Tunnel environment monitoring and early warning system and method based on edge computing and multi-modal fusion

Abstract

The present application belongs to the technical field of tunnel engineering safety monitoring, and particularly relates to a tunnel environment monitoring and early warning system and method based on edge computing and multi-modal fusion, comprising: a multi-source perception collection module arranged at different monitoring sections in the tunnel to collect original monitoring data of the tunnel; an edge computing and event generation module connected to the multi-source perception collection module, which pre-processes the collected original monitoring data, screens abnormal data, and generates event data packets; a reliable communication module connected to the edge computing and event generation module, which transmits the event data packets and the original monitoring data to the cloud; a cloud fusion decision module which receives data through the reliable communication module, calculates a comprehensive risk score in combination with a digital twin model and a multi-modal fusion algorithm; and a hierarchical early warning and linkage control module connected to the cloud fusion decision module and field execution equipment, which performs hierarchical early warning and linkage control according to the comprehensive risk score.

Description

Technical Field

[0001] This invention belongs to the field of tunnel engineering safety monitoring technology, specifically relating to a tunnel environment monitoring and early warning system and method based on edge computing and multimodal fusion. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Tunnel construction may involve safety risks such as the accumulation of toxic and harmful gases, sudden inrush of groundwater, instability of the tunnel face, deformation of surrounding rock, cracking of the lining, and personnel accidentally entering dangerous areas. During tunnel operation, problems may arise such as fire smoke, abnormal vehicle stops, road surface flooding, ventilation failure, abnormal lighting, long-term structural deformation, and leakage. These risks are characterized by their suddenness, interconnectedness, and concealment, placing high demands on the real-time performance, accuracy, and coordinated response capabilities of the monitoring system.

[0004] Existing tunnel environmental monitoring systems mostly employ single sensors or distributed subsystems. In practical applications, environmental monitoring, structural monitoring, water hazard monitoring, and video surveillance often operate independently, lacking unified fusion and analysis of multi-source heterogeneous data. This makes it difficult to promptly identify complex risks such as structural deformation accompanied by leakage, abnormal gas concentration accompanied by ventilation failure, and road surface water accumulation accompanied by drainage system blockage. The tunnel interior is long, narrow, and enclosed, and communication links are significantly affected by factors such as deployment conditions, construction disturbances, electromagnetic interference, and insufficient public network coverage. Uploading all raw video streams and high-frequency sensor data to the cloud not only consumes a large amount of bandwidth but also easily leads to delays and data loss risks. Most systems use static threshold alarms, failing to fully consider the tunnel's engineering stage, spatial location, historical trends, and consistency between multi-source data. They generally remain at the stage of "anomaly detection - manual inspection - manual handling," lacking sufficient linkage capabilities with on-site equipment such as ventilation, drainage, lighting, traffic guidance, and audible and visual alarms, making it difficult to achieve rapid response in the early stages of risks. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes a tunnel environment monitoring and early warning system and method based on edge computing and multimodal fusion. The system performs data cleaning, spatiotemporal alignment, anomaly screening, and event packet generation at the edge, while multimodal fusion, trend analysis, digital twin mapping, and comprehensive risk scoring are performed at the cloud. Combined with safety interlocking conditions to control field equipment, it achieves timely identification, tiered early warning, and closed-loop handling of tunnel environmental risks. This solves the problems of fragmented multi-source data, high pressure on raw data transmission, delayed early warning response, poor adaptability to static thresholds, and lack of linked closed-loop control in existing tunnel monitoring systems.

[0006] According to some embodiments, the first solution of the present invention provides a tunnel environment monitoring and early warning system based on edge computing and multimodal fusion, which adopts the following technical solution: A tunnel environment monitoring and early warning system based on edge computing and multimodal fusion includes: The multi-source sensing and acquisition module is set up at different monitoring sections inside the tunnel to collect raw monitoring data of the tunnel; The edge computing and event generation module is connected to the multi-source sensing and acquisition module, which preprocesses the acquired raw monitoring data, filters abnormal data, and generates event data packets. A reliable communication module connects to the edge computing and event generation module, transmitting event data packets and raw monitoring data to the cloud. The cloud-based fusion decision-making module receives data through a reliable communication module and calculates a comprehensive risk score by combining a digital twin model and a multimodal fusion algorithm. The graded early warning and linkage control module is connected to the cloud-based integrated decision-making module and the on-site execution equipment, and performs graded early warning and linkage control based on the comprehensive risk score.

[0007] As a further technical limitation, the raw monitoring data acquired must include at least environmental parameters, hydrological parameters, structural state parameters, and visual image data.

[0008] As a further technical limitation, the multi-source sensing and acquisition module includes an environmental sensing unit, a hydrological sensing unit, a structural state sensing unit, and a machine vision sensing unit; wherein, the environmental sensing unit includes a CO sensor, an H2S sensor, a CH4 sensor, a temperature and humidity sensor, an anemometer, and a smoke detector; the hydrological sensing unit includes a water level gauge, a piezometer, a flow meter, a leakage monitoring sensor, or a surrounding rock moisture content sensor; the structural state sensing unit includes a vibrating wire strain gauge, a laser displacement gauge, a MEMS inclinometer, a crack gauge, a convergence meter, or a settlement monitoring device; the machine vision sensing unit includes a high-definition camera, an infrared thermal imager, or a low-light camera, used to identify events such as water inrush, water accumulation, smoke, open flame, personnel intrusion, abnormal vehicle parking, and abnormal equipment operation.

[0009] As a further technical limitation, the edge computing and event generation module includes a data access unit, a spatiotemporal alignment unit, an edge inference unit, an event packet generation unit, and a local caching unit. The data access unit receives sensor data, video stream data, and device operating status data. The spatiotemporal alignment unit synchronously processes multimodal data with different sampling frequencies and data formats according to a unified timestamp, spatial section number, and engineering stage label. The edge inference unit performs anomaly identification on video images and sensor data, and outputs anomaly type, anomaly location, anomaly confidence level, and keyframe images. The event packet generation unit encapsulates the anomaly identification results, sensor feature values, change trends, keyframe images, and preliminary handling suggestions into event data packets. The local caching unit caches monitoring data and event data packets when communication is interrupted or network quality falls below a set threshold, and retransmits them after communication is restored.

[0010] As a further technical limitation, the cloud-based fusion decision-making module includes a data storage unit, a digital twin mapping unit, a stage strategy library, a risk scoring unit, and a model update unit. The data storage unit stores historical monitoring data, event data packages, and handling records for the tunnel construction, operation, and maintenance periods. The digital twin mapping unit maps real-time monitoring data, event locations, and risk levels to a 3D tunnel model or a planar unfolded model. The stage strategy library stores monitoring indicators, sampling frequencies, early warning thresholds, risk weights, and linkage rules corresponding to different engineering stages. The risk scoring unit calculates a comprehensive risk score based on threshold deviation, changing trends, visual recognition confidence, engineering stage weights, data consistency, and equipment operating status. The model update unit updates the risk scoring parameters and recognition model based on historical events, manual review results, and on-site handling results.

[0011] As a further technical limitation, the graded early warning and linkage control module includes a risk level determination unit, an information release unit, a safety interlock determination unit, and an equipment linkage control unit. The risk level determination unit is used to classify the comprehensive risk score into general concern, yellow warning, orange warning, and red warning levels. The information release unit is used to release early warning information via audible and visual alarms, SMS, APP, monitoring screens, broadcast systems, or variable message signs. The safety interlock determination unit is used to determine whether the status of on-site equipment, personnel location, power supply status, ventilation status, and manual authorization conditions meet the linkage requirements. The equipment linkage control unit is used to control ventilation equipment, drainage equipment, lighting equipment, alarm equipment, traffic guidance equipment, or non-essential power circuits via PLC, DO interface, industrial gateway, or relays when the safety interlock conditions are met.

[0012] According to some embodiments, the second aspect of the present invention provides a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion, which adopts a tunnel environment monitoring and early warning system based on edge computing and multimodal fusion provided in the first aspect, and adopts the following technical solution: A tunnel environment monitoring and early warning method based on edge computing and multimodal fusion includes: Obtain raw monitoring data of the tunnel; Based on the acquired raw tunnel monitoring data, abnormal data is filtered out and event data packets are generated; A comprehensive risk score is calculated based on event data packets, digital twin models, and multimodal fusion algorithms. Based on the comprehensive risk score, graded early warning and coordinated control are implemented to complete tunnel environment monitoring and early warning based on edge computing and multimodal fusion.

[0013] According to some embodiments, a third aspect of the present invention provides a computer-readable storage medium, employing the following technical solution: A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion as described in the second aspect of the present invention.

[0014] According to some embodiments, the fourth aspect of the present invention provides an electronic device, which adopts the following technical solution: An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the tunnel environment monitoring and early warning method based on edge computing and multimodal fusion as described in the second aspect of the present invention.

[0015] According to some embodiments, the fifth aspect of the present invention provides a computer program product, which adopts the following technical solution: A computer program product includes software code, wherein the program in the software code performs the steps of a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion as described in the second aspect of the present invention.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention achieves timely identification, graded early warning, and closed-loop handling of tunnel environmental risks by performing data cleaning, spatiotemporal alignment, anomaly screening, and event packet generation at the edge, and multimodal fusion, trend analysis, digital twin mapping, and comprehensive risk scoring at the cloud. Combined with safety interlocking conditions to control field equipment, it reduces redundant transmission of tunnel monitoring data, improves the timeliness of anomaly identification and the accuracy of early warning in multi-hazard coupled scenarios, and is suitable for safety monitoring and risk early warning during tunnel construction, operation, and maintenance. Attached Figure Description

[0017] The accompanying drawings, which form part of this embodiment, are used to provide a further understanding of this embodiment. The illustrative embodiments and their descriptions are used to explain this embodiment and do not constitute an improper limitation of this embodiment.

[0018] Figure 1 This is a schematic diagram of the structure of a tunnel environment monitoring and early warning system based on edge computing and multimodal fusion according to Embodiment 1 of the present invention; Figure 2 This is a data processing flowchart of the edge computing and event generation module in Embodiment 1 of the present invention; Figure 3 This is a flowchart of the multimodal data fusion and comprehensive risk scoring process in Embodiment 1 of the present invention; Figure 4 This is a flowchart of the hierarchical early warning and linkage control in Embodiment 1 of the present invention; Figure 5 This is a flowchart of a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion, as described in Embodiment 2 of the present invention. Detailed Implementation

[0019] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0020] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0021] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0022] In this invention, terms such as "upper," "lower," "left," "right," "front," "back," "vertical," "horizontal," "side," and "bottom" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used only to facilitate the description of the structural relationships of the various components or elements of this invention and do not specifically refer to any component or element in this invention. They should not be construed as limiting the invention.

[0023] In this invention, terms such as "fixed connection," "connected," and "linked" should be interpreted broadly, indicating a fixed connection, an integral connection, or a detachable connection; a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can determine the specific meaning of these terms in this invention based on the specific circumstances, and they should not be construed as limitations on the invention.

[0024] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0025] Example 1 Embodiment 1 of this invention introduces a tunnel environment monitoring and early warning system based on edge computing and multimodal fusion.

[0026] like Figure 1 The tunnel environment monitoring and early warning system shown includes: The multi-source sensing and acquisition module is set up at different monitoring sections inside the tunnel to collect raw monitoring data of the tunnel; The edge computing and event generation module is connected to the multi-source sensing and acquisition module, which preprocesses the acquired raw monitoring data, filters abnormal data, and generates event data packets. A reliable communication module connects to the edge computing and event generation module, transmitting event data packets and raw monitoring data to the cloud. The cloud-based fusion decision-making module receives data through a reliable communication module and calculates a comprehensive risk score by combining a digital twin model and a multimodal fusion algorithm. The graded early warning and linkage control module is connected to the cloud-based integrated decision-making module and the on-site execution equipment, and performs graded early warning and linkage control based on the comprehensive risk score.

[0027] As one or more implementation methods, the multi-source sensing and acquisition module in this embodiment includes an environmental sensing unit, a hydrological sensing unit, a structural state sensing unit, and a machine vision sensing unit. The environmental sensing unit includes a CO sensor, an H2S sensor, a CH4 sensor, a temperature and humidity sensor, an anemometer, and a smoke detector. The CO and H2S sensors employ electrochemical sensing methods, while the CH4 sensor employs infrared, catalytic combustion, or semiconductor combustible gas sensing methods. The hydrological sensing unit includes a water level gauge, a piezometer, a flow meter, a leakage monitoring sensor, or a surrounding rock moisture content sensor. The structural state sensing unit includes a vibrating wire strain gauge, a laser displacement gauge, a MEMS inclinometer, a crack gauge, a convergence meter, or a settlement monitoring device. The machine vision sensing unit includes a high-definition camera, an infrared thermal imager, or a low-light camera, used to identify events such as water inrush, water accumulation, smoke, open flames, personnel intrusion, abnormal vehicle parking, and abnormal equipment operation.

[0028] This embodiment uses a multi-source sensing and acquisition module to collect tunnel environmental parameters, hydrological parameters, structural state parameters, and visual image data. Among them, environmental parameters include CO, H2S, CH4, temperature, humidity, wind speed, wind direction, and smoke concentration; hydrological parameters include water level, seepage pressure, flow rate, leakage status, and water content of surrounding rock; structural state parameters include strain, displacement, dip angle, crack width, convergence deformation, and settlement changes; and visual image data includes visible light images, infrared thermal imaging images, and video streams.

[0029] As one or more implementation methods, the edge computing and event generation module in this embodiment is as follows: Figure 2 As shown, deployed at the convergence node in the tunnel site, the edge computing and event generation module performs local cleaning, noise reduction, outlier removal, time synchronization, spatial cross-section alignment, and feature extraction on the collected data. For video image data, the edge computing and event generation module uses a lightweight image recognition model to identify water inrush, water accumulation, smoke, open flames, personnel intrusion, abnormal vehicle stops, and equipment malfunctions. For sensor data, the edge computing and event generation module performs initial anomaly screening based on threshold deviation, trend abrupt changes, slope changes, and multi-source consistency judgments. After processing, the edge computing and event generation module generates event data packets.

[0030] The event data package in this embodiment includes event number, event type, occurrence time, occurrence location, monitoring section, engineering stage label, sensor feature value, visual recognition confidence level, key frame image, risk confidence level, and preliminary handling suggestions.

[0031] In this embodiment, the edge computing and event generation module includes a data access unit, a spatiotemporal alignment unit, an edge inference unit, an event packet generation unit, and a local caching unit. The data access unit receives sensor data, video stream data, and device operating status data. The spatiotemporal alignment unit synchronizes multimodal data with different sampling frequencies and data formats according to a unified timestamp, spatial section number, and engineering stage label. The edge inference unit identifies anomalies in video images and sensor data and outputs the anomaly type, anomaly location, anomaly confidence level, and keyframe images. The event packet generation unit encapsulates the anomaly identification results, sensor feature values, change trends, keyframe images, and preliminary handling suggestions into event data packets. The local caching unit caches monitoring data and event data packets when communication is interrupted or network quality is below a set threshold, and retransmits them after communication is restored.

[0032] It should be noted that the edge computing and event generation module assigns a unified timestamp, spatial section number and engineering stage label to the collected data, and converts the multimodal raw data into event data packages containing event type, event location, risk confidence, change trend, key frame image and suggested handling strategy, so as to realize the edge side preliminary judgment of tunnel environmental risks, cloud side comprehensive analysis and on-site closed-loop response.

[0033] As one or more implementation methods, the reliable communication module in this embodiment uses a combination of wired main link and wireless backup link to achieve data transmission; wherein, data aggregation between sensor nodes and edge computing nodes can be carried out using RS485, CAN, industrial Ethernet, LoRa or LoRa-Mesh; data backhaul between edge computing nodes and cloud fusion decision module can be carried out using fiber optic ring network, 4G, 5G or NB-IoT backup link; when the communication link is abnormal, the system automatically switches to the backup communication method and reduces the risk of data loss through local caching and breakpoint resume mechanism.

[0034] As one or more implementation methods, the cloud-based fusion decision module in this embodiment is as follows: Figure 3 As shown, the system includes a data storage unit, a digital twin mapping unit, a stage strategy library, a risk scoring unit, and a model update unit. The data storage unit stores historical monitoring data, event data packages, and handling records for the tunnel construction, operation, and maintenance periods. The digital twin mapping unit maps real-time monitoring data, event locations, and risk levels to the tunnel's 3D model or planar unfolded model. The stage strategy library stores monitoring indicators, sampling frequencies, early warning thresholds, risk weights, and linkage rules corresponding to different engineering stages. The risk scoring unit calculates a comprehensive risk score based on threshold deviation, changing trends, visual recognition confidence, engineering stage weights, data consistency, and equipment operating status. The model update unit updates the risk scoring parameters and recognition model based on historical events, manual review results, and on-site handling results.

[0035] In this embodiment, the cloud-based fusion decision module receives event data packets and monitoring data, and calculates a comprehensive risk score based on a historical monitoring database, an engineering phase strategy library, a digital twin model, and a multimodal fusion algorithm. The engineering phase strategy library includes monitoring indicators, sampling frequencies, risk weights, alarm thresholds, and linkage control rules corresponding to the construction excavation phase, initial support phase, secondary lining construction phase, operation phase, and maintenance phase. The digital twin model maps monitoring data, abnormal events, and risk levels to a three-dimensional or planar unfolded model of the tunnel, assisting managers in quickly determining the location and scope of risk.

[0036] In this embodiment, the comprehensive risk score R is determined according to the following factors, namely R = f(S, T, V, G, C, E); where S represents the degree of deviation of the monitored value from the safety threshold, T represents the trend of the monitored data within the set time window, V represents the confidence level of the machine vision recognition result, G represents the risk weight corresponding to the current engineering stage, C represents the degree of consistency or conflict between multi-source data, and E represents the operating status of the on-site execution equipment. When the comprehensive risk score R reaches the set level threshold, the system generates a warning message of the corresponding level and calls the corresponding linkage control rules.

[0037] As one or more implementation methods, the hierarchical early warning and linkage control module in this embodiment is as follows: Figure 4 As shown, it includes a risk level determination unit, an information release unit, a safety interlock determination unit, and an equipment linkage control unit. The risk level determination unit is used to classify the comprehensive risk score into general concern, yellow alert, orange alert, and red alert levels. The information release unit is used to release warning information via audible and visual alarms, SMS, APP, monitoring screens, broadcast systems, or variable message signs. The safety interlock determination unit is used to determine whether the status of on-site equipment, personnel location, power supply status, ventilation status, and manual authorization conditions meet the linkage requirements. The equipment linkage control unit is used to control ventilation equipment, drainage equipment, lighting equipment, alarm equipment, traffic guidance equipment, or non-essential power circuits via PLC, DO interface, industrial gateway, or relays when the safety interlock conditions are met.

[0038] In this embodiment, the graded early warning and linkage control module determines the early warning level based on a comprehensive risk score and disseminates early warning information via SMS, APP, audible and visual alarms, broadcasting systems, monitoring screens, and variable message signs. When the early warning level reaches the set conditions, the system judges the status of on-site equipment, personnel location, power supply status, and manual authorization conditions according to safety interlock rules. After the interlock conditions are met, it activates ventilation equipment, drainage equipment, lighting equipment, traffic guidance equipment, alarm equipment, or designated non-essential power circuits to achieve a closed-loop risk management system.

[0039] The tunnel environment intelligent monitoring and early warning system based on edge computing and multimodal fusion in this embodiment automatically or manually switches monitoring strategies according to different engineering stages of the tunnel—including the construction and excavation stage, the secondary lining support stage, and the traffic operation stage. The switching of monitoring strategies includes adjusting the sampling frequency, changing the early warning threshold, and redefining the alarm response process to meet the safety management needs of different stages. Specifically: During the construction and excavation phase, the system increases the sampling frequency and risk weight for monitoring gas concentration, water inrush, surrounding rock deformation, anomalies in the working face image, and intrusion by construction personnel. During the initial support and secondary lining construction stages, the sampling frequency and risk weight of convergence deformation, cracks, lining structure status, temperature and humidity, and leakage monitoring are increased. During the operation phase, the system increases the risk weights for monitoring fire smoke, CO concentration, traffic incidents, road water accumulation, ventilation and lighting equipment status, and long-term structural deformation. During the maintenance phase, the system adjusts the monitoring frequency, early warning thresholds, and inspection tasks based on the maintenance plan and historical disease locations.

[0040] Example 1: Environmental and Water Inrush Risk Monitoring in Drill-and-Blast Tunnel Construction This embodiment is applied to tunnel construction using the drill-and-blast method. During tunnel construction, there are risks such as sudden water inrush, accumulation of toxic and harmful gases, diffusion of smoke and dust after blasting, deformation of surrounding rock, and personnel accidentally entering dangerous areas at the tunnel face and behind it.

[0041] like Figure 1 As shown, the multi-source sensing and acquisition modules are deployed near the working face, the initial support area, the invert arch construction area, and near the secondary lining trolley. The environmental sensing unit includes a CO sensor, an H2S sensor, a CH4 sensor, a temperature and humidity sensor, and an anemometer; the hydrological sensing unit includes a water level gauge, a piezometer, and a leakage monitoring sensor; the structural status sensing unit includes a convergence meter, a crack gauge, and a MEMS inclinometer; and the machine vision sensing unit includes a low-light high-definition camera and an infrared thermal imager.

[0042] Edge computing and event generation modules are deployed within the tunnel near electrical distribution boxes or monitoring boxes to receive sensor data and video streams. For example... Figure 2 As shown, the edge computing and event generation module first performs outlier removal and filtering noise reduction on the sensor data, and then performs spatiotemporal alignment according to a unified timestamp and monitoring section number. For video images, the edge computing and event generation module identifies events such as water seepage at the working face, dripping water from the arch wall, water accumulation on the ground, smoke diffusion, and personnel approaching the danger zone, and extracts keyframes.

[0043] When the system detects rapid changes in water level or seepage pressure near the working face, and the visual image identifies continuous water flow textures at the working face or arch foot, the edge computing and event generation module generates a "suspected sudden water inrush event" data packet. This event data packet includes the location of the incident, image keyframes, water level change trends, seepage pressure change trends, and risk confidence levels. Combined with... Figure 3 The cloud-based integrated decision-making module increases the weight of water hazard risk based on the strategy library for the construction excavation stage, and calculates a comprehensive risk score by combining historical water inrush records. (Refer to...) Figure 4When the score reaches the orange alert level, the system will push warning information through the sound and light alarm and the APP, and start the drainage pump and emergency lighting after the safety interlock conditions are met; when the score reaches the red alert level, the system will further trigger broadcast prompts and on-site evacuation prompts.

[0044] This embodiment can comprehensively identify environmental, water hazards and structural anomalies in tunnels constructed using the drill-and-blast method, avoiding the delay in early warning caused by relying solely on a single water level threshold or manual inspection.

[0045] This embodiment of the intelligent tunnel environment monitoring and early warning system based on edge computing and multimodal fusion collects multimodal data, including gas concentration, temperature and humidity, wind speed, smoke, water level, seepage, structural deformation, crack changes, and video images, by deploying environmental, hydrological, structural, and visual sensing devices at different monitoring sections of the tunnel. The edge computing and event generation module performs data cleaning, spatiotemporal alignment, feature extraction, anomaly screening, and keyframe extraction on the collected data, generating event data packages containing event type, location, risk confidence level, trend characteristics, and handling suggestions. The cloud-based fusion decision-making module calculates a comprehensive risk score by combining historical monitoring data, engineering stage strategy library, and digital twin model. The hierarchical early warning and linkage control module links ventilation, drainage, lighting, audible and visual alarms, and information dissemination equipment according to risk level and safety interlocking conditions. This system can reduce redundant transmission of tunnel monitoring data, improve the timeliness of anomaly identification, and enhance the accuracy of early warning in multi-hazard coupled scenarios. It is suitable for safety monitoring and risk early warning during tunnel construction, operation, and maintenance.

[0046] Example 2 Embodiment 2 of the present invention introduces a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion.

[0047] like Figure 5 The tunnel environment monitoring and early warning method based on edge computing and multimodal fusion, as shown, includes: Obtain raw monitoring data of the tunnel; Based on the acquired raw tunnel monitoring data, abnormal data is filtered out and event data packets are generated; A comprehensive risk score is calculated based on event data packets, digital twin models, and multimodal fusion algorithms. Based on the comprehensive risk score, graded early warning and coordinated control are implemented to complete tunnel environment monitoring and early warning based on edge computing and multimodal fusion.

[0048] The detailed steps are the same as those of the tunnel environment monitoring and early warning system based on edge computing and multimodal fusion provided in Example 1, and will not be repeated here.

[0049] Example 3 Embodiment 3 of the present invention provides a computer-readable storage medium.

[0050] A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion as described in Embodiment 2 of the present invention.

[0051] The detailed steps are the same as those provided in Example 2 for a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion, and will not be repeated here.

[0052] Example 4 Embodiment 4 of the present invention provides an electronic device.

[0053] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the tunnel environment monitoring and early warning method based on edge computing and multimodal fusion as described in Embodiment 2 of the present invention.

[0054] The detailed steps are the same as those provided in Example 2 for a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion, and will not be repeated here.

[0055] Example 5 Embodiment 5 of the present invention provides a computer program product.

[0056] A computer program product includes software code, wherein the program in the software code performs the steps of a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion as described in Embodiment 2 of the present invention.

[0057] The detailed steps are the same as those provided in Example 2 for a tunnel environment monitoring and early warning method based on edge computing and multimodal fusion, and will not be repeated here.

[0058] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0059] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0060] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0061] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0062] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0063] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0064] The above description is merely a preferred embodiment of this practice and is not intended to limit the scope of this practice. Various modifications and variations can be made to this practice by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this practice should be included within the protection scope of this practice.

Claims

1. A tunnel environment monitoring and early warning system based on edge computing and multimodal fusion, characterized in that, include: The multi-source sensing and acquisition module is set up at different monitoring sections inside the tunnel to collect raw monitoring data of the tunnel; The edge computing and event generation module is connected to the multi-source sensing and acquisition module, which preprocesses the acquired raw monitoring data, filters abnormal data, and generates event data packets. A reliable communication module connects to the edge computing and event generation module, transmitting event data packets and raw monitoring data to the cloud. The cloud-based fusion decision-making module receives data through a reliable communication module and calculates a comprehensive risk score by combining a digital twin model and a multimodal fusion algorithm. The graded early warning and linkage control module is connected to the cloud-based integrated decision-making module and the on-site execution equipment, and performs graded early warning and linkage control based on the comprehensive risk score.

2. The tunnel environment monitoring and early warning system based on edge computing and multimodal fusion as described in claim 1, characterized in that, The raw monitoring data obtained includes at least environmental parameters, hydrological parameters, structural state parameters, and visual image data.

3. The tunnel environment monitoring and early warning system based on edge computing and multimodal fusion as described in claim 1, characterized in that, The multi-source sensing and acquisition module includes an environmental sensing unit, a hydrological sensing unit, a structural state sensing unit, and a machine vision sensing unit. The environmental sensing unit includes a CO sensor, an H2S sensor, a CH4 sensor, a temperature and humidity sensor, an anemometer, and a smoke detector. The hydrological sensing unit includes a water level gauge, a piezometer, a flow meter, a leakage monitoring sensor, or a surrounding rock moisture content sensor. The structural state sensing unit includes a vibrating wire strain gauge, a laser displacement gauge, a MEMS inclinometer, a crack gauge, a convergence meter, or a settlement monitoring device. The machine vision sensing unit includes a high-definition camera, an infrared thermal imager, or a low-light camera, used to identify events such as water inrush, water accumulation, smoke, open flames, personnel intrusion, abnormal vehicle stops, and abnormal equipment operation.

4. The tunnel environment monitoring and early warning system based on edge computing and multimodal fusion as described in claim 1, characterized in that, The edge computing and event generation module includes a data access unit, a spatiotemporal alignment unit, an edge inference unit, an event packet generation unit, and a local caching unit. The data access unit receives sensor data, video stream data, and device operating status data. The spatiotemporal alignment unit synchronizes multimodal data with different sampling frequencies and data formats according to a unified timestamp, spatial section number, and engineering stage label. The edge inference unit identifies anomalies in video images and sensor data and outputs the anomaly type, anomaly location, anomaly confidence level, and keyframe images. The event packet generation unit encapsulates the anomaly identification results, sensor feature values, change trends, keyframe images, and preliminary handling suggestions into event data packets. The local caching unit caches monitoring data and event data packets when communication is interrupted or network quality falls below a set threshold, and retransmits them after communication is restored.

5. The tunnel environment monitoring and early warning system based on edge computing and multimodal fusion as described in claim 1, characterized in that, The cloud-based fusion decision-making module includes a data storage unit, a digital twin mapping unit, a stage strategy library, a risk scoring unit, and a model update unit. The data storage unit stores historical monitoring data, event data packages, and handling records for the tunnel construction, operation, and maintenance periods. The digital twin mapping unit maps real-time monitoring data, event locations, and risk levels to a 3D or planar unfolded model of the tunnel. The stage strategy library stores monitoring indicators, sampling frequencies, early warning thresholds, risk weights, and linkage rules corresponding to different engineering stages. The risk scoring unit calculates a comprehensive risk score based on threshold deviation, changing trends, visual recognition confidence, engineering stage weights, data consistency, and equipment operating status. The model update unit updates the risk scoring parameters and recognition model based on historical events, manual review results, and on-site handling results.

6. The tunnel environment monitoring and early warning system based on edge computing and multimodal fusion as described in claim 1, characterized in that, The graded early warning and linkage control module includes a risk level determination unit, an information release unit, a safety interlock determination unit, and an equipment linkage control unit. The risk level determination unit classifies the comprehensive risk score into four levels: general concern, yellow alert, orange alert, and red alert. The information release unit releases early warning information via audible and visual alarms, SMS, APP, monitoring screens, broadcasting systems, or variable message signs. The safety interlock determination unit determines whether the status of on-site equipment, personnel location, power supply status, ventilation status, and manual authorization conditions meet the linkage requirements. When the safety interlock conditions are met, the equipment linkage control unit controls ventilation equipment, drainage equipment, lighting equipment, alarm equipment, traffic guidance equipment, or non-essential power circuits via PLC, DO interface, industrial gateway, or relays.

7. A tunnel environment monitoring and early warning method based on edge computing and multimodal fusion, comprising a tunnel environment monitoring and early warning system based on edge computing and multimodal fusion as described in any one of claims 1-6, characterized in that, include: Obtain raw monitoring data of the tunnel; Based on the acquired raw tunnel monitoring data, abnormal data is filtered out and event data packets are generated; A comprehensive risk score is calculated based on event data packets, digital twin models, and multimodal fusion algorithms. Based on the comprehensive risk score, graded early warning and coordinated control are implemented to complete tunnel environment monitoring and early warning based on edge computing and multimodal fusion.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the tunnel environment monitoring and early warning method based on edge computing and multimodal fusion as described in any one of claims 7.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the program, it implements the steps of the tunnel environment monitoring and early warning method based on edge computing and multimodal fusion as described in any one of claims 7.

10. A computer program product, comprising software code, characterized in that, The program in the software code executes the steps of the tunnel environment monitoring and early warning method based on edge computing and multimodal fusion as described in any one of claims 7.

Images