Traffic early warning system and control method thereof

Abstract

The invention relates to a traffic early warning system and a control method thereof, and the system comprises an intelligent spike which collects a road inclination angle, fence vibration and vehicle collision sound waves in real time, transmits the collected sound waves to an edge gateway, receives an instruction of the edge gateway, and forms an alarm; the camera is used for acquiring a pavement video and transmitting the pavement video to the edge gateway; and the edge gateway is used for accurately judging whether collapse or traffic accidents occur or not, and issuing a warning instruction to the intelligent spikes when the collapse or the traffic accidents are judged to occur. The method has the beneficial effects that 1, the road travel safety is improved, and the social positive value of a traffic system is improved; 2, the collected data can be shared to a highway operation unit, maintenance and analysis are carried out by the operation unit, and the operation efficiency is improved; 3, light induction is carried out on vehicles on the road, and the vehicles can be remotely guided to run according to lanes; 4, the video which is marked and stored can be used as a judgment basis of traffic accident liability; and 5, the economic effect of 84,000,000-11,000,000 yuan can be brought.

Description

A traffic early warning system and its control method Technical Field

[0001] This invention relates to the field of traffic technology, specifically to a traffic early warning system and its control method. Background Technology

[0002] Highways, as arteries for efficient logistics and travel, present particularly acute safety challenges due to the high speeds of vehicles. Data shows that rear-end collisions are the leading cause of accidents on highways, accounting for 42.61% of all accidents and causing a staggering 42.83% of fatalities.

[0003] An even more critical problem lies in the significant time lag in the traditional "manual discovery-alarm-response" response model after the initial accident. This causes the accident scene to transform into a dangerous "secondary accident trap" during the golden period for handling the situation: vehicles and pedestrians from the initial accident are exposed in the driving lane, while vehicles following behind, unable to receive timely warning information and lacking sufficient distance and time to brake effectively, are highly susceptible to triggering more destructive chain-reaction collisions, leading to increased casualties and exacerbated losses. This delay in warning from "incident" to "early warning" to "control" is the most fatal weakness in the current highway safety control system. Summary of the Invention

[0004] In summary, to address the shortcomings of existing technologies, the technical problem this invention aims to solve is to provide a traffic early warning system and its control method that addresses the problem of delayed road accident early warning response. It proposes a system capable of achieving second-level response and proactive early warning, fundamentally reducing "early warning delay" and preventing secondary accident chains.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: A traffic early warning system, comprising:

[0006] Smart road studs are installed on roads or guardrails at preset intervals. They collect road surface tilt angles, guardrail vibrations, and vehicle collision sound waves in real time and transmit them to the edge gateway. They also receive commands from the edge gateway and generate warnings.

[0007] The camera captures video of the road surface and transmits it to the edge gateway;

[0008] The edge gateway receives information data from smart road studs and cameras, and performs collaborative analysis on the dual-source data provided by the smart road studs and cameras. Based on the analysis results, it accurately determines whether a landslide or traffic accident has occurred. When a landslide or traffic accident is determined to have occurred, it issues a warning command to the smart road studs.

[0009] Based on the above technical solution, the present invention can be further improved as follows:

[0010] Furthermore, the smart road stud includes:

[0011] The sensor module collects information data in real time, including road surface tilt angle, guardrail vibration, and vehicle collision sound waves.

[0012] It includes a tilt sensor, a vibration sensor, and a microphone, wherein the tilt sensor, the vibration sensor, and the microphone respectively collect information data on road tilt angle, guardrail vibration, and vehicle collision sound waves in real time;

[0013] The control module is electrically connected to the LoRaWAN communication module and generates an early warning control signal after receiving an alarm command from the edge gateway through the LoRaWAN communication module.

[0014] The LED light module is electrically connected to the control module and performs a fast flashing action when it receives a warning control signal output by the control module.

[0015] The LoRaWAN communication module is electrically connected to the control module, receives data and instructions output by the control module, sends data to the edge gateway, listens for warning instructions issued by the edge gateway, and transmits the instructions to the control module.

[0016] The power supply module converts solar energy into electrical energy and provides power to the sensor module, the control module, the LED light module, and the LoRaWAN communication module.

[0017] Furthermore, the sensor module includes a tilt sensor, a vibration sensor, and a microphone, which respectively collect information data on road surface tilt angle, fence vibration, and vehicle collision sound waves in real time.

[0018] Furthermore, the power supply module includes a solar photovoltaic panel and a battery; the solar photovoltaic panel converts solar energy into electrical energy and charges the battery, and the battery provides power to the sensor module, the control module, the LED light module, and the LoRaWAN communication module respectively.

[0019] The present invention also provides a control method for a traffic warning system, which is implemented using the above-mentioned traffic warning system and includes the following steps:

[0020] Step 1: Deploy smart road studs, cameras, and edge gateways on the road;

[0021] Step 2: The smart road stud collects data on road surface tilt angle, road fence vibration and vehicle collision sound in real time. At the same time, the camera collects road video in real time, and the data collected by the smart road stud and the camera are transmitted to the edge gateway.

[0022] Step 3: The edge gateway identifies and analyzes the received data, and determines whether a landslide or traffic accident has occurred based on the analysis results.

[0023] Step 4: If the edge gateway determines that a landslide or traffic accident has occurred, it sends instructions to the smart road studs around the accident section.

[0024] Step 5: After receiving the command from the edge gateway, the control module of the smart road stud controls the LED light module to start flashing rapidly, reminding drivers of vehicles behind to pay attention to the road conditions.

[0025] Step 6: After seeing the LED light module flashing rapidly as a warning, the driver of the vehicle behind will increase their attention and slow down to observe the road conditions, ultimately reducing the probability of a secondary accident.

[0026] Step 7: After the incident is handled, staff access the edge gateway or mobile app via the web and send a recovery signal to all LED light modules in the warning state. After receiving the signal, the control module of the smart road stud controls the LED light modules to stop flashing rapidly.

[0027] Based on the above technical solution, the present invention can be further improved as follows:

[0028] Furthermore, the smart road studs transmit data wirelessly to the edge gateway via LoRaWAN, while the cameras transmit data to the edge gateway via wired Ethernet or 4G wireless network.

[0029] Furthermore, step 3 specifically includes:

[0030] Step 3.1: After receiving the sensor data from the smart road beacon, the edge gateway compares it with a preset threshold. If the data does not exceed the threshold, it is considered normal and continues to wait for the next round of data; if the data exceeds the threshold, it is marked as a suspected accident and enters the video review stage.

[0031] Step 3.2: The edge gateway decodes the video stream transmitted by the camera, extracts frames at 1-second intervals, and inputs the extracted frame images into a YOLO instance segmentation small model to initially identify vehicle stops, rollovers, collisions, and other abnormal events. If the small model identifies an anomaly, it is then re-checked through a large model to confirm the authenticity of the abnormal event. After the accident is finally confirmed, the abnormal event is labeled, encoded, and an alarm image or short video is generated, stored, and pushed to the management platform. If the verification fails to confirm, it is judged as a false alarm and no warning is triggered.

[0032] The beneficial effects of this invention are:

[0033] 1. It can strongly attract the attention of drivers of vehicles behind in low visibility conditions such as at night, giving them valuable extra reaction time, guiding vehicles to slow down in advance and maintain a safe distance, thereby eliminating the risk of secondary accidents in the bud, significantly improving the proactive safety control capabilities of highways and urban roads, improving road travel safety, and enhancing the positive social value of the transportation system.

[0034] 2. Smart road studs can be equipped with various sensors to monitor the condition of the road surface. The collected data can be shared with highway operators for maintenance and analysis, thereby improving operational efficiency.

[0035] 3. It can be connected to the traffic control center, which can directly take over the smart road studs and provide light guidance for vehicles on the road, and remotely guide vehicles to drive in the correct lanes.

[0036] 4. It can be connected to the traffic control center. The video of the camera recognizing and marking the sudden accident can be used as the basis for determining liability in traffic accidents.

[0037] 5. As of October 4, 2025, China's expressway network reached 191,000 kilometers. Assuming a 15%-20% probability of these sections being high-accident areas, 28,000 to 38,000 monitoring points would be needed. With an installation and deployment cost of approximately 300,000 yuan per point, this traffic early warning system and control method could generate an economic benefit of 8.4-11.4 billion yuan. Expanding to high-accident sections of national highways would bring even greater economic benefits. Attached Figure Description

[0038] Figure 1 shows the network topology of the traffic warning system;

[0039] Figure 2 is a structural block diagram of the smart road stud;

[0040] Figure 3 is a flowchart of the edge gateway analyzing the input video. Detailed Implementation

[0041] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0042] As shown in Figure 1, a traffic early warning system includes:

[0043] 1. Smart road studs

[0044] The smart road studs are installed at preset intervals on roads or guardrails. They collect real-time data on road surface tilt, guardrail vibration, and vehicle collision sound waves, transmitting this data to an edge gateway. Simultaneously, they receive commands from the edge gateway and generate warning signals. As shown in Figure 2, the smart road studs include a sensor module, a control module, an LED light module, a LoRaWAN communication module, and a power supply module. The sensor module collects real-time data on road surface tilt, guardrail vibration, and vehicle collision sound waves. The sensor module includes a tilt sensor, a vibration sensor, and a microphone, which respectively collect real-time data on road surface tilt, guardrail vibration, and vehicle collision sound waves. The control module is electrically connected to the LoRaWAN communication module and generates a warning control signal after receiving a warning command from the edge gateway via the LoRaWAN communication module. The LED light module is electrically connected to the control module and performs a fast-flashing action when it receives the warning control signal output by the control module. The LoRaWAN communication module is electrically connected to the control module and receives data and commands output by the control module. It sends data to the edge gateway and listens for warning commands issued by the edge gateway, transmitting these commands to the control module. This wirelessly triggers the warning control signal to activate the LED light modules of smart road studs within a certain range. The power supply module converts solar energy into electrical energy and provides power to the sensor module, control module, LED light module, and LoRaWAN communication module. The power supply module includes a solar photovoltaic panel and a battery. The solar photovoltaic panel converts solar energy into electrical energy and charges the battery, which in turn provides power to the sensor module, control module, LED light module, and LoRaWAN communication module.

[0045] 2. Camera

[0046] The camera is responsible for collecting video information about the road surface and transmitting the data back to the edge gateway via a wired or wireless network. The camera should be a high-definition (2MP) camera or higher and support the Onvif protocol.

[0047] 3. Edge Gateway

[0048] The edge gateway receives information data from smart road studs and cameras, and performs collaborative analysis on the dual-source data provided by the smart road studs and cameras. Based on the analysis results, it accurately determines whether a landslide or traffic accident has occurred, and sends a warning command to the smart road studs when a landslide or traffic accident is detected. The LoRaWAN edge gateway is installed on the side of the road, powered by mains electricity, and operates 24 hours a day.

[0049] This traffic warning system consists of smart road studs, cameras, and an edge gateway. The smart road studs collect various road surface data, including road slope, fence vibration amplitude, and sound waves from severe collisions or explosions, and communicate with the edge gateway via the LoRaWAN wireless protocol. The smart cameras collect video footage of the road surface, including vehicle collisions, rollovers, and other emergencies, and communicate with the edge gateway via Ethernet or a 4G wireless network. The edge gateway processes and analyzes the collected data, identifies the data, and determines whether it is an emergency. If an emergency is identified, it immediately activates the smart road studs to issue a light warning.

[0050] A control method for a traffic early warning system, implemented using the aforementioned traffic early warning system, includes the following steps:

[0051] Step 1: Deploy smart road studs, cameras, and edge gateways on the road.

[0052] Smart road studs are deployed at intervals along the road, for example, two rows of smart road studs spaced 10 or 20 meters apart on a one-way lane. An edge gateway is installed on the roadside to ensure that the smart road studs within a certain range can connect to the gateway normally. Cameras are installed at intervals (e.g., 50 or 100 meters) according to the relative driving direction of vehicles to ensure that clear images of vehicles can be obtained.

[0053] Step 2: The smart road stud collects real-time data on road surface tilt angle, road fence vibration, and vehicle collision sounds. Simultaneously, a camera captures real-time video of the road surface. Both the smart road stud and the camera transmit the data to the edge gateway. Preferably, the smart road stud transmits data to the edge gateway wirelessly via LoRaWAN, while the camera transmits data to the edge gateway via wired Ethernet or a 4G wireless network. The smart road stud transmits the collected sensor data (vibration, sound, tilt angle) to the edge gateway via the LoRaWAN wireless protocol, leveraging LoRaWAN's long-range and low-power characteristics to ensure fast and stable data transmission without delay or loss. The camera continuously transmits high-definition video streams to the edge gateway via a pre-set wired or wireless communication link, ensuring the real-time nature and integrity of the video data to meet the needs of subsequent decoding and recognition.

[0054] Step 3: The edge gateway identifies and analyzes the received data, and determines whether a landslide or traffic accident has occurred based on the analysis results. The edge gateway uses algorithms to analyze the data and verifies the data collected by the smart road studs, confirming the data from multiple dimensions to improve the accuracy of accident determination. Step 3 specifically includes:

[0055] Step 3.1: After receiving the sensor data from the smart road beacon, the edge gateway compares it with a preset threshold. If the data does not exceed the threshold, it is considered normal (e.g., the collision sound does not reach 100dB), and the system continues to wait for the next round of data. If the data exceeds the threshold, it is marked as a suspected accident and enters the video review stage. The results are compared with the camera image recognition results for multi-dimensional judgment, reducing the false alarm rate.

[0056] Step 3.2, as shown in Figure 3, involves the edge gateway decoding the video stream transmitted from the camera, extracting frames at 1-second intervals, and inputting the extracted frames into a YOLO instance segmentation small model to initially identify vehicle stops, rollovers, collisions, and other abnormal events. If the small model identifies an anomaly, it is then re-examined using a larger model (combining video context and image details) to confirm the authenticity of the abnormal event. After final confirmation of the incident, the abnormal event is labeled, encoded, and used to generate alarm images or short videos for storage and push to the management platform. If the verification fails to confirm the incident, it is determined to be a false alarm and no warning is triggered.

[0057] Step 4: If the edge gateway determines that a landslide or traffic accident has occurred, it sends instructions to the smart road studs around the accident section.

[0058] Step 5: After receiving the command from the edge gateway, the control module of the smart road stud starts to flash the LED light module rapidly to remind drivers of vehicles behind to pay attention to the road conditions.

[0059] Step 6: After seeing the LED light module flashing rapidly as a warning, the driver of the vehicle behind will increase their attention and slow down to observe the road conditions, ultimately reducing the probability of a secondary accident.

[0060] Step 7: After the incident is handled, staff access the edge gateway or mobile app via the web and send a recovery signal to all LED light modules in the warning state. After receiving the signal, the control module of the smart road stud controls the LED light modules to stop flashing rapidly.

[0061] The beneficial effects of the present invention will be illustrated below through specific examples:

[0062] Example 1:

[0063] This early warning system consists of multiple smart road studs and several gateways. It is deployed in landslide-prone sections (such as mountainous sections of highways and slope sections of national highways). Deploying them in two rows yields the best results. The deployment interval is adjusted according to the risk level of the road section (one stud every 1 meter in high-risk sections, and one stud every 10 meters in low-risk sections). The system's workflow is as follows:

[0064] In normal low-power mode, when there is no landslide, each smart road stud enters a low-power mode under the control of the control module. The tilt sensor module intermittently collects data (e.g., once every 30 seconds) and transmits the data back to the edge gateway for judgment. If the data is normal (not exceeding the threshold), low power consumption continues. The LoRaWAN module is in standby mode, maintaining only basic receiving functions. The LED light module is off. At the same time, the solar photovoltaic panel continuously charges the battery module to ensure sufficient battery power.

[0065] Edge gateway: Can be installed in a safe area near high-risk road sections, powered by mains electricity, maintaining uninterrupted power supply 24 / 7. Continuously identifies and analyzes acquired road data.

[0066] Landslide data collection: When a landslide or collapse occurs on a section of road where a smart road stud is located, the ground tilt angle changes drastically. The tilt angle sensing module will collect the data and send it back to the edge gateway for judgment and analysis.

[0067] Wide-area synchronous early warning: The edge gateway compares the data acquired from the smart road studs. If the data exceeds the threshold, it is judged as a road collapse accident. The edge gateway immediately issues an instruction to link other smart road studs within a 2-kilometer radius through the LoRaWAN wireless module on the gateway. After receiving the accident signal, the LoRaWAN modules of the surrounding road studs transmit the signal to their respective control modules. The control modules determine that it is a "surrounding accident" and then control their own LED light modules to start flashing rapidly. In this way, all road studs within a 2-kilometer radius flash rapidly in unison, forming a "wide-area warning zone" to ensure that following vehicles receive a warning before approaching the accident point, have enough reaction time, and increase their attention to slow down and avoid the accident.

[0068] System recovery: Once the incident has been resolved, staff can send a "recovery signal" to the smart road stud via web access to the edge gateway or APP. After receiving the signal, the control module will stop the LED light module from flashing rapidly, and the smart road stud will return to its normal low-power state.

[0069] Example 2:

[0070] This early warning system consists of multiple high-definition cameras, multiple smart road spikes, and several gateways. It is deployed in accident-prone sections (such as lane merging or diverging points, sharp turns, or consecutive sharp bends) where rear-end collisions and other accidents are common (e.g., at lane merging or diverging points, sharp turns, or consecutive sharp bends). The smart road spikes with integrated microphone modules are primarily placed on road fencing (e.g., one every 10 meters). They are then combined with ordinary smart road spikes placed along the road divider in the middle of the road. For optimal effect, a deployment of two or more rows is recommended. Based on actual performance, a placement of one every 10 meters is suggested. The system workflow is as follows:

[0071] In normal low-power mode, when there is no landslide, each smart road stud enters a low-power mode under the control of the control module. The microphone module intermittently collects data (e.g., 30 seconds of data collection followed by 30 seconds of sleep), with multiple smart road studs collecting data at staggered intervals to complement each other and avoid missing any monitoring. If the data is normal (not exceeding the threshold), low power consumption continues; the LoRaWAN module is in standby mode, maintaining only basic reception functions; the LED light module is off. Simultaneously, the solar photovoltaic panel continuously charges the battery module to ensure sufficient battery power.

[0072] High-definition cameras: The cameras are installed on the roadside and are powered by either mains electricity or solar power. They monitor accident-prone road sections and collect video images 24 / 7. The collected video is then transmitted back to the edge gateway for identification.

[0073] Edge gateway: Can be installed in a safe area near accident-prone sections of road, powered by mains electricity, maintaining uninterrupted power supply 24 / 7. Continuously identifies and analyzes acquired road data.

[0074] Collision Detection: When a collision occurs on a section of road where a smart road beacon is located, a high-definition camera transmits real-time video of the collision back to the edge gateway for identification. A small model identifies the collision as a vehicle collision, which is then verified by a larger model. Simultaneously, if the sound decibel value collected by the microphone module exceeds a preset threshold by the control module, the edge gateway will make a comprehensive judgment based on multiple dimensions to confirm a collision.

[0075] Wide-area synchronous early warning: After analyzing and determining that a collision has occurred, the edge gateway immediately issues a command to coordinate with other smart road studs within a 2-kilometer radius via its LoRaWAN wireless module. Upon receiving the accident signal, the LoRaWAN modules of the surrounding road studs transmit the signal to their respective control modules. These control modules, recognizing the incident as a "surrounding accident," immediately activate their own LED light modules to flash rapidly. This process continues, with all road studs within a 2-kilometer radius flashing synchronously, forming a "wide-area warning zone." This ensures that following vehicles receive a warning before approaching the accident site, have sufficient reaction time, and increase their attention to slow down and avoid the collision.

[0076] System recovery: Once the incident has been resolved, staff can send a "recovery signal" to the smart road stud via web access to the edge gateway or APP. After receiving the signal, the control module will stop the LED light module from flashing rapidly, and the smart road stud will return to its normal low-power state.

[0077] 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 within the protection scope of the present invention.

Claims

1. A traffic early warning system, characterized in that, include: Smart road studs are installed at preset intervals on roads or guardrails. They collect real-time data on road surface inclination, guardrail vibration, and vehicle collision sound waves, transmitting this data to an edge gateway. Simultaneously, they receive commands from the edge gateway and generate warnings. Cameras collect road surface video and transmit it to the edge gateway. The edge gateway receives data from both the smart road studs and the cameras. It also performs collaborative analysis of the dual-source data from the smart road studs and cameras, accurately determining whether a landslide or traffic accident has occurred based on the analysis results. When a landslide or traffic accident is detected, it issues a warning command to the smart road studs.

2. The traffic early warning system according to claim 1, characterized in that, The smart road stud includes: a sensor module that collects real-time information data on road surface tilt angle, guardrail vibration, and vehicle collision sound waves; a control module electrically connected to a LoRaWAN communication module, which receives warning commands from the edge gateway via the LoRaWAN communication module and generates a warning control signal; an LED light module electrically connected to the control module, which performs a fast-flash action when it receives a warning control signal output by the control module; a LoRaWAN communication module electrically connected to the control module, which receives data and commands output by the control module, sends the data to the edge gateway, and listens for warning commands issued by the edge gateway, transmitting the commands to the control module; and a power supply module that converts solar energy into electrical energy and provides power to the sensor module, the control module, the LED light module, and the LoRaWAN communication module.

3. The traffic early warning system according to claim 2, characterized in that, The sensor module includes a tilt sensor, a vibration sensor, and a microphone. The tilt sensor, the vibration sensor, and the microphone respectively collect information data on road tilt angle, fence vibration, and vehicle collision sound waves in real time.

4. The traffic early warning system according to claim 2, characterized in that, The power supply module includes a solar photovoltaic panel and a battery; the solar photovoltaic panel converts solar energy into electrical energy and charges the battery, and the battery provides power to the sensor module, the control module, the LED light module and the LoRaWAN communication module respectively.

5. A control method for a traffic early warning system, implemented using the traffic early warning system according to any one of claims 1-4, characterized in that, The process includes the following steps: Step 1, deploying smart road studs, cameras, and edge gateways on the road; Step 2, the smart road studs collect real-time data on road surface tilt, road fence vibration, and vehicle collision sounds, while the cameras collect real-time road video, and the data collected by the smart road studs and cameras is transmitted to the edge gateway; Step 3, the edge gateway identifies and analyzes the received data, and determines whether a landslide or traffic accident has occurred based on the analysis results; Step 4, if the edge gateway determines that a landslide or traffic accident has occurred, it issues a command to the smart road studs around the accident site; Step 5, after receiving the command from the edge gateway, the control module of the smart road studs controls the LED light modules to start flashing rapidly, reminding drivers of following vehicles to pay attention to the road conditions; Step 6, after seeing the rapidly flashing LED light modules as a warning, drivers of following vehicles increase their attention and slow down to observe the road conditions, ultimately reducing the probability of secondary accidents; Step 7, after the accident is handled, staff access the edge gateway via a web interface or a mobile APP to send a recovery signal to all LED light modules in the warning state, and the control module of the smart road studs receives the signal and controls the LED light modules to stop flashing rapidly.

6. The control method for the traffic early warning system according to claim 5, characterized in that, The smart road studs transmit data wirelessly to the edge gateway via LoRaWAN, while the cameras transmit data to the edge gateway via wired Ethernet or 4G wireless network.

7. The control method for the traffic early warning system according to claim 5, characterized in that, Step 3 specifically includes: Step 3.1: After receiving the sensor data from the smart road beacon, the edge gateway compares it with a preset threshold. If the threshold is not exceeded, it is considered normal, and the system continues to wait for the next round of data; if the threshold is exceeded, it is marked as a suspected accident and enters the video review stage; Step 3.2: The edge gateway decodes the video stream transmitted by the camera, extracts frames at 1-second intervals, and inputs the extracted frame images into a YOLO instance segmentation small model to initially identify vehicle stops, rollovers, collisions, and other abnormal events; if the small model identifies an anomaly, it is then reviewed through a large model to confirm the authenticity of the abnormal event; after the accident is finally confirmed, the abnormal event is marked, encoded to generate an alarm image or short video for storage, and pushed to the management platform; if the review fails to confirm, it is determined to be a false alarm and no warning is triggered.

Images