Highway disaster event rapid identification alarm system
By using AI vision technology and segmented alarm mechanism in the rapid identification alarm system for road disaster incidents, the problem of manual delays in the existing technology is solved, and rapid and effective disasters are achieved Alarm and risk aversion reminders reduce the risks of casualties and secondary disasters.
Patent Information
- Application Number
- CN202411214071.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-31
- Publication Date
- 2025-01-03
AI Technical Summary
The existing road disaster event identification alarm system relies on manual confirmation when disaster emerges, resulting in delays and inability to issue hazard alarms to vehicles in a timely manner, and cannot effectively reduce the impact of casualties and secondary disasters.
A quick identification alarm system for road disaster events was designed, using cameras to collect image information, automatically judge disaster events through AI vision technology and trigger segmented alarm mechanisms, including sound and light alarm modules and intelligence boards, reminding vehicles close to the disaster location. Emergency braking or diversion, long-distance vehicles publish alarm information through text messages to reduce manual intervention and information delays.
It realizes rapid identification and alarm within 1 second after the disaster, reduces the impact of casualties and secondary disasters, has high system reliability, low cost, and strong applicability. It can effectively remind vehicles to avoid hazards during poor night vision.
Smart Images

Figure CN119251982A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of highway disaster alarm, and in particular to a highway disaster event rapid identification and alarm system. Background Art
[0002] With the rapid development of information technology, the hazard warning technology of structures has reached a relatively mature stage and can predict and identify potential structural risks. However, although the warning technology is very advanced in theory, in practical applications, its high cost often becomes an obstacle to popularization, resulting in the inability of transportation infrastructure such as highway bridges to achieve comprehensive risk coverage through early warning technology. Therefore, although early warning technology has great potential in preventing disaster accidents, disaster events such as highway bridge collapse are still inevitable and occur from time to time. For example, the road collapse of Meida Expressway in May 2024, the unilateral collapse of the highway bridge of Shuiyang section of Danning Expressway in July, and the sudden collapse of the tunnel bridge of Sichuan Yakang Expressway in August. The above accidents all occurred at night when the visibility was poor. When a disaster event suddenly occurs, the initial casualties may be relatively small, but after the accident, due to obstructed vision, poor vision, lack of disaster alarm, and delayed information transmission, subsequent vehicles are unaware of the disaster ahead and cannot avoid risks in time, which increases the casualty rate and leads to further expansion of the situation. Therefore, under the premise that disasters are inevitable, it is particularly necessary to establish a timely and effective highway disaster event rapid identification alarm device and system.
[0003] At present, there is a Chinese invention patent with application number CN202310314154.7 that uses AI analysis of road videos to timely warn of abnormal situations on the system platform, and intervenes in advance after manual confirmation to avoid traffic accidents. However, this method is mostly applicable to disaster warning. When a disaster occurs, manual confirmation delays the timing, and this method only alarms on the system platform. Subsequent vehicles are unaware of the disaster ahead and cannot avoid danger in time, and secondary casualties cannot be avoided. The Chinese invention patent application with application number CN202310948226.3 provides a landslide disaster identification method and system based on deep learning. The camera captures the data to the IoT video gateway for data displacement comparison and analysis, and the platform issues warning information, but it also only alarms on the system platform, and cannot issue risk avoidance alarms in time at the disaster site, and cannot avoid secondary casualties. In addition, the above patent cannot remind vehicles on the disaster driving route to leave from a distance, which can easily cause vehicle congestion on the road section near the disaster, increasing the difficulty of disaster rescue. The above are all situations that people in this technical field do not want to see again and hope to be improved. Summary of the invention
[0004] In view of the above, it is necessary to provide a highway disaster event rapid identification and alarm system, which can quickly identify safety hazards and disaster accidents, and use a variety of alarm methods to remind vehicles on the highway section to emergency brake or divert, etc., which can greatly reduce casualties and the impact of secondary disasters.
[0005] In order to achieve the above object, the technical solution adopted by the present invention is:
[0006] A highway disaster event rapid identification and alarm system, comprising
[0007] A power supply device is used to provide power support for each power-consuming module of the system;
[0008] Cameras, installed on the highway route, are used to collect image information of the highway;
[0009] An alarm device is installed on the highway route, including an audible and visual alarm module and an information board, wherein the audible and visual alarm module is installed on both sides of the highway, and the information board is installed at the entrance and ramp of the highway;
[0010] The alarm control integrated machine is used to receive the image information collected by the camera, analyze the image information data through AI visual technology, automatically determine whether a disaster event occurs on the highway, and locate the location information of the disaster when a disaster event is determined to occur, and trigger the disaster alarm mechanism to send warning information to the outside world, and at the same time synchronize the information to the system platform; wherein the disaster alarm mechanism is a segmented alarm mechanism, specifically, 1) for the road section within 1km from the disaster location, the alarm control integrated machine drives the sound and light alarm module of the alarm device to start to remind the vehicles within the road section to take emergency braking and avoidance measures; 2) for the road section within 1~10km from the disaster location, the alarm control integrated machine drives the information board of the alarm device to start to induce and remind the vehicles within the road section to take diversion measures; 3) for the road section beyond 10km from the disaster location, the alarm control integrated machine sends warning information to the driver, passengers and supervisors by sending text messages;
[0011] The system platform, as the system master control, is used to receive and view various information uploaded by the alarm control integrated machine, and can issue control instructions to the alarm control integrated machine.
[0012] Preferably, the power supply device is a solar energy device consisting of a solar photovoltaic panel and a solar energy controller.
[0013] Preferably, the camera uses a disaster identification camera with 4 million pixels, 23 times optical zoom, focal length 4mm, and infrared night vision of 150 meters.
[0014] Preferably, the power supply device, camera, and alarm control integrated machine at the same position are installed on the main control monitoring pole.
[0015] Preferably, the sound and light alarm module includes but is not limited to a tweeter, a flashing light, a red and green warning light, and a projection spotlight.
[0016] Preferably, the alarm device has a manual alarm trigger button, which can be controlled to start and stop by the manual alarm trigger button.
[0017] Preferably, the alarm control integrated machine is connected to each alarm device via lora wireless communication.
[0018] Compared with the prior art, the present invention has the following beneficial effects:
[0019] 1. The system of the present invention collects image information on the highway route through a camera, and drives the alarm device to send out alarm information based on the automatic analysis result of the image information by the alarm control integrated machine, wherein the vehicles near the accident are reminded to brake urgently by means of sound and light alarm, and the vehicles on the driving route far from the accident site are reminded to take measures to leave and divert by means of information release such as information boards, and at the same time, the information is synchronized to the system platform to inform the management and maintenance unit to make emergency plans. Such a segmented alarm mechanism can minimize casualties and the impact of secondary disasters.
[0020] 2. The present invention discloses a rapid identification and alarm system for highway disaster events, which does not require excessive manual intervention. When necessary, only on-duty personnel are required to confirm the accident, thereby reducing problems such as information delays caused by human judgment and reaction. In addition, the high-performance edge computing and alarm integrated equipment used in the present invention can realize identification and alarm within 1 second after the disaster occurs. The alarms are independent of each other, and single or multiple failures will not affect the device alarm, thereby ensuring the reliability of the complete set of equipment.
[0021] 3. The present invention is simple and has strong applicability. Compared with real-time structural monitoring equipment, the present invention can achieve rapid disaster identification and alarm while having a lower cost and has a high promotion value. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] Figure 1 It is a schematic diagram of the segmented alarm mechanism of the present invention.
[0023] Figure 2 Schematic diagram of the interface of the system platform of the present invention, wherein (a) is a statistical dashboard and (b) is an alarm record display interface.
[0024] Figure 3 This is the installation structure diagram of the equipment on the main control monitoring pole.
[0025] Figure 4It is a topological diagram of the alarm control integrated machine associated with multiple alarm devices through LoRa wireless communication.
[0026] Figure 5 This is a simple schematic diagram of the system setup.
[0027] Figure 6 It is a schematic diagram of hardware construction used on actual roads.
[0028] Figure 7 This is a schematic diagram of the platform disaster alarm information interface for the G358 (Shishui Line) roadbed slope disaster alarm.
[0029] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION
[0030] In a preferred embodiment of the present invention, a highway disaster event rapid identification and alarm system comprises:
[0031] The power supply device is used to provide power support for each power module of the system. Specifically, the power supply device is preferably a solar device consisting of a solar photovoltaic panel and a solar controller, so as to power the power modules of the system through solar energy, provide a stable power supply, achieve energy conservation and emission reduction, and be green and environmentally friendly.
[0032] The camera is installed on the highway route and is used to collect image information of the highway to provide image information for subsequent accurate and rapid identification of disasters and alarms. Preferably, the camera uses a disaster identification camera with 4 million pixels, 23x optical zoom, focal length of 4mm, and infrared night vision of 150 meters. In actual application, if a video surveillance system has been installed on the highway route, the image information collected by the installed video surveillance system can be used instead of the camera of the present invention to collect image information. At this time, real-time road monitoring video images can be obtained through the existing video surveillance system interface, which can provide data resources for intelligent disaster detection.
[0033] The alarm device is installed on the highway route, including an audible and visual alarm module and an information board, wherein the audible and visual alarm module is installed on both sides of the highway, and the information board is installed at the entrance and ramp of the highway; in this way, the alarm device can remind vehicles near the accident to brake urgently through audible and visual alarms, and remind vehicles on the driving route to take measures such as leaving or diverting by publishing information on the information board. Preferably, the audible and visual alarm module includes but is not limited to a tweeter, a flashing light, a red and green warning light, and a projection spotlight.
[0034] The alarm control integrated machine is used to receive the image information collected by the camera, analyze the image information data through AI visual technology, automatically determine whether a disaster event occurs on the highway, and when a disaster event is determined to occur, locate the location information of the disaster, and trigger the disaster alarm mechanism to send warning information to the outside world, and at the same time synchronize the information to the system platform; wherein the disaster alarm mechanism is a segmented alarm mechanism, specifically, 1) for the section within 1km of the disaster location, the alarm control integrated machine drives the sound and light alarm module of the alarm device to start to remind the vehicles within the section to take emergency braking and avoidance measures; 2) for the section within 1~10km of the disaster location, the alarm control integrated machine drives the information board of the alarm device to start to induce and remind the vehicles within the section to take diversion measures; 3) for the section beyond 10km from the disaster location, the alarm control integrated machine sends warning information to the driver, passengers and supervisors by sending text messages. The specific schematic diagram can be seen Figure 1 . That is, the alarm control integrated machine of the present invention is responsible for edge computing of the camera captured images, issuing control instructions to the alarm device and synchronizing information to the system platform. Among them, the alarm control integrated machine automatically determines whether landslides, landslides, collapses and other disaster events occur in roads, bridges, slopes and other land transportation infrastructure based on AI visual technology, and triggers the disaster alarm mechanism. The AI visual technology used is to optimize the AI model algorithm, which is improved on the basis of the existing AI model algorithm. The specific improvements are: (1) adding image segmentation and target detection: among them, image segmentation is used to distinguish road areas, and target detection is used to detect disaster targets; (2) using OpenVINO technology to optimize the deployment model: OpenVINO is a unified deep learning deployment solution launched by Intel, which aims to optimize the model and improve the performance on various Intel hardware, and improve the computing speed by about 8-10 times. (3) Add various road and bridge data, as well as video data under various weather, lighting and other environments, to continuously optimize the training model. (4) For the relatively small amount of real road disaster data, image synthesis technology is used to simulate disaster samples and continuously optimize the training model. Therefore, the AI visual technology used in the present invention can improve the recognition rate of disasters. It relies on maintenance technicians to continuously mark and identify disaster samples, and forms a target detection model, image large model, road boundary recognition model, road scene detection model and other AI model algorithm libraries based on video images according to different disaster characteristics and recognition difficulty, which can realize efficient and accurate identification and alarm of highway disasters. Furthermore, the disaster alarm mechanism triggered by the alarm control integrated machine of the present invention adopts a segmented alarm mechanism, so that different alarm methods can be used in different sections of the road, so as to achieve effective reminders and warnings, and minimize casualties and the impact of secondary disasters.
[0035] The system platform, as the system master control, is used to receive and view various information uploaded by the alarm control integrated machine, and can issue control instructions to the alarm control integrated machine. That is, the system platform can realize emergency response functions such as rapid alarm and information release of accidents, and can also facilitate real-time viewing of risk status information of the entire highway, and realize the viewing and control of alarm equipment, cameras, information boards and other electromechanical equipment along the entire road section. In actual application, the system platform can use "one picture" to view the risk status information of the entire highway and detailed monitoring information of a single structure in real time; the system platform can realize the viewing and control of alarm equipment, cameras, information boards and other electromechanical equipment along the entire road section. In addition, the system platform includes functions such as emergency plan management, rapid warning response to emergency events, and multi-department coordination and command and dispatch. Based on disaster warning information, combined with transportation infrastructure disaster identification, rapid alarm technology and equipment and software systems, it realizes alarm information release and emergency disposal, disaster reporting management and other business applications. The specific system platform can be as follows Figure 2 shown.
[0036] In addition, it should be noted that in this embodiment, the power supply device, camera, and alarm control integrated machine at the same position are installed on the main control monitoring pole, and the main control monitoring pole is placed in a position where a panoramic view of the highway can be captured. At this time, the installation structure diagram of the equipment on the main control monitoring pole is as follows: Figure 3 As shown, this makes it convenient for the power supply device to provide power support for the camera and alarm control all-in-one machine here.
[0037] In this embodiment, the alarm device has a manual alarm trigger button, which can be controlled to start and stop by the manual alarm trigger button, that is, through the setting of the manual alarm trigger button, the alarm device of the present invention can also be manually triggered. In this way, the alarm device of the present invention has three trigger mechanisms: the first is fully automatic triggering, which automatically triggers the alarm mechanism after the AI visual technology identifies the occurrence of disasters on bridges and slopes. This triggering mechanism can reduce the problems such as information delay caused by human judgment and reaction, and reduce the damage caused by secondary disasters; the second is human-computer interaction triggering, after the AI visual technology identifies the occurrence of disaster damage, it is transmitted to the system platform through the cloud. The system platform is a 24h monitoring center. In this way, the monitoring personnel trigger the alarm mechanism after retrieving the scene picture for secondary confirmation. The secondary confirmation can confirm whether the disaster section problem is solved and whether it is still necessary to continue to issue disaster alarms; the third is manual triggering, which is triggered manually by monitoring, maintenance or driver personnel at the system platform through the monitoring center to view the disaster accident or the disaster accident scene. The three triggering mechanisms work together to achieve the fastest speed and maximum avoidance of the expansion of casualties after the disaster.
[0038] Finally, it should be noted that the alarm control integrated machine is connected to each alarm device through lora wireless communication. The various types of alarm devices (tweeters, flashing lights, red and green alarm lights, projection spotlights and information boards) are independent of each other, such as Figure 4 The alarm devices at the same location are of various types or in various numbers, and more than one information board is set at the same location, so that if an alarm device or information board is damaged and cannot work, it will not affect the operation of other alarm devices.
[0039] Based on the above system, a simple system structure diagram is built, as shown Figure 5 In addition, the system proposed by the present invention is also applied to actual highways, specifically the G358 National Highway maintained by the Guixi Highway Development Center. When applied, sound and light alarm devices are installed within a certain distance on both sides of the road according to the highway design speed. After a disaster occurs, the tweeter and warning lights can be used to effectively warn and remind passing vehicles in time to stop the vehicles in time. The hardware construction in the specific implementation is as follows Figure 6 As shown, the platform disaster alarm information is as follows Figure 7 shown.
[0040] The above description is a detailed description of the preferred feasible embodiments of the present invention, but the embodiments are not intended to limit the scope of the patent application of the present invention. All equivalent changes or modified changes completed under the technical spirit suggested by the present invention should fall within the patent scope covered by the present invention.
Claims
1. A highway disaster event rapid identification and alarm system, characterized by: include A power supply device is used to provide power support for each power-consuming module of the system; Cameras, installed on the highway route, are used to collect image information of the highway; An alarm device is installed on the highway route, including an audible and visual alarm module and an information board, wherein the audible and visual alarm module is installed on both sides of the highway, and the information board is installed at the entrance and ramp of the highway; The alarm control integrated machine is used to receive the image information collected by the camera, analyze the image information data through AI visual technology, automatically determine whether a disaster event occurs on the highway, and locate the location information of the disaster when a disaster event is determined to occur, and trigger the disaster alarm mechanism to send warning information to the outside world, and at the same time synchronize the information to the system platform; wherein the disaster alarm mechanism is a segmented alarm mechanism, specifically, 1) for the road section within 1km from the disaster location, the alarm control integrated machine drives the sound and light alarm module of the alarm device to start to remind the vehicles within the road section to take emergency braking and avoidance measures; 2) for the road section within 1~10km from the disaster location, the alarm control integrated machine drives the information board of the alarm device to start to induce and remind the vehicles within the road section to take diversion measures; 3) for the road section beyond 10km from the disaster location, the alarm control integrated machine sends warning information to the driver, passengers and supervisors by sending text messages; The system platform, as the system master control, is used to receive and view various information uploaded by the alarm control integrated machine, and can issue control instructions to the alarm control integrated machine.
2. A highway disaster event rapid identification and alarm system as claimed in claim 1, characterized in that: The power supply device is a solar energy device consisting of a solar photovoltaic panel and a solar energy controller.
3. A highway disaster event rapid identification and alarm system as claimed in claim 1, characterized in that: The camera uses a disaster identification camera with 4 million pixels, 23x optical zoom, focal length 4mm, and infrared night vision of 150 meters.
4. A highway disaster event rapid identification and alarm system as claimed in claim 1, characterized in that: The power supply device, camera and alarm control all-in-one machine at the same position are installed on the main control monitoring pole.
5. A highway disaster event rapid identification and alarm system as claimed in claim 1, characterized in that: The sound and light alarm module includes but is not limited to a tweeter, a flashing light, a red and green warning light, and a projection spotlight.
6. A highway disaster event rapid identification and alarm system as claimed in claim 1, characterized in that: The alarm device has a manual alarm trigger button, which can be controlled to start and stop by the manual alarm trigger button.
7. A highway disaster event rapid identification and alarm system as claimed in claim 1, characterized in that: The alarm control integrated machine is connected to each alarm device through lora wireless communication.
