight parameter, ||·|| 2 is the norm;

[0034] The tunnel status feature data extracted in real time is input into the tunnel status prediction deep learning model. The tunnel status prediction deep learning model outputs the probability value of the abnormal state. Combined with the dynamic threshold adjustment mechanism, it determines whether to trigger an early warning. When the early warning is triggered, tunnel early warning data including the abnormality type, location, and time information is generated;

[0035] The judgment formula for determining whether an early warning is triggered is:

[0036]

[0037] If P ≥ θ, trigger the warning;

[0038] Among them, P is the average probability value of abnormal state in the latest k time windows, p t is the abnormal probability value output by the model at time t, T is the current time point, and θ is the dynamic threshold.

[0039] Based on the above technical solution, preferably, the tunnel status feature data and tunnel warning data are pushed to the management personnel through multi-channel information push, so that the management personnel can monitor the tunnel status of the smart tunnel in real time, including:

[0040] According to the level and type of tunnel warning data, a hierarchical warning mechanism is triggered to determine the push priority of warning information;

[0041] Through multi-channel information push, tunnel status feature data and tunnel warning data are sent to designated managers and system terminals.

[0042] In a second aspect, the present invention further provides a monitoring and management system for a smart tunnel based on the Internet of Things, the system comprising:

[0043] Data acquisition module, used to collect multi-source heterogeneous monitoring data in the tunnel through multiple types of sensor nodes;

[0044] A data processing module, configured to perform standardization processing on the multi-source heterogeneous monitoring data to obtain standardized tunnel monitoring data;

[0045] A synchronization and alignment module is used to synchronize and align the standardized tunnel monitoring data, process time differences, missing values and abnormal values of different data sources, and obtain tunnel monitoring time synchronization data;

[0046] A feature extraction module is used to extract multi-dimensional features from the standardized tunnel monitoring data to obtain tunnel status feature data;

[0047] A status warning module is used to build a deep learning warning model to identify the tunnel status feature data and obtain tunnel warning data;

[0048] The monitoring and management module is used to push tunnel status feature data and tunnel warning data to management personnel through multi-channel information push, so that management personnel can monitor the tunnel status of the smart tunnel in real time.

[0049] In a third aspect, the present invention further provides an electronic device comprising: at least one processor, at least one memory, a communication interface, and a bus;

[0050] Among them, the processor, memory, and communication interface communicate with each other through the bus, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to implement steps such as a monitoring and management method for a smart tunnel based on the Internet of Things.

[0051] In a fourth aspect, the present invention further provides a computer-readable storage medium, which stores computer instructions, and the computer instructions enable a computer to implement steps of a monitoring and management method for a smart tunnel based on the Internet of Things.

[0052] The monitoring and management method and system of a smart tunnel based on the Internet of Things of the present invention have the following beneficial effects compared with the prior art:

[0053] (1) By collecting multi-source heterogeneous data such as environment, structure, video, and traffic flow through multiple types of sensors, and through standardization, time synchronization alignment, feature extraction, and deep learning warning model analysis, real-time monitoring and early warning of tunnel status are achieved. Abnormal conditions in tunnel operation can be discovered in a timely manner and early warning information can be pushed through multiple channels, effectively improving the safety of tunnel operation management and providing managers with timely and accurate tunnel status monitoring and management information;

[0054] (2) By extracting time domain features and frequency domain features from tunnel monitoring time synchronization data and combining them with correlation analysis between multiple parameters, key correlation features are extracted to generate tunnel status feature data. This fully captures the multi-dimensional information of the tunnel operation status, which can not only reflect the static and dynamic characteristics of the tunnel status, but also reveal the correlation between different monitoring data, such as the correlation between environmental data and traffic flow data, thereby improving the reliability of tunnel abnormal status identification;

[0055] (3) By adopting multi-layer neural networks and long short-term memory networks to construct a deep learning model for tunnel state prediction, feature extraction and time series modeling are performed on historical tunnel state feature data, and supervised learning is combined to optimize model parameters, thereby improving the accuracy of abnormal state identification. The abnormal probability of the model output is judged through a dynamic threshold adjustment mechanism, achieving accurate early warning of tunnel abnormal states, effectively capturing the dynamic change characteristics of tunnel states, and improving the real-time prediction of tunnel abnormal states. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0057] Figure 1 This is a flow chart of a monitoring and management method for a smart tunnel based on the Internet of Things of the present invention;

[0058] Figure 2 This is a structural diagram of a monitoring and management system for a smart tunnel based on the Internet of Things of the present invention. DETAILED DESCRIPTION

[0059] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the embodiments described 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 making creative efforts are within the scope of protection of the present invention.

[0060] See also Figure 1 The present invention provides a monitoring and management method for a smart tunnel based on the Internet of Things, comprising the following steps:

[0061] Collecting multi-source heterogeneous monitoring data in the tunnel through multi-type sensor nodes, wherein the multi-source heterogeneous monitoring data includes environmental data, structural data, video data, and traffic flow data;

[0062] Standardizing the multi-source heterogeneous monitoring data, including unified data model conversion and multi-protocol adaptation, to obtain standardized tunnel monitoring data;

[0063] Synchronizing and aligning the standardized tunnel monitoring data, processing time differences, missing values, and outliers from different data sources, and obtaining tunnel monitoring time-synchronized data;

[0064] Extracting multidimensional features from the standardized tunnel monitoring data, including time domain features, frequency domain features, and correlation features, to obtain tunnel status feature data;

[0065] Constructing a deep learning early warning model to identify the tunnel status feature data and obtain tunnel early warning data;

[0066] Through multi-channel information push, tunnel status feature data and tunnel warning data are pushed to management personnel, so that management personnel can monitor the tunnel status of the smart tunnel in real time.

[0067] Specifically, this embodiment uses multiple types of sensors to collect multi-source heterogeneous data such as environment, structure, video and traffic flow, and through standardized processing, time synchronization alignment, feature extraction and deep learning warning model analysis, it realizes real-time monitoring and early warning of tunnel status. It can promptly detect abnormal conditions in tunnel operation and push early warning information through multiple channels, effectively improving the safety of tunnel operation management and providing managers with timely and accurate tunnel status monitoring and management information.

[0068] The method of collecting multi-source heterogeneous monitoring data in the tunnel by using multi-type sensor nodes includes:

[0069] According to the tunnel structure characteristics and monitoring requirements, a sensor deployment plan is set up to determine the layout and installation method of various types of sensors in the tunnel.

[0070] In a specific embodiment, the sensor deployment solution includes:

[0071] Environmental monitoring sensors are installed at the tunnel entrance, exit, and every 50-200 meters on the tunnel wall to monitor environmental parameters such as temperature, humidity, CO, and NO2;

[0072] Structural monitoring sensors are placed at key locations of the tunnel structure, including the vault, sidewalls, and joints, to monitor strain, displacement, and inclination data;

[0073] Video surveillance equipment will be evenly distributed along the tunnel's longitudinal direction to ensure there are no blind spots, and infrared imaging equipment will be added in accident-prone areas.

[0074] Traffic flow monitoring equipment is installed at tunnel entrances, exits and key sections to count vehicles, identify speeds and vehicle types.

[0075] Differentiated data collection frequencies and rules are set for different types of monitoring data to comprehensively monitor the tunnel status.

[0076] In a specific embodiment, the differentiated data collection rules include:

[0077] Environmental data is collected periodically, once every 10-30 minutes under normal conditions, and automatically increased to once every 1-5 minutes when an anomaly is detected.

[0078] Structural data is collected using a combination of triggered and periodic collection modes, with periodic collection occurring 1-4 times a day. Intensive collection is triggered when sudden increases in traffic volume, extreme weather, or geological activity are detected.

[0079] Video data uses a combination of continuous acquisition and intelligent recognition, storing and transmitting complete video clips to the data center only when the system identifies an abnormal event;

[0080] Traffic flow data is collected in real time, and the frequency of data upload is dynamically adjusted according to the volume of traffic to reduce network load.

[0081] Specifically, this embodiment achieves comprehensive coverage and precise monitoring of the tunnel's multi-dimensional environment, structure, and traffic flow by designing a sensor deployment plan based on the tunnel's structural characteristics and monitoring requirements, rationally placing environmental monitoring sensors, structural monitoring sensors, video surveillance equipment, and traffic flow monitoring equipment. Furthermore, by employing differentiated data collection frequencies and rules, and dynamically adjusting the collection mode based on the characteristics of different monitoring data, this ensures the real-time and integrity of monitoring data while effectively reducing the system's network load and resource consumption, thereby improving the efficiency and reliability of the tunnel monitoring system.

[0082] The step of standardizing the multi-source heterogeneous monitoring data to obtain standardized tunnel monitoring data includes:

[0083] Perform protocol parsing and data format conversion on multi-source heterogeneous monitoring data, and convert data of different protocols and formats into a unified data model.

[0084] In a specific embodiment, the protocol parsing and the data format conversion include:

[0085] Design multi-protocol adapters for mainstream communication protocols such as Modbus, OPC UA, and MQTT to automatically analyze data from different protocols;

[0086] According to different sensor data formats, data mapping rules are adopted to convert the data into a unified structured format. The data model includes data source identification, timestamp, data value, unit and status code fields.

[0087] The data after protocol parsing is verified and cleaned to eliminate erroneous data, redundant data and invalid data to obtain standardized tunnel monitoring data.

[0088] In a specific embodiment, the data verification and cleaning includes:

[0089] Use data integrity verification algorithm to verify the field integrity of collected data;

[0090] Identify and eliminate abnormal data that exceeds a reasonable range through outlier detection algorithms;

[0091] Missing data are supplemented by interpolation or historical data backfilling;

[0092] For redundant data, deduplication is performed based on timestamps and data source identifiers to ensure data uniqueness.

[0093] Specifically, this embodiment effectively addresses the compatibility issues of multi-source heterogeneous data access in the intelligent tunnel monitoring system through multi-protocol adapters and data mapping rules, enabling automatic parsing and unified conversion of data from different communication protocols such as Modbus, OPC UA, and MQTT. Furthermore, by establishing a standardized data model containing data source identifiers, timestamps, data values, units, and status code fields, a structurally consistent data foundation is provided. In particular, data validation and cleaning methods such as integrity checking, outlier detection, missing data completion, and redundant data deduplication are implemented to improve data quality and reliability, effectively reducing the risk of misjudgment due to poor data quality.

[0094] The step of synchronizing and aligning the standardized tunnel monitoring data, processing time differences, missing values, and abnormal values of different data sources, and obtaining tunnel monitoring time synchronization data includes:

[0095] Timestamp alignment is performed on the standardized tunnel monitoring data to resolve the time differences between different data sources and obtain a time-consistent dataset.

[0096] In a specific embodiment, the timestamp alignment includes:

[0097] Adopting a time window-based alignment algorithm to align data with different sampling frequencies to a unified time axis;

[0098] For high-frequency data, a downsampling method is used to obtain key data points;

[0099] For low-frequency data, an interpolation algorithm is used to generate supplementary data points to ensure time alignment with high-frequency data.

[0100] Missing values are filled and outliers are corrected for the time-consistent dataset to obtain complete and reliable tunnel monitoring time synchronization data.

[0101] In a specific embodiment, the missing value filling and outlier correction include:

[0102] Missing values are filled using regression models or interpolation algorithms based on historical data;

[0103] Use anomaly detection methods based on statistical analysis for outliers, including box plot method or Z score method, to identify and correct abnormal data that exceeds the reasonable range;

[0104] Duplicate and redundant data are deduplicated based on timestamps and data source identifiers to ensure data uniqueness and accuracy.

[0105] Specifically, this embodiment effectively addresses the time discrepancy problem caused by inconsistent sampling frequencies across different data sources through a time-window-based alignment algorithm and differentiated processing strategies, including downsampling of high-frequency data and interpolation of low-frequency data. Furthermore, it employs regression models based on historical data and statistical analysis methods, such as boxplots and Z-scores, to fill missing values and correct outliers. Furthermore, it uses timestamps and data source identifiers for deduplication, ensuring data integrity, accuracy, and consistency, thereby improving the data reliability and analytical accuracy of the tunnel monitoring system.

[0106] The step of extracting multi-dimensional features from the standardized tunnel monitoring data to obtain tunnel status feature data includes:

[0107] Extract time domain features and frequency domain features from tunnel monitoring time synchronization data, including:

[0108] Extracting time domain features by statistical analysis methods, wherein the time domain features include mean, variance, peak, skewness and kurtosis; extracting frequency domain features by fast Fourier transform, wherein the frequency domain features include main frequency, spectral energy distribution and spectral density;

[0109] Based on the correlation analysis between multiple parameters, the Pearson correlation coefficient is used to calculate the correlation characteristics between environmental data and traffic flow data;

[0110] Based on the dynamic time warping algorithm, the dynamic correlation between time series data is analyzed, key correlation features are extracted, and tunnel status feature data is obtained;

[0111] The calculation formula of the Pearson correlation coefficient is:

[0112]

[0113] Where r is the Pearson correlation coefficient, x i and y i are the values of the environmental data variable and the traffic data variable in the i-th sample, N is the total number of samples, is the mean of the environmental data variable, is the mean of the traffic flow data variable.

[0114] Specifically, this embodiment extracts time domain features and frequency domain features from tunnel monitoring time synchronization data, and combines correlation analysis between multiple parameters to extract key correlation features, generate tunnel status feature data, and comprehensively capture multi-dimensional information of the tunnel operation status. It can not only reflect the static and dynamic characteristics of the tunnel status, but also reveal the correlation between different monitoring data, such as the correlation between environmental data and traffic flow data, thereby improving the reliability of identifying abnormal tunnel conditions.

[0115] The deep learning warning model is constructed to identify the tunnel status feature data to obtain tunnel warning data, including:

[0116] A pre-trained model is constructed using a multi-layer neural network to extract features from historical tunnel status feature data. Based on a long-short-term memory network, time series modeling is performed on the feature data to capture the dynamic characteristics of tunnel status changes. The pre-trained model is trained using labeled warning event data through supervised learning methods, and model parameters are optimized to improve the accuracy of abnormal state identification, resulting in a deep learning model for tunnel status prediction.

[0117] The loss function of the tunnel state prediction deep learning model is calculated as:

[0118]

[0119] Among them, L is the loss function, M is the total number of samples, z j is the true label of the j-th sample, is the probability value predicted by the model, λ is the regularization coefficient, W is the model weight parameter, ||·|| 2 is the norm;

[0120] The tunnel status feature data extracted in real time is input into the tunnel status prediction deep learning model. The tunnel status prediction deep learning model outputs the probability value of the abnormal state. Combined with the dynamic threshold adjustment mechanism, it determines whether to trigger an early warning. When the early warning is triggered, tunnel early warning data including the abnormality type, location, and time information is generated;

[0121] The judgment formula for determining whether an early warning is triggered is:

[0122]

[0123] If P ≥ θ, trigger the warning;

[0124] Among them, P is the average probability value of abnormal state in the latest k time windows, p t is the abnormal probability value output by the model at time t, T is the current time point, and θ is the dynamic threshold.

[0125] Specifically, this embodiment constructs a deep learning model for tunnel state prediction by adopting a multi-layer neural network and a long short-term memory network, performs feature extraction and time series modeling on historical tunnel state feature data, and optimizes model parameters in combination with supervised learning, thereby improving the accuracy of abnormal state identification. The abnormal probability of the model output is judged through a dynamic threshold adjustment mechanism, thereby achieving accurate early warning of tunnel abnormal states, effectively capturing the dynamic change characteristics of the tunnel state, and improving the real-time prediction of tunnel abnormal states.

[0126] The tunnel status feature data and tunnel warning data are pushed to management personnel through multi-channel information push, so that management personnel can monitor the tunnel status of the smart tunnel in real time, including:

[0127] According to the level and type of tunnel warning data, the hierarchical warning mechanism is triggered to determine the push priority of the warning information.

[0128] In a specific embodiment, the hierarchical warning mechanism includes:

[0129] According to the abnormality of tunnel warning data, the warning level is divided into three levels: low, medium and high;

[0130] When the warning level is low, only the warning information is recorded and pushed to the monitoring system backend;

[0131] When the warning level is medium, push warning information to relevant managers and recommend inspection measures;

[0132] When the warning level is high, emergency warning information will be sent immediately to all relevant management personnel, and emergency equipment in the tunnel (such as warning lights, broadcasting systems, etc.) will be linked to alert vehicles and personnel in the tunnel.

[0133] Through multi-channel information push, tunnel status feature data and tunnel warning data are sent to designated managers and system terminals.

[0134] In a specific embodiment, the multi-channel information push method includes:

[0135] Push early warning information to managers through SMS, email, mobile app notifications, etc.

[0136] Real-time display of tunnel status characteristic data and early warning data on the monitoring center display screen;

[0137] Through the Internet of Things platform interface, early warning information is pushed to the tunnel emergency management system and traffic management system to achieve information sharing and coordinated response.

[0138] Specifically, this embodiment effectively addresses the issues of inaccurate warning information processing and untimely response in traditional tunnel monitoring systems by establishing a hierarchical warning mechanism and a multi-channel information push system. By categorizing warning levels into low, medium, and high and associating them with different handling strategies, differentiated management of tunnel abnormalities is achieved, preventing excessive interference from low-level warnings on management personnel while ensuring that high-risk situations receive an immediate response. In particular, the design of high-level warnings automatically links to in-tunnel emergency equipment, automating the entire process from monitoring and warning to emergency response.

[0139] Utilizing multiple channels for information push, such as SMS, email, app notifications, and display screens, we ensure that warning information is delivered promptly and reliably to relevant personnel through multiple channels, significantly improving the reliability and coverage of information delivery. Furthermore, through the IoT platform interface, connectivity with the tunnel emergency management system and traffic management system enables cross-system information sharing and coordinated response, improving the overall efficiency of handling tunnel safety incidents.

[0140] See also Figure 2 The present invention also provides a monitoring and management system for a smart tunnel based on the Internet of Things, the system comprising:

[0141] Data acquisition module, used to collect multi-source heterogeneous monitoring data in the tunnel through multiple types of sensor nodes;

[0142] A data processing module, configured to perform standardization processing on the multi-source heterogeneous monitoring data to obtain standardized tunnel monitoring data;

[0143] A synchronization and alignment module is used to synchronize and align the standardized tunnel monitoring data, process time differences, missing values and abnormal values of different data sources, and obtain tunnel monitoring time synchronization data;

[0144] A feature extraction module is used to extract multi-dimensional features from the standardized tunnel monitoring data to obtain tunnel status feature data;

[0145] A status warning module is used to build a deep learning warning model to identify the tunnel status feature data and obtain tunnel warning data;

[0146] The monitoring and management module is used to push tunnel status feature data and tunnel warning data to management personnel through multi-channel information push, so that management personnel can monitor the tunnel status of the smart tunnel in real time.

[0147] Specifically, the monitoring and management system of a smart tunnel based on the Internet of Things in this embodiment realizes the unified collection of multi-source heterogeneous data through the data acquisition module, ensures the standardized conversion of data by using the data processing module, solves the time difference problem of different data sources based on the synchronization alignment module, realizes the intelligent extraction of multi-dimensional features through the feature extraction module, uses the status warning module to provide accurate warning identification through the deep learning model, and uses the monitoring management module to ensure the timely delivery of information. This modular system architecture design not only improves the maintainability and scalability of the system, but also realizes the seamless connection between various functional modules, making the system operation more stable and reliable. Through the collaborative work of various modules, the system improves the automation level of tunnel monitoring.

[0148] The present invention also discloses an electronic device, comprising: at least one processor, at least one memory communication interface and a bus: wherein the processor, memory and communication interface communicate with each other through the bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to implement a monitoring and management method for a smart tunnel based on the Internet of Things.

[0149] The present invention also discloses a computer-readable storage medium storing computer instructions that cause the computer to implement all or part of the steps of the IoT-based smart tunnel monitoring and management method described in an embodiment of the present invention. The storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0150] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A monitoring and management method for a smart tunnel based on the Internet of Things, characterized in that: The following steps are involved: Collect multi-source heterogeneous monitoring data in the tunnel through multi-type sensor nodes; Standardizing the multi-source heterogeneous monitoring data to obtain standardized tunnel monitoring data; Synchronizing and aligning the standardized tunnel monitoring data, processing time differences, missing values, and outliers from different data sources, and obtaining tunnel monitoring time-synchronized data; Extracting multidimensional features from the standardized tunnel monitoring data to obtain tunnel status feature data; Constructing a deep learning early warning model to identify the tunnel status feature data and obtain tunnel early warning data; Through multi-channel information push, tunnel status feature data and tunnel warning data are pushed to management personnel, so that management personnel can monitor the tunnel status of the smart tunnel in real time.

2. The monitoring and management method of a smart tunnel based on the Internet of Things according to claim 1, characterized in that: The method of collecting multi-source heterogeneous monitoring data in the tunnel by using multi-type sensor nodes includes: According to the tunnel structure characteristics and monitoring requirements, a sensor deployment plan is set up to determine the layout and installation method of various types of sensors in the tunnel; Differentiated data collection frequencies and rules are set for different types of monitoring data to comprehensively monitor the tunnel status.

3. The monitoring and management method of a smart tunnel based on the Internet of Things according to claim 1, characterized in that: The step of standardizing the multi-source heterogeneous monitoring data to obtain standardized tunnel monitoring data includes: Perform protocol parsing and data format conversion on multi-source heterogeneous monitoring data, and convert data of different protocols and formats into a unified data model; The data after protocol parsing is verified and cleaned to eliminate erroneous data, redundant data and invalid data to obtain standardized tunnel monitoring data.

4. The monitoring and management method of a smart tunnel based on the Internet of Things according to claim 1, characterized in that: The step of synchronizing and aligning the standardized tunnel monitoring data, processing time differences, missing values, and abnormal values of different data sources, and obtaining tunnel monitoring time synchronization data includes: Perform timestamp alignment on the standardized tunnel monitoring data to obtain a time-consistent dataset; The time-consistent data set is filled with missing values and corrected for outliers to obtain the tunnel monitoring time-synchronized data.

5. The monitoring and management method of a smart tunnel based on the Internet of Things according to claim 4, characterized in that: The step of extracting multi-dimensional features from the standardized tunnel monitoring data to obtain tunnel status feature data includes: Extract time domain features and frequency domain features from tunnel monitoring time synchronization data, including: Extracting time domain features by statistical analysis methods, wherein the time domain features include mean, variance, peak, skewness and kurtosis; extracting frequency domain features by fast Fourier transform, wherein the frequency domain features include dominant frequency, spectral energy distribution and spectral density; Based on the correlation analysis between multiple parameters, the Pearson correlation coefficient is used to calculate the correlation characteristics between environmental data and traffic flow data; Based on the dynamic time warping algorithm, key correlation features are extracted to obtain tunnel status feature data; The calculation formula of the Pearson correlation coefficient is: Where r is the Pearson correlation coefficient, x i and y i are the values of the environmental data variable and the traffic data variable in the i-th sample, N is the total number of samples, is the mean of the environmental data variable, is the mean of the traffic flow data variable.

6. The monitoring and management method of a smart tunnel based on the Internet of Things according to claim 1, characterized in that: The deep learning warning model is constructed to identify the tunnel status feature data to obtain tunnel warning data, including: A multi-layer neural network was used to construct a pre-trained model, which extracted features from historical tunnel status data. Time series modeling was then performed on the feature data using a long-short-term memory network. The pre-trained model was trained using labeled warning event data through supervised learning, and model parameters were optimized to obtain a deep learning model for tunnel status prediction. The loss function of the tunnel state prediction deep learning model is calculated as: Among them, L is the loss function, M is the total number of samples, z j is the true label of the j-th sample, is the probability value predicted by the model, λ is the regularization coefficient, W is the model weight parameter, ||·|| 2 is the norm; The tunnel status feature data extracted in real time is input into the tunnel status prediction deep learning model. The tunnel status prediction deep learning model outputs the probability value of the abnormal state. Combined with the dynamic threshold adjustment mechanism, it determines whether to trigger an early warning. When the early warning is triggered, tunnel early warning data including the abnormality type, location, and time information is generated; The judgment formula for determining whether an early warning is triggered is: If P ≥ θ, trigger the warning; Among them, P is the average probability value of abnormal state in the latest k time windows, p t is the abnormal probability value output by the model at time t, T is the current time point, and θ is the dynamic threshold.

7. The monitoring and management method of a smart tunnel based on the Internet of Things according to claim 1, characterized in that: The tunnel status feature data and tunnel warning data are pushed to management personnel through multi-channel information push, so that management personnel can monitor the tunnel status of the smart tunnel in real time, including: According to the level and type of tunnel warning data, a hierarchical warning mechanism is triggered to determine the push priority of warning information; Through multi-channel information push, tunnel status feature data and tunnel warning data are sent to designated managers and system terminals.

8. A monitoring and management system for a smart tunnel based on the Internet of Things, used to execute the monitoring and management method for a smart tunnel based on the Internet of Things according to any one of claims 1 to 7, characterized in that: The system comprises: Data acquisition module, used to collect multi-source heterogeneous monitoring data in the tunnel through multiple types of sensor nodes; A data processing module, configured to perform standardization processing on the multi-source heterogeneous monitoring data to obtain standardized tunnel monitoring data; A synchronization and alignment module is used to synchronize and align the standardized tunnel monitoring data, process time differences, missing values and abnormal values of different data sources, and obtain tunnel monitoring time synchronization data; A feature extraction module is used to extract multi-dimensional features from the standardized tunnel monitoring data to obtain tunnel status feature data; A status warning module is used to build a deep learning warning model to identify the tunnel status feature data and obtain tunnel warning data; The monitoring and management module is used to push tunnel status feature data and tunnel warning data to management personnel through multi-channel information push, so that management personnel can monitor the tunnel status of the smart tunnel in real time.

9. An electronic device, characterized in that: include: at least one processor, at least one memory, a communication interface, and a bus; The processor, memory, and communication interface communicate with each other via the bus, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions enable a computer to implement the method according to any one of claims 1 to 7.

Images