Multi-section cooperative early warning method and platform based on multi-source traffic meteorological data

Abstract

The application relates to a multi-section cooperative early warning method and platform based on multi-source traffic meteorological data, and belongs to the technical field of intelligent traffic. The platform comprises a multi-source data acquisition system, a traffic meteorological monitoring and early warning decision system and a variable information board joint control system; the multi-source data acquisition system comprises a data quality control module, a traffic meteorological monitoring module and a traffic monitoring video AI identification module; the traffic meteorological monitoring and early warning decision system comprises a data receiving module, a data visualization module, a decision support module and a prevention and control equipment linkage module. The application can realize multi-point, cooperative, dynamic, accurate and stable and reliable active early warning under severe weather, so as to improve the driving safety of the expressway.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation technology, specifically relating to a multi-segment collaborative early warning method and platform based on multi-source traffic and meteorological data. Background Technology

[0002] Currently, highway safety systems have deployed various traffic and weather monitoring modules and variable message signs, among other information dissemination equipment. However, existing technologies generally suffer from the following problems:

[0003] (1) Lack of effective coordination: The existing traffic meteorological monitoring system and the existing roadside information dissemination and active prevention and control equipment system are usually independent of each other. Meteorological monitoring data cannot automatically and in real time trigger the front-end information dissemination equipment to issue early warnings. Most of the time, it depends on the manual operation of the monitoring center personnel, which results in a delayed response and cannot play a timely early warning role in the early stage of disasters.

[0004] (2) Inaccurate warning information: Traditional warning methods are usually "point-to-point", that is, if an anomaly is detected at a certain monitoring point, at most a nearby information board will be triggered to display a fixed and general warning, such as "Dense fog ahead, drive with caution". This information does not specify the exact distance of the disaster point and has limited reference value for drivers.

[0005] (3) Insufficient early warning coverage: Regional coordinated early warning has not been formed. When a meteorological disaster occurs at a location, it is impossible to coordinate the control of multiple information boards at upstream locations (in both directions of traffic), which prevents drivers at remote locations from knowing the road conditions ahead in advance and causes them to miss the best opportunity to slow down and change lanes in advance.

[0006] (4) Uncontrollable data quality: Some existing systems directly use data from front-end monitoring devices, but lack effective data quality control. Errors or abnormal monitoring data may lead to false alarms or missed alarms, affecting the reliability of the system.

[0007] (5) Simple warning logic and poor stability: Most existing automatic warning logics are based on simple threshold judgments, that is, a warning is triggered once the monitored value exceeds the threshold, and it is canceled when it is below the threshold. This design will cause the warning information to appear and disappear frequently on the information board (i.e., "flickering") in critical states, which will confuse the driver and reduce the authority and reliability of the system.

[0008] CN202110215150.4 discloses a "Adaptive Variable Speed ​​Limit and Information Collaborative Control System for Highways in Adverse Weather," which includes a Variable Message System (VMS), a Variable Speed ​​Limit Sign (VLS), a Traffic Weather Monitoring Station (RWS), a Traffic Flow Detection Device (TRA), and a collaborative controller. The VMS is used to publish the current severe weather type, location, and control measures; the VLS is used to publish the real-time speed limit value of the current speed limit zone; the TRA is used to collect traffic flow information; and the RWS is used to collect meteorological parameters of the road segment in real time. The disadvantages of this technology are: First, its core purpose is the collaborative adjustment of variable speed limits, and it cannot perform collaborative dissemination of early warning information. Second, its data sources are limited. Third, it cannot effectively solve the problem of frequent on / off (i.e., strobeing) of early warning information under critical conditions, and its simple threshold triggering logic has low reliability in practical applications.

[0009] CN202010386291.8 discloses a "Highway Meteorological Early Warning System and Method," which includes a roadside system and an early warning cloud platform that are communicatively connected to a meteorological monitoring system. The roadside system is communicatively connected to the early warning cloud platform and also to all vehicles within a preset communication range. The early warning cloud platform receives meteorological data for a preset area from the meteorological monitoring system, analyzes the data to obtain weather condition data for each section of the highway within a preset time period, and then sends the weather condition data to the roadside system. The roadside system generates early warning information based on the weather condition data and sends it to all vehicles within the preset communication range to remind drivers to pay attention to driving safety. By directly sending early warning information to vehicles, the problem of inadequate early warning information delivery during severe weather is avoided, thereby improving driving safety on highways during severe weather. The disadvantages of this technology are: firstly, its early warning strategy is relatively simple; when a disaster occurs, it cannot coordinate and control multiple upstream information boards for a coordinated regional early warning strategy. Secondly, the content of the issued early warning information is fixed and general, failing to provide dynamic and personalized information display for each information board. Finally, this technology also failed to solve the problem of frequent switching on and off (i.e., flickering) of warning information in critical states.

[0010] On October 6, 2018, the publication "A Severe Weather Early Warning Management System and Method Based on Four-Dimensional Real-Scene Traffic Simulation" was released. This system includes a multi-element, all-around tracking and detection radar sensor, a data acquisition and analysis processor, a cloud processing server, and a four-dimensional simulation management workstation. The radar tracks vehicles and pedestrians on the road and acquires their location information, simultaneously triggering license plate capture cameras to capture images of vehicles in real time. The system acquires node weather information for the current road segment, performs four-dimensional simulation of the traffic environment, and activates early warning rules. By acquiring the location information of vehicles and pedestrians, it determines the road segment that vehicles and pedestrians will be passing through, and sends alarm information to vehicles and pedestrians about to pass through that segment via point-to-point or broadcast methods. Based on the four-dimensional real-scene traffic simulation, it provides road managers with information for developing emergency plans, road control, and accident vehicle and pedestrian rescue, pushing information to plan alternative routes and reduce the risk of high-speed travel under severe weather conditions. The drawback of this technology is that its core technology lies in constructing a complex four-dimensional real-scene traffic simulation system, which cannot implement multi-point collaborative information dissemination strategies on upstream two-way roads for specific disaster points, nor can it generate early warning information containing dynamic and precise distances. More importantly, this technology also lacks a stable release mechanism based on configurable trigger and revocation rules designed to ensure the stability of early warning releases.

[0011] Therefore, how to achieve deep integration between traffic meteorological monitoring and front-end prevention and control equipment, and provide accurate, timely, coordinated and stable regional proactive early warning, is a technical problem that urgently needs to be solved. Summary of the Invention

[0012] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a multi-segment collaborative early warning method and platform based on multi-source traffic meteorological data.

[0013] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0014] A multi-segment collaborative early warning platform based on multi-source traffic and meteorological data includes a multi-source data acquisition system, a traffic and meteorological monitoring and early warning decision system, and a variable message sign control system.

[0015] The multi-source data acquisition system includes a data quality control module, a traffic and meteorological monitoring module, and a traffic surveillance video AI recognition module;

[0016] The data quality control module is connected to the weather sensor, roadside monitoring camera, traffic meteorological monitoring module, and traffic monitoring video AI recognition module, respectively. It is used to perform quality inspection on the data collected by the weather sensor and the video image data collected by the roadside monitoring camera. After removing invalid data through preset quality control rules (range validity, change stability, multi-source consistency, data collection continuity), the valid data collected by the weather sensor is transmitted to the traffic meteorological monitoring module, and the valid video image data collected by the roadside monitoring camera is transmitted to the traffic monitoring video AI recognition module.

[0017] The traffic weather monitoring module is used to receive data collected by weather sensors that have passed quality inspection, and to determine whether severe weather conditions have occurred by comparing the collected data with preset severe weather judgment thresholds.

[0018] The traffic monitoring video AI recognition module is equipped with an AI model that can identify severe weather. The traffic monitoring video AI recognition module is used to receive video images from roadside monitoring cameras that have passed quality inspection. Through the AI ​​model that can identify severe weather, the video images are analyzed and the recognition results of severe weather conditions are directly output.

[0019] The traffic meteorological monitoring, early warning and decision-making system includes a data receiving module, a data visualization module, a decision support module, and a prevention and control equipment linkage module;

[0020] The data receiving module is connected to the traffic meteorological monitoring module and the traffic monitoring video AI recognition module respectively, and is used to receive severe weather status data identified by the traffic meteorological monitoring module and the traffic monitoring video AI recognition module.

[0021] The data visualization module is connected to the data receiving module and is used to visualize the received severe weather data.

[0022] The decision support module, connected to the data receiving module, is used to determine the status of each type of meteorological disaster at each monitoring location based on its built-in early warning triggering rules and early warning cancellation rules, and according to the received severe weather status data, and to give the corresponding early warning triggering or cancellation instructions. Then, based on the multi-segment variable information collaborative control strategy, it determines the variable information board to receive the early warning triggering or cancellation instructions and the corresponding display content.

[0023] The prevention and control equipment linkage module is connected to the decision support module and is used to send the early warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display.

[0024] Furthermore, preferably, the weather sensors include a visibility meter, a rain gauge, and a road condition sensor.

[0025] Furthermore, preferably, severe weather conditions include dense fog, blizzards, heavy rain, icy roads, strong winds, high road temperatures, sandstorms, and thunderstorms.

[0026] Furthermore, preferably, the warning triggering rule of the decision support module is as follows: when the system detects that a certain location meets the disaster triggering conditions N times consecutively, and the time elapsed since the last successful triggering of a similar warning at that location exceeds a duration T, then the warning is deemed valid and a warning triggering command is generated.

[0027] Furthermore, preferably, the warning cancellation rule is as follows: when a certain location is detected to have failed to meet the disaster triggering conditions N times consecutively, it is determined that the disaster has ended, and a warning cancellation command is immediately generated.

[0028] Furthermore, preferably, the multi-segment variable information cooperative control strategy is as follows:

[0029] a. Determine the location of the disaster point: Based on the source of the data corresponding to the early warning triggering instruction, locate the mileage marker of the meteorological disaster occurrence point;

[0030] b. Filter target information boards: With the disaster point as the center, automatically search and filter all variable information boards within a preset distance range upstream of the two-way traffic direction;

[0031] c. Generate dynamic warning or cancellation information: If a warning is triggered, the distance to the disaster point is automatically calculated for each selected variable message sign based on its mileage marker, and a warning message including the distance is generated; if a warning is cancelled, an instruction to clear the corresponding warning is generated.

[0032] Furthermore, preferably, the variable message sign control system includes a data receiving and parsing module, a data display module, and a control module;

[0033] The data receiving and parsing module receives instructions and corresponding display content from the prevention and control equipment linkage module, and then sends the parsed content to the control module. The control module controls the data display module to display the corresponding content based on the parsed content.

[0034] This invention also provides a multi-segment collaborative early warning method based on multi-source traffic and meteorological data, employing the aforementioned multi-segment collaborative early warning platform based on multi-source traffic and meteorological data, and includes the following steps:

[0035] Step (1), Raw data acquisition and quality control: Acquire raw data from weather sensors and / or roadside monitoring cameras, and first perform quality checks on the raw data to remove invalid data that exceeds the normal range or contains logical errors;

[0036] Step (2), severe weather condition identification: The valid sensor data that has passed quality inspection is sent to the traffic meteorological monitoring module for threshold comparison to identify severe weather conditions; the valid video image data is sent to the traffic monitoring video AI recognition module for analysis by the AI ​​model to output the severe weather condition.

[0037] Step (3), warning logic judgment: Based on the built-in warning triggering rules and warning cancellation rules, and according to the identified severe weather status data, the status judgment of each type of meteorological disaster at each monitoring location is performed, and corresponding warning triggering or cancellation instructions are given;

[0038] Step (4), Cooperative control strategy execution: Determine the variable information board to receive the warning trigger or cancellation command and the corresponding display content according to the multi-segment variable information cooperative control strategy;

[0039] Step (5), Information Linkage Release: Send the warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display.

[0040] Furthermore, preferably, the generation of the early warning triggering instruction in step (3) specifically includes: when a certain location is detected to meet the disaster triggering conditions N times consecutively, and the current time is more than the time T since the last successful triggering of a similar early warning at that location, an early warning triggering instruction is generated.

[0041] This invention solves the problems of lack of linkage between traffic meteorological monitoring and information release systems, inaccurate early warning information, limited early warning range, inability to guarantee data quality, and information flickering caused by simple early warning logic. This invention can realize multi-point, coordinated, dynamic, accurate and stable proactive early warning under severe weather conditions, thereby improving the driving safety of highways.

[0042] In this invention, multiple information boards within a preset range upstream (for vehicles traveling in both directions) are automatically identified and controlled to provide synchronous early warnings, with the disaster site as the center.

[0043] This invention dynamically calculates the distance between each controlled information board and the disaster point, and generates personalized early warning information containing that precise distance.

[0044] In this invention, triggering an early warning requires simultaneously meeting two conditions: "a threshold for the number of times the trigger condition is met consecutively" and "a threshold for the time elapsed since the last successful triggering moment." However, canceling an early warning only requires meeting one condition: "a threshold for the number of times the trigger condition is not met consecutively." Moreover, these thresholds can be independently configured and modified for different monitoring locations and disaster types.

[0045] The traffic monitoring video AI recognition module includes a pre-trained AI model capable of recognizing severe weather. This invention does not limit the specific AI model; any AI model capable of recognizing severe weather is acceptable.

[0046] In a multi-source data acquisition system, the data quality control module, as the first step in data processing, performs quality checks on the raw data collected from front-end weather sensors and surveillance cameras. Invalid data is determined using one or more of the following methods:

[0047] (1) Validity control of the range: Determine whether the original data value is within the preset valid physical range (for example, the visibility value cannot be negative, and the road surface temperature should be between -40℃ and 80℃).

[0048] (2) Change stability control: Determine whether the rate of change of the original data in a short period of time exceeds the preset threshold in order to eliminate instantaneous jump data caused by signal interference.

[0049] (3) Multi-source consistency quality control: When there are more than one source of data for the same monitoring point (such as sensor data and AI recognition results at the same time), the results from different sources are compared; if there is an obvious contradiction in the results (for example, the sensor reports that the road surface is icy while the AI ​​recognizes that the road surface is dry), the data set is determined to be abnormal and an alarm can be triggered to prompt manual intervention.

[0050] (4) Data acquisition continuous quality control: Determine whether the value of a certain data source remains unchanged for a period of time that is far longer than normal, so as to identify faults such as equipment crashes or communication interruptions.

[0051] This invention first obtains severe weather conditions (including dense fog, blizzard, heavy rain, road icing, strong winds, high road surface temperature, sandstorm, and thunderstorm) on highways through two or more methods.

[0052] One method is to acquire real-time sensor data from traditional meteorological monitoring equipment (such as visibility meters and road surface sensors) deployed along the route.

[0053] Secondly, an AI visual recognition model that does not impose restrictions on specific algorithms is used to analyze and identify weather conditions (dense fog, blizzard, heavy rain, icy roads, strong winds, high road temperatures, sandstorms, and thunderstorms) in real-time video images from existing roadside surveillance cameras.

[0054] When a meteorological disaster is confirmed at a precise mileage marker location through one or two fusion-verified data sources, a stable release strategy incorporating trigger sensitivity and trigger interval thresholds is used to determine the need for an early warning. Once a warning is deemed necessary, the core "multi-segment variable message sign (VMS) collaborative control strategy" is automatically executed. This strategy is not a traditional "point-to-point" warning; instead, it automatically selects all VMS signs within a preset distance range in both directions upstream of the disaster site, centered on the disaster site. Then, for each selected VMS sign, the precise distance between it and the disaster site is independently and dynamically calculated, generating highly personalized warning information containing the specific distance (e.g., generating a message like "Dense fog 8 kilometers ahead, please slow down" for a VMS sign 8 kilometers from the disaster site). Finally, these different warning messages are synchronously and collaboratively sent and displayed on their respective VMS signs, forming a large-scale, accurate, and timely collaborative warning network upstream of the disaster area, effectively overcoming the shortcomings of existing technologies such as delayed warnings, vague information, and insufficient coverage.

[0055] Compared with the prior art, the beneficial effects of this invention are as follows:

[0056] (1) The early warning is stable and reliable, avoiding information flickering: By introducing an early warning triggering mechanism based on the dual conditions of "trigger sensitivity" and "trigger interval", as well as an cancellation mechanism based on "trigger sensitivity", the instantaneous fluctuation of data is effectively filtered out, avoiding the problem of early warning information repeatedly appearing and disappearing on the information board in the critical state, which greatly improves the stability of the early warning and the driver's trust in the system.

[0057] (2) Abundant monitoring methods and wide coverage: Using the roadside surveillance cameras that have been deployed in large numbers as the data source for meteorological monitoring is cost-effective and greatly expands the coverage and density of monitoring.

[0058] (3) Higher reliability: By integrating two different data sources, sensors and AI visual recognition, cross-validation can be performed, which can effectively improve the accuracy of severe weather identification and reduce the risk of false alarms and missed alarms.

[0059] (4) Timely response and high degree of automation: It realizes end-to-end automated closed-loop control from monitoring and decision-making to release, which greatly shortens the early warning response time and can provide second-level early warning in the early stage of severe weather events;

[0060] (5) Accurate and dynamic early warning information: The issued early warning information includes the type of disaster and the precise distance to the disaster site. The information content is dynamically generated, providing drivers with clearer and more valuable driving guidance;

[0061] (6) Collaborative and coordinated, with a wide range of early warning: Through the “multi-segment variable information collaborative control strategy”, the impact range of a single disaster point is expanded, and multiple information boards in the upstream two-way traffic direction are synchronously and collaboratively warned, realizing “regionalized” prevention and control.

[0062] (7) Enhance safety: Through timely, accurate, wide-ranging and highly reliable proactive early warning, the traffic accident rate caused by severe weather can be effectively reduced, and the traffic safety level of highways can be significantly improved.

[0063] (8) Compared with traditional manual intervention or simple single-threshold automatic early warning methods, the present invention has made significant progress in the following aspects:

[0064] In terms of response efficiency: The automated closed-loop control of this invention reduces the average time for issuing early warning information from 10-15 minutes required by traditional manual operation to less than 1 minute, improving efficiency by more than 90%.

[0065] Regarding stability of information dissemination: Under simulated critical weather conditions (e.g., visibility fluctuating around a threshold), a simple single threshold judgment logic may cause the information board information to "flicker" 5-10 times within 30 minutes, while the stable triggering and cancellation mechanism of this invention can reduce the number of "flickering" times to 0, avoiding causing inconvenience to drivers.

[0066] Regarding the accuracy and safety of early warnings: Through cross-validation of multi-source data, the overall false alarm rate for severe weather can be reduced by more than 5%.

[0067] Meanwhile, compared to general information such as "fog ahead", the dynamic distance warning provided by this invention can make the driver's deceleration behavior smoother in simulation tests, reducing the incidence of sudden braking events near the warning point by about 15-20%, thereby effectively improving driving safety. Attached Figure Description

[0068] Figure 1 This is a schematic diagram of the multi-segment collaborative early warning platform based on multi-source traffic meteorological data of the present invention;

[0069] Figure 2 This is a schematic diagram of a multi-segment variable information collaborative control strategy;

[0070] Figure 3 Flowchart for the decision-making process of triggering and canceling early warnings;

[0071] Figure 4 This is a schematic diagram of the login interface for the traffic meteorological monitoring, early warning, and decision-making system.

[0072] Figure 5 This is a schematic diagram of the main interface of the variable message sign control system;

[0073] Figure 6 This is a schematic diagram of the information board warning information editing interface that includes dynamic distance parameters;

[0074] Figure 7 Screenshot of the early warning information transmission log to demonstrate the multi-segment coordinated control strategy;

[0075] Figure 8 These are on-site photos showing the actual application of the joint control system on roads. Detailed Implementation

[0076] The present invention will now be described in further detail with reference to the embodiments.

[0077] Those skilled in the art will understand that the following embodiments are for illustrative purposes only and should not be construed as limiting the scope of the invention. Where specific techniques or conditions are not specified in the embodiments, they are performed in accordance with the techniques or conditions described in the literature in the field or according to the product instructions. Materials or equipment whose manufacturers are not specified are all conventional products that can be obtained by purchase.

[0078] Example 1

[0079] like Figure 1 As shown, the multi-segment collaborative early warning platform based on multi-source traffic meteorological data includes a multi-source data acquisition system, a traffic meteorological monitoring and early warning decision system, and a variable message sign control system.

[0080] The multi-source data acquisition system includes a data quality control module, a traffic and meteorological monitoring module, and a traffic surveillance video AI recognition module;

[0081] The data quality control module is connected to the weather sensor, roadside monitoring camera, traffic meteorological monitoring module, and traffic monitoring video AI recognition module, respectively. It is used to perform quality inspection on the data collected by the weather sensor and the video image data collected by the roadside monitoring camera. After removing invalid data through preset quality control rules, the valid data collected by the weather sensor is transmitted to the traffic meteorological monitoring module, and the valid video image data collected by the roadside monitoring camera is transmitted to the traffic monitoring video AI recognition module.

[0082] The traffic weather monitoring module is used to receive data collected by weather sensors that have passed quality inspection, and to determine whether severe weather conditions have occurred by comparing the collected data with preset severe weather judgment thresholds.

[0083] The traffic monitoring video AI recognition module includes an AI model capable of identifying severe weather. This module receives video images from roadside monitoring cameras that have passed quality inspection. Using the AI ​​model, it analyzes the video images and directly outputs the identification results of the severe weather conditions. This invention does not limit the specific AI model used; for example, it can employ an image classification model based on a convolutional neural network (CNN), a vision model based on a Transformer, or other deep learning models suitable for image-based weather condition recognition. The model can be trained using publicly available or self-collected traffic monitoring image datasets containing different weather conditions (such as dense fog, heavy rain, and icy roads).

[0084] The traffic meteorological monitoring, early warning and decision-making system includes a data receiving module, a data visualization module, a decision support module, and a prevention and control equipment linkage module;

[0085] The data receiving module is connected to the traffic meteorological monitoring module and the traffic monitoring video AI recognition module respectively, and is used to receive severe weather status data identified by the traffic meteorological monitoring module and the traffic monitoring video AI recognition module.

[0086] The data visualization module is connected to the data receiving module and is used to visualize the received severe weather data.

[0087] The decision support module, connected to the data receiving module, is used to determine the status of each type of meteorological disaster at each monitoring location based on its built-in early warning triggering rules and early warning cancellation rules, and according to the received severe weather status data, and to give the corresponding early warning triggering or cancellation instructions. Then, based on the multi-segment variable information collaborative control strategy, it determines the variable information board to receive the early warning triggering or cancellation instructions and the corresponding display content.

[0088] The prevention and control equipment linkage module is connected to the decision support module and is used to send the early warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display.

[0089] The multi-segment collaborative early warning method based on multi-source traffic meteorological data, as described in this embodiment, uses a multi-segment collaborative early warning platform based on multi-source traffic meteorological data and includes the following steps:

[0090] Step (1), Raw data acquisition and quality control: Acquire raw data from weather sensors and / or roadside monitoring cameras, and first perform quality checks on the raw data to remove invalid data that exceeds the normal range or contains logical errors;

[0091] Step (2), severe weather condition identification: The valid sensor data that has passed quality inspection is sent to the traffic meteorological monitoring module for threshold comparison to identify severe weather conditions; the valid video image data is sent to the traffic monitoring video AI recognition module for analysis by the AI ​​model to output the severe weather condition.

[0092] Step (3), warning logic judgment: Based on the built-in warning triggering rules and warning cancellation rules, and according to the identified severe weather status data, the status judgment of each type of meteorological disaster at each monitoring location is performed, and corresponding warning triggering or cancellation instructions are given;

[0093] Step (4), Cooperative control strategy execution: Determine the variable information board to receive the warning trigger or cancellation command and the corresponding display content according to the multi-segment variable information cooperative control strategy;

[0094] Step (5), Information Linkage Release: Send the warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display.

[0095] Example 2

[0096] like Figure 1 As shown, the multi-segment collaborative early warning platform based on multi-source traffic meteorological data includes a multi-source data acquisition system, a traffic meteorological monitoring and early warning decision system, and a variable message sign control system.

[0097] The multi-source data acquisition system includes a data quality control module, a traffic and meteorological monitoring module, and a traffic surveillance video AI recognition module;

[0098] The data quality control module is connected to the weather sensor, roadside monitoring camera, traffic meteorological monitoring module, and traffic monitoring video AI recognition module, respectively. It is used to perform quality inspection on the data collected by the weather sensor and the video image data collected by the roadside monitoring camera. After removing invalid data through preset quality control rules, the valid data collected by the weather sensor is transmitted to the traffic meteorological monitoring module, and the valid video image data collected by the roadside monitoring camera is transmitted to the traffic monitoring video AI recognition module.

[0099] The traffic weather monitoring module is used to receive data collected by weather sensors that have passed quality inspection, and to determine whether severe weather conditions have occurred by comparing the collected data with preset severe weather judgment thresholds.

[0100] The traffic monitoring video AI recognition module is equipped with an AI model that can identify severe weather. The traffic monitoring video AI recognition module is used to receive video images from roadside monitoring cameras that have passed quality inspection. Through the AI ​​model that can identify severe weather, the video images are analyzed and the recognition results of severe weather conditions are directly output.

[0101] The traffic meteorological monitoring, early warning and decision-making system includes a data receiving module, a data visualization module, a decision support module, and a prevention and control equipment linkage module;

[0102] The data receiving module is connected to the traffic meteorological monitoring module and the traffic monitoring video AI recognition module respectively, and is used to receive severe weather status data identified by the traffic meteorological monitoring module and the traffic monitoring video AI recognition module.

[0103] The data visualization module is connected to the data receiving module and is used to visualize the received severe weather data.

[0104] The decision support module, connected to the data receiving module, is used to determine the status of each type of meteorological disaster at each monitoring location based on its built-in early warning triggering rules and early warning cancellation rules, and according to the received severe weather status data, and to give the corresponding early warning triggering or cancellation instructions. Then, based on the multi-segment variable information collaborative control strategy, it determines the variable information board to receive the early warning triggering or cancellation instructions and the corresponding display content.

[0105] The prevention and control equipment linkage module is connected to the decision support module and is used to send the early warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display.

[0106] Weather sensors include visibility meters, rain gauges, and road condition sensors.

[0107] Severe weather conditions include dense fog, blizzards, heavy rain, icy roads, strong winds, high road temperatures, sandstorms, and thunderstorms.

[0108] The specific rules for triggering early warnings in the decision support module are as follows: when the system detects that a certain location meets the disaster triggering conditions N times consecutively, and the time elapsed since the last successful triggering of a similar early warning at that location exceeds a time interval T, then the early warning is deemed valid, and an early warning triggering command is generated.

[0109] The specific rules for canceling the warning are as follows: when a location is detected to have failed to meet the disaster triggering conditions N times consecutively, it is determined that the disaster has ended, and a warning cancellation command is immediately generated.

[0110] The multi-segment variable information cooperative control strategy is as follows:

[0111] a. Determine the location of the disaster point: Based on the source of the data corresponding to the early warning triggering instruction, locate the mileage marker of the meteorological disaster occurrence point;

[0112] b. Filter target information boards: With the disaster point as the center, automatically search and filter all variable information boards within a preset distance range upstream of the two-way traffic direction;

[0113] c. Generate dynamic warning or cancellation information: If a warning is triggered, the distance to the disaster point is automatically calculated for each selected variable message sign based on its mileage marker, and a warning message including the distance is generated; if a warning is cancelled, an instruction to clear the corresponding warning is generated.

[0114] The variable message sign control system includes a data receiving and parsing module, a data display module, and a control module;

[0115] The data receiving and parsing module receives instructions and corresponding display content from the prevention and control equipment linkage module, and then sends the parsed content to the control module. The control module controls the data display module to display the corresponding content based on the parsed content.

[0116] The multi-segment collaborative early warning method based on multi-source traffic meteorological data, as described in this embodiment, uses a multi-segment collaborative early warning platform based on multi-source traffic meteorological data and includes the following steps:

[0117] Step (1), Raw data acquisition and quality control: Acquire raw data from weather sensors and / or roadside monitoring cameras, and first perform quality checks on the raw data to remove invalid data that exceeds the normal range or contains logical errors;

[0118] Step (2), severe weather condition identification: The valid sensor data that has passed quality inspection is sent to the traffic meteorological monitoring module for threshold comparison to identify severe weather conditions; the valid video image data is sent to the traffic monitoring video AI recognition module for analysis by the AI ​​model to output the severe weather condition.

[0119] Step (3), warning logic judgment: Based on the built-in warning triggering rules and warning cancellation rules, and according to the identified severe weather status data, the status judgment of each type of meteorological disaster at each monitoring location is performed, and corresponding warning triggering or cancellation instructions are given;

[0120] Step (4), Cooperative control strategy execution: Determine the variable information board to receive the warning trigger or cancellation command and the corresponding display content according to the multi-segment variable information cooperative control strategy;

[0121] Step (5), Information Linkage Release: Send the warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display.

[0122] Example 3

[0123] like Figure 1 As shown, the multi-segment collaborative early warning platform based on multi-source traffic meteorological data includes a multi-source data acquisition system, a traffic meteorological monitoring and early warning decision system, and a variable message sign control system.

[0124] The multi-source data acquisition system includes a data quality control module, a traffic and meteorological monitoring module, and a traffic surveillance video AI recognition module;

[0125] The data quality control module is connected to the weather sensor, roadside monitoring camera, traffic meteorological monitoring module, and traffic monitoring video AI recognition module, respectively. It is used to perform quality inspection on the data collected by the weather sensor and the video image data collected by the roadside monitoring camera. After removing invalid data through preset quality control rules, the valid data collected by the weather sensor is transmitted to the traffic meteorological monitoring module, and the valid video image data collected by the roadside monitoring camera is transmitted to the traffic monitoring video AI recognition module.

[0126] The traffic weather monitoring module is used to receive data collected by weather sensors that have passed quality inspection, and to determine whether severe weather conditions have occurred by comparing the collected data with preset severe weather judgment thresholds.

[0127] The traffic monitoring video AI recognition module is equipped with an AI model that can identify severe weather. The traffic monitoring video AI recognition module is used to receive video images from roadside monitoring cameras that have passed quality inspection. Through the AI ​​model that can identify severe weather, the video images are analyzed and the recognition results of severe weather conditions are directly output.

[0128] The traffic meteorological monitoring, early warning and decision-making system includes a data receiving module, a data visualization module, a decision support module, and a prevention and control equipment linkage module;

[0129] The data receiving module is connected to the traffic meteorological monitoring module and the traffic monitoring video AI recognition module respectively, and is used to receive severe weather status data identified by the traffic meteorological monitoring module and the traffic monitoring video AI recognition module.

[0130] The data visualization module is connected to the data receiving module and is used to visualize the received severe weather data.

[0131] The decision support module, connected to the data receiving module, is used to determine the status of each type of meteorological disaster at each monitoring location based on its built-in early warning triggering rules and early warning cancellation rules, and according to the received severe weather status data, and to give the corresponding early warning triggering or cancellation instructions. Then, based on the multi-segment variable information collaborative control strategy, it determines the variable information board to receive the early warning triggering or cancellation instructions and the corresponding display content.

[0132] The prevention and control equipment linkage module is connected to the decision support module and is used to send the early warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display.

[0133] Weather sensors include visibility meters, rain gauges, and road condition sensors.

[0134] Severe weather conditions include dense fog, blizzards, heavy rain, icy roads, strong winds, high road temperatures, sandstorms, and thunderstorms.

[0135] The specific rules for triggering early warnings in the decision support module are as follows: when the system detects that a certain location meets the disaster triggering conditions N times consecutively, and the time elapsed since the last successful triggering of a similar early warning at that location exceeds a time interval T, then the early warning is deemed valid, and an early warning triggering command is generated.

[0136] The specific rules for canceling the warning are as follows: when a location is detected to have failed to meet the disaster triggering conditions N times consecutively, it is determined that the disaster has ended, and a warning cancellation command is immediately generated.

[0137] The multi-segment variable information cooperative control strategy is as follows:

[0138] a. Determine the location of the disaster point: Based on the source of the data corresponding to the early warning triggering instruction, locate the mileage marker of the meteorological disaster occurrence point;

[0139] b. Filter target information boards: With the disaster point as the center, automatically search and filter all variable information boards within a preset distance range upstream of the two-way traffic direction;

[0140] c. Generate dynamic warning or cancellation information: If a warning is triggered, the distance to the disaster point is automatically calculated for each selected variable message sign based on its mileage marker, and a warning message including the distance is generated; if a warning is cancelled, an instruction to clear the corresponding warning is generated.

[0141] The variable message sign control system includes a data receiving and parsing module, a data display module, and a control module;

[0142] The data receiving and parsing module receives instructions and corresponding display content from the prevention and control equipment linkage module, and then sends the parsed content to the control module. The control module controls the data display module to display the corresponding content based on the parsed content.

[0143] The traffic meteorological monitoring, early warning, and decision-making system, as the core of the system, is deployed in the control center. This system is responsible for receiving and processing multi-source data, making early warning logic judgments and decisions based on configurable strategy tables, and generating and sending control commands.

[0144] Variable message sign control system: Deployed on the side of the highway, it is responsible for receiving and executing instructions from the central system to issue or clear warning information to passing vehicles.

[0145] These two systems are connected through data transmission and automatic control signals, forming a closed-loop intelligent interconnected control system.

[0146] The multi-segment collaborative early warning method based on multi-source traffic meteorological data, as described in this embodiment, uses a multi-segment collaborative early warning platform based on multi-source traffic meteorological data and includes the following steps:

[0147] Step (1), Raw data acquisition and quality control: Acquire raw data from weather sensors and / or roadside monitoring cameras, and first perform quality checks on the raw data to remove invalid data that exceeds the normal range or contains logical errors;

[0148] Step (2), severe weather condition identification: The valid sensor data that has passed quality inspection is sent to the traffic meteorological monitoring module for threshold comparison to identify severe weather conditions; the valid video image data is sent to the traffic monitoring video AI recognition module for analysis by the AI ​​model to output the severe weather condition.

[0149] Step (3), warning logic judgment: Based on the built-in warning triggering rules and warning cancellation rules, and according to the identified severe weather status data, the status judgment of each type of meteorological disaster at each monitoring location is performed, and corresponding warning triggering or cancellation instructions are given;

[0150] Specifically, the decision support module assesses the status of each type of meteorological disaster at each monitoring location based on a pre-set, editable, and maintainable "early warning triggering rule strategy table." The core parameters of this strategy table include:

[0151] Trigger sensitivity (number threshold N): Defines how many consecutive times a disaster status needs to be detected before an event is confirmed to have occurred or ended.

[0152] Trigger interval (duration threshold T): Defines the shortest time interval that must be allowed between two successful triggers of the same type of event.

[0153] This module performs the following logical judgment:

[0154] Warning Trigger Judgment (Message Sending): The system confirms the validity of the warning and proceeds to the next step only when a location is detected to meet the disaster triggering conditions N times consecutively, and the time elapsed since the last successful triggering of a similar warning at that location exceeds a duration T. Both conditions must be met simultaneously.

[0155] Warning Cancellation Judgment (Information Clearance): When a location is detected to have failed to meet the disaster triggering conditions N times consecutively, the system confirms that the disaster has ended, immediately generates a cancellation command, and proceeds to the next step. This judgment only requires one condition to be met.

[0156] Step (4), Cooperative control strategy execution: Determine the variable information board to receive the warning trigger or cancellation command and the corresponding display content according to the multi-segment variable information cooperative control strategy;

[0157] Once the decision support module determines, based on the above logic, that it needs to activate or cancel an early warning, it will execute the "multi-segment variable information collaborative control strategy".

[0158] The strategy is as follows:

[0159] a. Determine the location of the disaster: Based on the source of the data corresponding to the early warning trigger command, accurately locate the mileage marker of the meteorological disaster occurrence point.

[0160] b. Target Information Board Filtering: Centered on the disaster point, the system automatically retrieves and filters all variable message signs within a preset distance (e.g., 10 kilometers) upstream in both directions of traffic. This preset distance is flexible and not fixed, depending on the road grade, design speed, terrain (e.g., mountainous or plain areas), and the severity of the meteorological disaster.

[0161] c. Generate dynamic warning or cancellation information: If a warning is triggered, the distance to the disaster point is automatically calculated for each selected variable message sign based on its mileage marker, and personalized warning information containing the precise distance is generated; if a warning is cancelled, an instruction to clear the corresponding warning is generated.

[0162] Step (5), Information Linkage and Release: Send the warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display. Specifically:

[0163] The prevention and control equipment linkage module sends the warning trigger or cancellation command and corresponding display content to all selected target variable message signs via the network. After receiving and parsing the command, the variable message sign linkage system immediately publishes the corresponding dynamic warning information or cancellation information on the display screen (data display module).

[0164] This invention obtains weather condition data along highways through at least one of the following methods:

[0165] Sensor data acquisition: Real-time acquisition of weather sensor data (such as visibility meters, rain gauges, road condition sensors, etc.) that have passed quality inspection, through traffic and meteorological monitoring modules deployed along the highway.

[0166] Traffic surveillance video AI recognition: This invention utilizes AI visual recognition models deployed at the central or edge levels to analyze real-time video images transmitted from roadside surveillance cameras that have passed quality inspection, identifying severe weather conditions. This invention does not limit the specific algorithm or type of this AI visual recognition model; any model capable of quickly and accurately identifying one or more severe weather conditions from traffic surveillance video images can be applied within the technical framework of this invention.

[0167] The acquired severe weather data, along with its location information, will be transmitted at a fixed frequency (e.g., once per minute) to the central traffic meteorological monitoring, early warning, and decision-making system.

[0168] Table 1 shows the display content under different severe weather types, including the control measures to be taken and remarks.

[0169] Table 1

[0170]

[0171]

[0172] Table 2 shows the core parameters, such as "trigger sensitivity" (threshold for consecutive times) and "trigger interval" (threshold for duration), that are used to define and configure warnings and cancel warnings for different types of meteorological disasters at different monitoring locations.

[0173] Table 2

[0174]

[0175] A schematic diagram of the multi-segment variable information cooperative control strategy of this invention is shown below. Figure 2 As shown in the diagram, a two-way highway is displayed, indicating a meteorological disaster (such as dense fog) at a certain kilometer marker. Upstream from the disaster point for both directions of traffic, multiple variable message signs at different distances will be marked, schematically displaying the warning information shown on each sign, which includes different distance information.

[0176] The flowchart for the decision-making process of triggering and canceling early warnings is as follows: Figure 3 As shown in the flowchart, the system details the entire business loop after receiving monitoring data, from data quality inspection and status judgment to finally deciding whether to issue or cancel an alert based on two core parameters: "trigger sensitivity" (number of consecutive times N) and "trigger interval" (duration T).

[0177] Application Example: Collaborative Early Warning of Heavy Fog on Highways Based on Trigger and Cancellation Strategies

[0178] This example scenario involves a meteorological monitoring device deployed at kilometer marker K100+000 on a highway heading north (mileage markers from smallest to largest). The monitoring data is collected every minute. The "Early Warning Trigger Rule Strategy Table" in the system backend configures the following rules for the "dense fog" disaster type at this location:

[0179] Trigger sensitivity (N): 3 times

[0180] Trigger interval (T): 30 minutes

[0181] Step (1), Raw data acquisition and quality control: Acquire raw visibility data collected by meteorological sensors at K100+000 and / or raw video streams from roadside monitoring cameras; The data quality control module first checks these raw data and removes invalid data that exceeds the normal range or has logical errors; For example, it determines whether the visibility reading is negative or an abnormal value that is far beyond the range to confirm that the data is valid.

[0182] Step (2), severe weather condition identification: The valid sensor data that has passed quality inspection is sent to the traffic meteorological monitoring module for threshold comparison to identify severe weather conditions; the valid video image data is sent to the traffic monitoring video AI recognition module for analysis by the AI ​​model to output the severe weather condition.

[0183] The confirmation of dense fog at K100+000 can be achieved through any one or two of the following methods:

[0184] Method 1 (Sensor): When the visibility meter of the meteorological monitoring station deployed at this location measures data below the threshold, the device will report the "strong dense fog" status and location information.

[0185] Method 2 (AI Visual Recognition): The traffic monitoring video AI recognition module continuously analyzes the video stream transmitted from the roadside monitoring camera at K100+000. Based on image features (such as clarity, visibility of reference objects, etc.), it determines that the area is in a state of "strong dense fog". The system records this state and the corresponding camera location information.

[0186] The data is reported in real time to the traffic and meteorological monitoring, early warning and decision-making system deployed in the central computer room.

[0187] Step (3), warning logic judgment: After the traffic meteorological monitoring and warning decision system receives the "strong dense fog" status data, it makes a status judgment on each type of meteorological disaster at each monitoring location based on its built-in warning triggering rules and warning cancellation rules, and according to the data that has been quality checked, and gives the corresponding warning triggering or cancellation instructions.

[0188] Scenario 1: Warning Triggered

[0189] At 10:00, the system received the "dense fog" status data at K100+000 for the first time. The system's internal counter recorded the number of consecutive occurrences of "dense fog" as 1.

[0190] At 10:01, the system received "dense fog" status data again. The cumulative number of consecutive occurrences is 2.

[0191] At 10:02, the system received the "dense fog" status data for the third time. The cumulative number of consecutive occurrences is 3. At this point, 3 ≥ N(3), and the "trigger sensitivity" condition is met. The system then checks the "trigger interval" condition and queries the database, finding that there are no "dense fog" warning records for this location within 30 minutes, and the "trigger interval" condition is also met.

[0192] At this point, both triggering conditions have been met. The decision support module confirms the validity of this "dense fog" warning, records the successful trigger time as 10:02, and passes the warning instruction to the next step.

[0193] Scenario 2: Warning Cancellation

[0194] At 11:30, the weather improved, and the system received "no dense fog" status data at K100+000. The system's internal counter recorded the number of consecutive occurrences of "no dense fog" as 1.

[0195] At 11:31, the system received "no dense fog" status data again. The cumulative number of consecutive occurrences is 2.

[0196] At 11:32, the system received the "no dense fog" status data for the third time. The cumulative number of consecutive counts is 3. At this point, 3 ≥ N(3), and the "trigger sensitivity" condition for cancellation is met.

[0197] The decision support module immediately confirms that the "dense fog" disaster has ended, generates a cancellation command, and passes the command to the next step.

[0198] Step (4), Cooperative control strategy execution: Determine the variable information board to receive the warning trigger or cancellation command and the corresponding display content according to the multi-segment variable information cooperative control strategy;

[0199] When the warning was triggered at 10:02, the system identified the disaster location as K100+000. The system set the coordinated warning range to 12 kilometers uphill. The system automatically retrieved all variable message signs from the database between K112+000 and K100+000 in the uphill direction, and between K88+000 and K100+000 in the downhill direction.

[0200] The search results are as follows: There are information boards at K103+000 and K108+000 in the uplink direction; there is an information board at K92+000 in the downlink direction.

[0201] The system dynamically generates commands for these three information boards:

[0202] For the information board at K108+000: the calculated distance is 108-100=8 kilometers. The generated message is: "Dense fog 8 kilometers ahead, please slow down."

[0203] For the information board at K103+000: the calculated distance is 103-100=3 kilometers. The generated message is: "Dense fog 3 kilometers ahead, please slow down."

[0204] For the information board at K92+000 (for oncoming traffic): the calculated distance is 100-92=8 kilometers. The generated message is: "Dense fog 8 kilometers ahead, please slow down."

[0205] When the warning was lifted at 11:32, the system also locked these three information boards and generated instructions for them to clear the aforementioned dense fog warning.

[0206] Step (5), Information Linkage Release: Send the warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display.

[0207] At 10:02, the system's prevention and control equipment linkage module sent the early warning command to the three information boards, and the information boards displayed the corresponding early warning information almost simultaneously.

[0208] At 11:32, the module sends the cancellation command to the three information boards, which then clear the warning information and restore normal display.

[0209] Through the above methods, the issuance and revocation of warnings are subject to "delayed confirmation," ensuring the stability of information and avoiding confusion caused by brief fluctuations in data near critical values.

[0210] Comparative experiment

[0211] To further demonstrate the superiority of the stable early warning release strategy proposed in this invention based on "trigger sensitivity (number of consecutive times N)" and "trigger interval (duration T)", the following comparative experiment was set up.

[0212] Experimental setup:

[0213] Experimental group: The complete technical solution of this invention is adopted. Taking the scenario in the aforementioned application example as an example, the warning triggering rule is configured as follows: trigger sensitivity N = 3 times, trigger interval T = 30 minutes.

[0214] Control group: A simplified warning logic is used, i.e., without using the stable release strategy of this invention. The logic is as follows: once a single visibility data point is detected to be below a threshold (e.g., 50 meters), a "dense fog" warning is immediately triggered; once a single data point is detected to be above the threshold, the warning is immediately revoked. This is a common and simple threshold determination method in the prior art.

[0215] Simulated Scenario: A scenario is simulated near a traffic weather monitoring point (K100+000) where weather conditions are at a critical point. Visibility monitoring data fluctuates slightly around the 50-meter threshold within 15 minutes. The specific data sequence is as follows:

[0216] 10:00-65m (above the threshold)

[0217] 10:01-48m (below the threshold)

[0218] 10:02-52m (above the threshold)

[0219] 10:03-45m (below the threshold)

[0220] 10:04-49m (below the threshold)

[0221] 10:05-60m (above the threshold)

[0222] (Data for the next 9 minutes all exceeded the threshold)

[0223] Comparison of experimental results:

[0224] 1. Performance of the control group (simplified logic):

[0225] At 10:01, the system detected a visibility of 48m and immediately triggered an early warning. The upstream information board displayed "Dense fog X kilometers ahead".

[0226] At 10:02, the system detected a visibility of 52m, immediately canceled the warning, and cleared the information board information.

[0227] At 10:03, the system detected a visibility of 45m and triggered another warning. The information board once again displayed "Dense fog X kilometers ahead".

[0228] At 10:04, the system detected a visibility of 49m, and the warning continued to be displayed.

[0229] At 10:05, the system detected a visibility of 60m, and the warning was revoked again, and the information on the information board was cleared again.

[0230] Results Analysis: Within just 5 minutes, the warning information on the information board experienced two frequent on / off cycles: "display -> clear -> display -> clear." This "flickering" phenomenon severely interferes with the driver's judgment and reduces their trust in the warning system.

[0231] 2. Performance of the experimental group (the scheme of this invention):

[0232] At 10:01, the system detected dense fog, and the internal counter recorded it as "1 consecutive time". The condition N=3 is not met, and no warning is triggered.

[0233] At 10:02, the system detected no dense fog and reset the internal counter to zero.

[0234] At 10:03, the system detected dense fog, and the internal counter recorded it as "1 time consecutively".

[0235] At 10:04, the system detected dense fog, and the internal counter incremented to "2 consecutive times". This still does not meet the condition N=3, so no warning is triggered.

[0236] At 10:05, the system detected no dense fog and reset the internal counter to zero.

[0237] Results Analysis: During the entire 15-minute critical fluctuation period, the system of this invention did not issue any early warning information. It successfully filtered out invalid data interference caused by brief fluctuations in weather near the critical point, avoiding the "flickering" phenomenon. Only when the weather continues to deteriorate steadily (e.g., visibility is below the threshold for 3 consecutive minutes or more) will the system trigger a stable and reliable early warning.

[0238] Experimental Conclusion: Comparative experiments show that omitting the core stable release strategy of this invention results in frequent and invalid alarms under critical weather conditions, posing a potential threat to driving safety. This invention, by introducing a dual judgment mechanism based on consecutive counts and trigger intervals, effectively ensures the stability and reliability of early warning information release. This demonstrates that the synergistic effect of various technical steps and parameters is essential for the superior effects of this invention, representing a significant advancement.

[0239] Practical application cases

[0240] To further demonstrate the feasibility and practicality of the technical solution of this invention, Figures 4 to 8 The document displays screenshots of the actual operating interface of the software platform developed based on the method of this invention (the platform is a scientific and technological achievement of the Yunnan Provincial Transportation Investment and Construction Group Co., Ltd. Science and Technology Innovation Project (Project Name: Research and Application of Joint Control Technology for Monitoring, Early Warning and Prevention of Traffic Meteorological Disasters on Expressways, Project No.: YCIC-YF-2021-08)). The system has been deployed and applied in traffic disaster-prone road sections in northeastern Yunnan Province.

[0241] Figure 4 This is a screenshot of the login interface for the traffic meteorological monitoring, early warning, and decision-making system. The system adopts a B / S architecture, which facilitates authorized users to access and manage it remotely via a browser.

[0242] Figure 5 The main interface for monitoring variable message sign (VMS) equipment in the VMS control system is showcased. On this interface, managers can intuitively view the real-time status, station location, and currently displayed content of all VMS signs within their jurisdiction, achieving centralized and visual management of the front-end equipment.

[0243] Figure 6This is a screenshot of the information editing interface for the information board. This interface demonstrates the high flexibility and configurability of the warning information in this invention. Users can not only select different warning rule types (such as heavy rain), but also customize the font, color, size, and other styles of the information. More importantly, the phrase "XX kilometers ahead" in the content editing box intuitively reflects the core of this invention.

[0244] The dynamic distance calculation and personalized information generation functions represent a significant improvement over the fixed and general early warning content in existing technologies.

[0245] Figure 7 The historical logs of warning messages sent from the information boards are displayed. The logs show that at a specific time point (e.g., 09:45:02 on April 8, 2025), the system automatically issued similar warning messages ("Dense fog 3 km ahead," "Dense fog 4 km ahead") to multiple information boards at different station numbers (e.g., downstream K1709+050k, downstream K1710+790k, etc.). This strongly demonstrates the actual operational effectiveness of the multi-segment collaborative control strategy of this invention, achieving regionalized and networked early warning for a disaster point, highlighting the superiority of this invention.

[0246] Figure 8 These are photos taken on actual roads showing the display effect of information boards. The photos accurately reflect that the warning information from this invention's system has been successfully and clearly disseminated to road users, completing the entire closed-loop control from monitoring and decision-making to dissemination, demonstrating the high practical feasibility and application value of this invention.

[0247] In summary, this application example fully demonstrates the entire process of the present invention from backend management to frontend presentation, proving that the present invention is not only technically feasible, but also has a higher degree of automation, more accurate early warning information, and a wider range of collaborative coverage compared with existing technologies in practical applications.

[0248] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A multi-segment collaborative early warning platform based on multi-source traffic meteorological data, characterized in that, This includes a multi-source data acquisition system, a traffic and meteorological monitoring, early warning, and decision-making system, and a variable message sign control system; The multi-source data acquisition system includes a data quality control module, a traffic and meteorological monitoring module, and a traffic surveillance video AI recognition module; The data quality control module is connected to the weather sensor, roadside monitoring camera, traffic meteorological monitoring module, and traffic monitoring video AI recognition module, respectively. It is used to perform quality inspection on the data collected by the weather sensor and the video image data collected by the roadside monitoring camera. After removing invalid data through preset quality control rules, the valid data collected by the weather sensor is transmitted to the traffic meteorological monitoring module, and the valid video image data collected by the roadside monitoring camera is transmitted to the traffic monitoring video AI recognition module. The traffic weather monitoring module is used to receive data collected by weather sensors that have passed quality inspection, and to determine whether severe weather conditions have occurred by comparing the collected data with preset severe weather judgment thresholds. The traffic monitoring video AI recognition module is equipped with an AI model that can recognize severe weather; the traffic monitoring video AI recognition module is used to receive video images from roadside monitoring cameras that have passed quality inspection. The AI ​​model, capable of recognizing severe weather, analyzes video images and directly outputs the recognition results of severe weather conditions. The traffic meteorological monitoring, early warning and decision-making system includes a data receiving module, a data visualization module, a decision support module, and a prevention and control equipment linkage module; The data receiving module is connected to the traffic meteorological monitoring module and the traffic monitoring video AI recognition module respectively, and is used to receive severe weather status data identified by the traffic meteorological monitoring module and the traffic monitoring video AI recognition module. The data visualization module is connected to the data receiving module and is used to visualize the received severe weather data. The decision support module, connected to the data receiving module, is used to determine the status of each type of meteorological disaster at each monitoring location based on its built-in early warning triggering rules and early warning cancellation rules, and according to the received severe weather status data, and to give the corresponding early warning triggering or cancellation instructions. Then, based on the multi-segment variable information collaborative control strategy, it determines the variable information board to receive the early warning triggering or cancellation instructions and the corresponding display content. The prevention and control equipment linkage module is connected to the decision support module and is used to send the early warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display. The specific rules for triggering early warnings in the decision support module are as follows: when the system detects that a certain location meets the disaster triggering conditions N times consecutively, and the time elapsed since the last successful triggering of a similar early warning at that location has exceeded a duration T, then the early warning is deemed valid and an early warning triggering command is generated. The multi-segment variable information cooperative control strategy is as follows: a. Determine the location of the disaster point: Based on the source of the data corresponding to the early warning triggering instruction, locate the mileage marker of the meteorological disaster occurrence point; b. Filter target information boards: With the disaster point as the center, automatically search and filter all variable information boards within a preset distance range upstream of the two-way traffic direction; c. Generate dynamic early warning or cancellation information: If an early warning is triggered, the distance to the disaster point will be automatically calculated for each selected variable message sign based on its mileage marker, and early warning information including the distance will be generated. If the warning is to be revoked, a command to clear the corresponding warning will be generated.

2. The multi-segment collaborative early warning platform based on multi-source traffic meteorological data according to claim 1, characterized in that, Weather sensors include visibility meters, rain gauges, and road condition sensors.

3. The multi-segment collaborative early warning platform based on multi-source traffic meteorological data according to claim 1, characterized in that, Severe weather conditions include dense fog, blizzards, heavy rain, icy roads, strong winds, high road temperatures, sandstorms, and thunderstorms.

4. The multi-segment collaborative early warning platform based on multi-source traffic meteorological data according to claim 1, characterized in that, The specific rules for canceling the warning are as follows: when a location is detected to have failed to meet the disaster triggering conditions N times consecutively, it is determined that the disaster has ended, and a warning cancellation command is immediately generated.

5. The multi-segment collaborative early warning platform based on multi-source traffic meteorological data according to claim 1, characterized in that, The variable message sign control system includes a data receiving and parsing module, a data display module, and a control module; The data receiving and parsing module receives instructions and corresponding display content from the prevention and control equipment linkage module, and then sends the parsed content to the control module. The control module controls the data display module to display the corresponding content based on the parsed content.

6. A multi-segment collaborative early warning method based on multi-source traffic meteorological data, employing the multi-segment collaborative early warning platform based on multi-source traffic meteorological data as described in any one of claims 1 to 5, characterized in that, Includes the following steps: Step (1), Raw data acquisition and quality control: Acquire raw data from weather sensors and / or roadside monitoring cameras, and first perform quality checks on the raw data to remove invalid data that exceeds the normal range or contains logical errors; Step (2), severe weather condition identification: The valid sensor data that has passed quality inspection is sent to the traffic meteorological monitoring module for threshold comparison to identify severe weather conditions; the valid video image data is sent to the traffic monitoring video AI recognition module for analysis by the AI ​​model to output the severe weather condition. Step (3), warning logic judgment: Based on the built-in warning triggering rules and warning cancellation rules, and according to the identified severe weather status data, the status of each type of meteorological disaster at each monitoring location is judged, and corresponding warning triggering or cancellation instructions are given; Step (4), Cooperative control strategy execution: Determine the variable information board to receive the warning trigger or cancellation command and the corresponding display content according to the multi-segment variable information cooperative control strategy; Step (5), Information Linkage Release: Send the warning trigger or cancellation command and the corresponding display content to the designated variable message sign for display.

Images