Traffic emergency early warning management system suitable for severe weather

Abstract

The invention discloses a traffic emergency early warning management system suitable for severe weather, which relates to the technical field of traffic safety, and comprises a remote sensing monitoring unit for acquiring remote sensing data in a monitoring area in real time, extracting a wind and sand area through spectral characteristics and multi-time sequence image analysis, generating a wind and sand dynamic track and a flowing trend, and performing early warning and early warning. The prediction unit evaluates the influence degree of wind and sand on different road sections in combination with a traffic network of a monitoring area, and the early warning strategy unit dynamically plans a detouring route and speed limiting measures by utilizing an optimization algorithm in combination with a wind and sand influence evaluation result and real-time traffic flow. The remote sensing image technology is adopted, large-range monitoring data is obtained in real time, through spectral feature extraction and multi-time-sequence image analysis, extraction of wind and sand concentration distribution is achieved, the wind and sand diffusion process is analyzed through a fluid dynamic model and remote sensing data, the diffusion tracks of wind and sand in different time periods are simulated in combination with topographic conditions, and the wind and sand concentration distribution is obtained. And a result is evaluated based on real-time traffic flow data and sand wind influence.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic safety, and in particular to a traffic emergency warning management system suitable for severe weather. Background Art

[0002] Sandstorms are a common natural phenomenon in arid and semi-arid regions, significantly impacting regional traffic safety and operational efficiency. Specifically, large amounts of sand suspended in the air significantly reduce drivers' field of vision, degrading visibility. When visibility drops below 100 meters, drivers struggle to detect vehicles ahead and road conditions, which can easily lead to traffic accidents. Furthermore, sandstorms often cause traffic jams, forcing vehicles to slow down or stop, reducing road capacity. This can be particularly detrimental to the economy at transportation hubs or along key logistics corridors.

[0003] Existing methods for monitoring and assessing wind and sand are mostly based on distribution data on wind speed and sand concentration, lacking specificity for the monitored area. This means they lack a comprehensive analysis of wind and sand diffusion characteristics and terrain conditions, making it difficult to accurately characterize the impact of wind and sand on different traffic sections. Particularly in areas with complex terrain, the flow and diffusion of wind and sand are significantly affected by terrain characteristics. Existing methods struggle to fully account for this dynamic relationship, resulting in inaccurate risk assessments. Furthermore, when generating traffic emergency plans based on the diffusion trajectory of wind and sand, they often rely on static traffic network models, formulating emergency plans by analyzing the impact of wind and sand diffusion and flow on road networks at different distances. This ignores the real-time characteristics of vehicle flow and traffic dynamics, making it difficult for detour routes or speed limits to match actual traffic conditions. This can lead to emergency plans being unfeasible or inefficient. Summary of the Invention

[0004] 1) Technical issues solved The present invention provides a traffic emergency warning management system suitable for severe weather, which realizes full-chain management from monitoring and prediction to emergency decision-making and feedback optimization.

[0005] 2) Technical solution To achieve the above objectives, the present invention provides the following technical solutions: a traffic emergency warning management system suitable for severe weather, comprising: The remote sensing monitoring unit acquires remote sensing image data within the monitoring area in real time, extracts the wind-blown sand coverage area and its sand concentration distribution by calculating the spectral difference between the corresponding bands of sand and dust in the remote sensing image and the background objects; uses multi-time series remote sensing images to analyze the spatiotemporal variation characteristics of the sand concentration distribution, calculates the transfer rate and direction of the wind-blown sand, and combines it with a fluid dynamics model to simulate the movement trajectory of the wind-blown sand and its changing trend; The prediction unit integrates the dust concentration distribution data provided by the remote sensing monitoring unit with the terrain and traffic data of the monitoring area into a unified grid map, calculates the average wind-blown sand concentration and wind speed of the grid cells covered by each road based on the overlap ratio between road segments and each grid cell in the grid map, and assesses the impact of wind-blown sand on the traffic conditions and risk level of each road; An early warning decision unit generates a traffic emergency plan based on the wind and sand impact data evaluated by the prediction unit, aligning it with real-time traffic flow data in time and space. The plan selects a road route with the lowest detour time cost as the detour route plan, and a speed limit zone on a high-risk road section through an optimization algorithm. a data management unit that stores and analyzes the remote sensing monitoring data, sandstorm impact prediction results, and historical data on traffic emergency plans, and uses machine learning technology to optimize the path planning and speed limit calculation of the optimization algorithm in the early warning decision-making unit; The traffic guidance unit receives and publishes the traffic emergency plan generated by the early warning decision unit in real time, and pushes the detour route plan and speed limit zone information to the traffic management department and drivers in the monitoring area in real time.

[0006] Furthermore, the frequency of acquiring remote sensing images of the monitoring area by the remote sensing monitoring unit is set according to the period of frequent sandstorms, and after acquiring the images, the remote sensing data is corrected; The corrected remote sensing images are subjected to band fusion, the spectral index of the selected specific band of wind and sand is calculated, and the dust coverage range is extracted by setting the threshold segmentation range.

[0007] Furthermore, the remote sensing monitoring unit inverts the movement trajectory and flow trend of the wind and sand by comparing the position changes of the wind and sand area in different time periods based on the remote sensing image data of multiple time periods; specifically: Perform geometric correction on remote sensing images at different times to ensure the consistency of pixel positions; Using a change detection algorithm to identify changes in dust areas at different times, the dust areas at different times are superimposed, the movement direction and speed are calculated based on the time intervals, and a dynamic trend map of wind and sand is obtained by analyzing the dust movement trajectory; The wind speed is calculated based on the movement rate of the wind and sand area, the concentration changes are calibrated using spectral characteristics and ground reference points, and a sand concentration inversion model is established to obtain the concentration distribution of the wind and sand.

[0008] Furthermore, the prediction unit receives the wind and sand coverage area, wind speed and movement trajectory data provided by the remote sensing monitoring unit, and obtains the terrain data and traffic basic data of the monitoring area, projects the remote sensing data, terrain data and traffic data into a grid map with unified coordinates, simulates the diffusion process of wind and sand particles through a fluid dynamics model, and combines the terrain conditions and wind speed and direction characteristics to predict the spatial distribution of wind and sand in different time periods.

[0009] Furthermore, the prediction unit superimposes the wind and sand diffusion range with the traffic network, determines the affected road sections through spatial intersection analysis, evaluates the impact on road traffic based on wind and sand concentration, wind speed and terrain conditions, including the location of the road section and the type of impact, and sets impact thresholds to divide the sections in the monitoring area into high-risk, medium-risk and low-risk routes.

[0010] Furthermore, the prediction unit integrates the wind and sand movement trajectory, the topography and traffic data of the monitoring area to provide a wind and sand concentration distribution forecast map for several time periods in the future, and outputs a traffic risk level distribution map, marking high-risk sections and alternative routes.

[0011] Furthermore, the early warning decision unit integrates the wind and sand impact assessment results provided by the prediction unit with real-time traffic flow data to provide basic data for dynamic planning; specifically: Align wind and sand impact data with real-time traffic flow data in time and space, quantify the data uniformly based on road risk level and capacity, and construct a comprehensive assessment matrix; An optimization algorithm is used to calculate the path with the minimum detour time cost in the wind and sand covered area as the detour path plan; wherein, a list of high-risk road sections that are impassable and have speed limits is extracted from the evaluation results of the prediction unit, and a set of candidate routes for bypassing the high-risk routes is generated based on the topological structure of the traffic road network in the monitoring area. The optimization algorithm is used to select the path with the minimum overall travel time cost from the candidate routes as the detour path plan, and the safe driving speed threshold is calculated based on the wind and sand concentration, wind speed and road section attributes, so as to set the speed limit interval.

[0012] Furthermore, the receiving warning decision unit integrates the detour route, speed limit measures and real-time traffic flow information to generate a traffic emergency plan for the entire area, and sends the traffic emergency plan to the traffic guidance unit, which transmits it to the traffic management department and drivers traveling in the monitoring area.

[0013] Furthermore, the data management unit receives and stores the remote sensing image data and wind-blown sand concentration distribution results from the remote sensing monitoring unit, the road risk assessment results from the prediction unit, and the emergency plan generated by the early warning decision-making unit. The data management unit indexes the data by time and region to build a case library of historical emergency plans and implementation effects. The historical emergency plans were compared and analyzed with the actual implementation results to extract the impact indicators. Based on the case library, machine learning technology was used to optimize the path planning and speed limit algorithms in the early warning decision unit.

[0014] 3) Beneficial effects: Compared with the prior art, this invention has the following beneficial effects: This invention uses remote sensing imaging technology to obtain large-scale monitoring data in real time, and realizes the extraction of wind sand concentration distribution through spectral feature extraction and multi-time series image analysis. It uses fluid dynamics models and remote sensing data to deeply analyze the wind sand diffusion process, and combines terrain conditions and wind speed and direction characteristics to accurately simulate the diffusion trajectory of wind sand in different time periods.

[0015] Based on real-time traffic flow data and wind and sand impact assessment results, detour routes and speed limit measures are dynamically planned to balance traffic efficiency and safety. The data management unit stores and indexes historical disaster cases and emergency plans, and continuously optimizes path planning and speed limit algorithms through machine learning to improve the system's adaptability and safety management level. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A functional block diagram of a traffic emergency warning management system suitable for severe weather provided by an embodiment of the present invention; Figure 2 A schematic diagram showing the connections of various units in a traffic emergency warning management system suitable for severe weather conditions provided by an embodiment of the present invention; Figure 3 A schematic diagram of a process for acquiring remote sensing data in a traffic emergency warning management system applicable to severe weather provided by an embodiment of the present invention; Figure 4 A schematic diagram of a process for spatiotemporally matching wind and sand trajectories with traffic routes in a traffic emergency warning management system for severe weather provided by an embodiment of the present invention; Figure 5 A schematic diagram of a flow chart of a traffic emergency warning decision unit planning a traffic emergency plan in a traffic emergency warning management system applicable to severe weather provided by an embodiment of the present invention; In the picture: 10. Remote sensing monitoring unit; 20. Forecasting unit; 30. Early warning decision-making unit; 40. Traffic guidance unit; 50. Data management unit. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0018] In the description of the present invention, it should be understood that the terms "longitudinal", "transverse", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0019] In addition, the terms "first", "second", etc., if used, are merely used to distinguish and describe, and should not be understood as indicating or implying relative importance.

[0020] It should be noted that, in the absence of conflict, the features in the embodiments of the present invention may be combined with each other.

[0021] Severe weather, such as sandstorms, can reduce visibility and reduce road friction, which can easily lead to traffic accidents. Considering that the existing traffic emergency management system mainly obtains real-time traffic data and meteorological data, its combination of sandstorms and traffic data is limited in terms of its impact and early warning function, and it is unable to quickly generate targeted traffic emergency plans. Therefore, combined with Figures 1 to 5 As shown, embodiments of the present invention provide a traffic emergency warning and management system suitable for severe weather. This system is applicable to desertified and arid regions, as well as areas prone to frequent sandstorms. Based on real-time traffic flow data and its generated assessment of the impact of wind and sand on traffic, the system dynamically plans detour routes and speed limits, balancing traffic efficiency and safety. Furthermore, by storing and indexing historical disaster cases and emergency plans, the system leverages machine learning to continuously optimize route planning and speed limit algorithms, enhancing the system's adaptability and safety management capabilities.

[0022] Specifically, the first is a remote sensing monitoring unit 10 for acquiring data. This unit obtains remote sensing image data from a selected monitoring area in real time, calculates the spectral difference between the corresponding band of sand and dust in the remote sensing image and the background objects, extracts the wind-blown sand coverage area, and then calculates the dust concentration distribution within the coverage area based on the extracted wind-blown sand coverage area and its spectral index. The unit also uses multi-time series remote sensing images to analyze the spatiotemporal variation characteristics of the dust concentration distribution, thereby calculating the wind-blown sand transfer rate and movement direction in the wind-blown sand coverage area. Finally, a fluid dynamics model is introduced to simulate the movement trajectory of the wind-blown sand based on the transfer rate and direction of the wind-blown sand. The fluid dynamics model can also be used to simulate the changing trend of the wind-blown sand flow.

[0023] More specifically, the remote sensing monitoring unit 10 acquires remote sensing images. This data acquisition utilizes satellite sensors (such as Landsat and Sentinel-2) or unmanned aerial vehicles (UAVs) to receive electromagnetic wave information reflected from the ground surface. Specific wavelengths, such as visible light, near-infrared light, and short-wave infrared light, are used to record surface characteristics based on the reflectance spectrum of these wavelengths. More specifically, it can be understood that the spectral characteristics of wind-blown sand, such as low reflectivity and high scattering, allow it to be distinguished from its surroundings.

[0024] In some embodiments, satellite remote sensing or drone remote sensing is selected based on the monitoring scope and requirements. Satellite remote sensing is more suitable for large-scale regional monitoring due to its periodic acquisition characteristics, while drone remote sensing is more suitable for localized regional monitoring due to its flexibility and high resolution. The monitoring method is not specifically limited herein.

[0025] Regarding the setting of monitoring frequency, in some embodiments of the present invention, by analyzing the time periods of historical sandstorm events, such as seasonal variations, key periods can be identified for increased monitoring frequency. Alternatively, for certain high-frequency periods of sandstorm events, depending on the characteristics of regional sandstorm activity, in some embodiments, image acquisition can be set to occur daily or every few hours. It is important to note that the imagery must ensure complete coverage of the area and avoid cloud cover or other obstructions as much as possible.

[0026] Reference here Figure 3 After acquiring preliminary remote sensing images, preliminary correction is required. This includes radiometric, geometric, and atmospheric correction to eliminate image noise and distortion. Geometric correction of remote sensing images can be performed using ground control points (GCPs) to correct for geometric errors caused by satellite or drone tilt and the curvature of the earth. Alternatively, in some embodiments, a digital elevation model (DEM) can be used for refined correction to ensure accurate alignment between the image and geographic coordinates.

[0027] Regarding the radiometric correction of remote sensing images, this is to remove the effects of atmospheric scattering and absorption and restore the true surface reflectivity. In some embodiments, the original image data is corrected by using an atmospheric correction algorithm, such as FLAASH or the 6S model.

[0028] After preliminary correction of the remote sensing image, sensitive bands are selected to generate a multi-band composite image. This is because dust has a high visible light reflectivity, especially in the red band, and a lower near-infrared reflectivity, while also exhibiting specific absorption characteristics in the mid-infrared band. Therefore, the dust area can be extracted through band ratios or indices.

[0029] To summarize the above, in some embodiments of the present invention, the sensitive bands for wind and sand are set to the shortwave infrared and red light bands. Among them, the area covered by wind and sand usually exhibits a higher reflectivity in the shortwave infrared band due to the significant scattering characteristics of its particles. In the red light band, wind and sand exhibit a lower reflectivity, thereby distinguishing it from other land features such as vegetation or water bodies.

[0030] Regarding the extraction of spectral features of remote sensing images, the spectral index can enhance the characteristics of wind and sand and weaken the interference of other ground objects. More specifically, when the sensitive band is selected as short-wave infrared and red light band Finally, the shortwave infrared and red light bands are extracted from the remote sensing images, and the spatial resolution of different bands is unified.

[0031] For each pixel in the remote sensing image, or divide the remote sensing image into several sub-regions, calculate the spectral index for each sub-region The calculated spectrum result is a single-band image of the enhanced sandstorm area. :

[0032] About Spectral Index The value of is usually in the range of -1 to 1. The value of the sandy area is within a specific range and is usually positive.

[0033] When calculating the spectral index of each pixel of the remote sensing image, or in other words, when calculating the spectral index of each sub-region After setting the value of Thresholds are compared to distinguish between wind-blown sand areas and non-wind-blown sand areas. The thresholds can be calibrated using empirical formulas or ground-based data, and the specific values ​​are not limited here.

[0034] Finally, morphological operations such as dilation and erosion are used to optimize the extraction results, remove isolated pixels or noise, and generate a binary layer of the wind and sand coverage area to provide input for trajectory deduction.

[0035] Based on a dataset of remote sensing images acquired over multiple time periods, the movement trajectory and flow trends of windblown sand are inverted by comparing the positional changes of windblown sand areas in each remote sensing image over time. The movement of windblown sand is influenced by factors such as wind speed, topography, and vegetation cover, and its trajectory can be inverted through change detection in multiple time-series remote sensing images. This can be understood as temporal differences reflecting changes in the position of windblown sand. The windblown sand areas at different times are superimposed, and the movement direction and velocity are calculated by combining time intervals and spatial displacements.

[0036] What needs to be ensured here is that the remote sensing image data sets obtained above at different time periods need to have a consistent geographic coordinate system. In some embodiments, ground control points, such as obvious ground features, can be selected for geometric correction. After correction, the resolution of remote sensing images at different times is standardized.

[0037] The remote sensing images are then differentiated to identify pixels or sub-regions with significant differences, representing changes in the sandstorm area. Each moment's remote sensing image is individually classified to extract the sandstorm area. The classification results are then compared to identify the spread or movement of dust, generating a layer showing the location changes of the sandstorm area at different times.

[0038] The overall steps can be understood as superimposing the detected wind-blown sand areas, marking the position changes at each moment, and calculating the wind-blown sand movement rate using the time interval and spatial displacement. Vector field analysis methods can then be used to describe the movement direction and trajectory of the wind-blown sand. The trajectory data is interpolated to generate a wind-blown sand dynamic trend map. Regarding the interpolation method, Kriging spatial interpolation can be used in some embodiments.

[0039] Regarding the concentration distribution of wind-blown sand, the spectral characteristics and ground-measured data can be used to infer the concentration distribution of wind-blown sand. Specifically, typical pixels or sub-areas in the wind-blown sand covered area are selected, their spectral index values ​​are extracted, and they are calibrated with the ground monitoring data to establish a mathematical model of spectral index and concentration. For example, the monitoring data of PM10 (particulate matter concentration) from multiple ground monitoring stations can be selected as samples, and compared with the spectral index of the short infrared light band at the corresponding location in the remote sensing image, and combined with the extracted atmospheric particulate matter related parameters, a regression model is constructed to establish a regression relationship. The constructed regression model is then applied to estimate the concentration of PM10 in all wind-blown sand covered areas in the remote sensing image, thereby generating a wind-blown sand concentration distribution map for the entire area, with the concentration unit being . In addition, in some embodiments, during the testing phase of building a regression model, the model's prediction results can be compared with data from unmodeled ground monitoring stations, thereby adjusting the regression parameters to improve prediction accuracy. This allows for dynamic sand and dust change analysis based on a wind and sand concentration distribution map.

[0040] In summary, after acquiring remote sensing images of the monitored area, the remote sensing monitoring unit 10 analyzes and outputs the distribution of wind-blown sand concentration, including the sand movement rate and direction within the wind-blown sand-covered area, as well as the changing trend of sand concentration. This data provides spatiotemporal data on wind-blown sand movement, which is used to simulate future spread and assess its impact on traffic safety within the transportation network.

[0041] After acquiring the topographic and traffic data of the monitored area, the prediction unit 20 of the system integrates the dust concentration distribution data provided by the remote sensing monitoring unit 10, together with the topographic and traffic data, into a unified raster map. This allows the system to calculate the average wind and sand concentration and wind speed of the grid cells covered by each road in the raster map based on the overlap ratio between each road in the traffic route and each grid cell on the map. Given a known road risk assessment standard, the impact of wind and sand on the traffic conditions and risk level of each road can be assessed. It should be noted that this known risk assessment standard can be specifically set for areas with different accident frequencies and areas with different meteorological risks.

[0042] For details, please refer to Figure 4 The prediction unit 20 receives the coverage of the sandstorm area, the dynamic trajectory and movement rate of the sandstorm, and the dust concentration distribution map provided by the remote sensing monitoring unit 10, and obtains the terrain data and basic traffic data of the monitoring area. The terrain data includes elevation data, terrain slope, and slope aspect. The basic traffic data includes the regional transportation network, such as road location, grade, and flow data, as well as road facility information, such as bridges, tunnels, and key area markings.

[0043] The dust concentration distribution data, terrain data, and transportation infrastructure data (traffic network) are integrated into a unified grid map and then processed uniformly. The definition of grid here is also understood as dividing the monitoring area into grid units of equal size. For example, if the monitoring area is in kilometers, it can be divided into 10m 10m grid cell.

[0044] In the divided grid map, the information stored in each grid cell is: dust concentration ( )、Dust wind speed( ), terrain elevation value ( ) and road coverage information (the coverage ratio of roads in the grid).

[0045] Based on the obtained line segment coordinates of each road in the traffic network and the position coordinates of the grid cells, the overlap ratio between the road and each grid cell is calculated. In other words, the overlap area between the road and the grid cell is calculated as the fraction of the total area of ​​the grid cell. The windblown sand concentration and wind speed for each road are the weighted averages of the values ​​in the grid cells it covers.

[0046] After unified rasterization processing, the spatial distribution of traffic network nodes and roads is analyzed. Through the geographic information of the road network, such as road section ID and geographic coordinates, the wind and sand impact area is matched with the spatial location of traffic flow. Spatial cross-analysis is performed on the affected road nodes and traffic flow data to determine the extent of the impact of wind and sand on specific road sections.

[0047] The wind and sand assessment results (wind and sand concentration, wind speed) are then synchronized with the real-time traffic flow data along the time axis. Considering the predicted time window for wind and sand diffusion, the traffic flow data is divided into the same time granularity, such as 5 minutes or 10 minutes.

[0048] After analyzing the spatial distribution of wind and sand, the impact of wind and sand on different road sections is analyzed by combining the transportation network with this spatial distribution data. For example, impact levels are categorized based on road grade and wind and sand concentration. High-concentration areas in the wind and sand assessment results are marked as high-risk sections, and key sections and nodes severely affected by wind and sand, such as highways and important transportation hubs, are identified.

[0049] refer to Figure 5 After aligning the wind and sand impact data with real-time traffic flow data in time and space, the data is uniformly quantified according to indicators such as road importance, risk level, and traffic capacity to construct an evaluation matrix. Specifically, the constructed evaluation matrix maps the risk level and traffic capacity of each road section to a unified quantitative indicator, such as the quantitative traffic weight:

[0050] in, is the concentration of wind and sand, is the wind speed of the sandstorm, is the mean of the historical traffic flow data of the road.

[0051] According to the road risk assessment standard, the calculated sand concentration and wind speed values ​​are mapped to risk levels. The risk assessment standard can be formulated based on existing environmental monitoring and traffic safety research data indicators. For example, the impact of sand and dust concentration on traffic and health can refer to the World Health Organization and national air quality standards:

[0052] When the light is bright and clear, it will cause significant harm to health and affect visibility; When the pollution is extremely high, visibility is significantly reduced and the risk to traffic safety is high.

[0053] The impact of dust wind speed on traffic can be referred to the "Highway Wind Environment Design Specifications": lateral wind speed When the vehicle is moving, it will significantly affect the vehicle's handling performance; lateral wind speed When the vehicle is moving, it may easily cause the light vehicle to roll over or lose control.

[0054] Considering that this solution is mainly applicable to desertified areas, arid areas, and areas with frequent sandstorms, in some embodiments of experimental research areas with reference to open desertified areas, the impact of sandstorm wind speed on traffic is set as: Dust wind speed When the wind and sand begin to move significantly, affecting the sight and road surface; Dust wind speed When the wind and sand concentration and flow rate increase significantly, it poses a threat to road conditions.

[0055] Taking into account the cross-influence of dust concentration and wind speed, in some embodiments, the risk level of dust to roads is classified as shown in the following table.

[0056] For example, suppose the average sand concentration of the grid cells covered by a road is , wind speed is , the overall risk of the road is medium.

[0057] The system's early warning decision-making unit 30 generates a dynamic traffic emergency plan based on the sandstorm risk assessment results from the prediction unit 20, combined with real-time traffic flow data. Specifically, this is done by spatially and temporally aligning sandstorm impact data with real-time traffic flow data, and constructing a comprehensive assessment matrix based on the risk level and capacity of each road. An optimization algorithm is then used to calculate possible detour routes within the sandstorm-affected area. Taking into account both efficiency and safety, the route with the lowest time cost is selected as the detour route plan. Furthermore, based on the risk level of each road and combined with the aforementioned characteristics such as sandstorm concentration and sand transport rate, speed restrictions are set for high-risk sections.

[0058] Among them, the travel time is the total travel time of the predicted route, the safety is whether the safe driving speed of different sections meets the predicted safety threshold, and the detour cost is the additional mileage and time caused by the detour.

[0059] Path optimization involves planning paths in a traffic network affected by wind and sand using an optimization algorithm. The specific steps include extracting impassable sections, such as those with high risk levels and above, based on the evaluation results provided by the prediction unit 20, and marking these sections as being assessed to be high risk or above. For example, in the road network topology structure diagram of the acquired traffic road network map, candidate paths that bypass high-risk areas are searched within the road network topology structure. It should be noted that the candidate paths are selected based on avoiding high-risk sections and balancing travel time and driving safety.

[0060] In some embodiments, an optimization algorithm can be used to calculate the travel time cost of candidate paths in a traffic network affected by wind and sand. , its target optimization algorithm is: in, For the path The length of the road section, The safe driving speed for the road section.

[0061] Finally, the optimal algorithm is used to calculate and output the path with the minimum travel time cost as the recommended detour route.

[0062] Regarding the speed limit interval setting in the early warning strategy, it can be adjusted according to the wind and sand concentration. , wind speed of sandstorms And the road segment attributes, that is, the road safety level mentioned above, calculate the safe driving speed: in, The function is a safe driving speed model trained based on historical data, which takes into account factors such as visibility and wind speed. The specific form of the function is not limited.

[0063] Afterwards, based on the above-mentioned impact levels, graded speed limit intervals are set on high-risk sections, such as a speed limit gradient of every 20 kilometers per hour, and the traffic control system is notified to implement dynamic speed limits.

[0064] Finally, using the above main indicators, assign a weighted score to each candidate route : in, and Safety indicators and detour costs, 、 and It is a weight coefficient, which is adjusted according to actual conditions to prioritize safety.

[0065] In summary, it can be understood that through the above process of the early warning decision unit 30, the impact of wind and sand and traffic flow information can be efficiently integrated, and a traffic emergency plan that takes into account both safety and efficiency can be dynamically planned to minimize the impact of wind and sand on traffic.

[0066] The traffic guidance unit 40 of the system accurately and efficiently transmits the traffic emergency plan generated by the early warning decision unit 30 to relevant departments and drivers, ensuring that the implementation of the plan can quickly respond to the dynamic impact of wind and sand on traffic.

[0067] Specifically, the traffic guidance unit 40 receives the optimized traffic emergency plan from the warning decision unit 30, including detour routes, high-risk road section information, speed limit intervals, and safe driving tips. It is important to ensure that the received data is seamlessly connected to the interface of the real-time traffic monitoring system.

[0068] Emergency plans can be transmitted to traffic management departments via dedicated data communication interfaces for unified traffic flow control. These measures include, but are not limited to, setting up traffic signs in high-risk areas (such as speed limit warnings and detour indicators); adjusting traffic signal control logic to prioritize the evacuation of vehicles in affected areas; and activating emergency broadcast systems to notify drivers in real time.

[0069] Alternatively, detour routes and speed limit information can be pushed directly to the vehicle driver through a navigation system such as an in-vehicle GPS or a mobile phone navigation application.

[0070] The traffic guidance unit 40 continuously receives real-time updated data from the warning decision unit 30, and adjusts the information push strategy according to the wind and sand trajectory and traffic flow changes to ensure the timeliness and accuracy of the emergency plan.

[0071] The data management unit 50 of the system is the core data hub of the system. It is responsible for receiving, storing, processing and analyzing remote sensing monitoring data, wind and sand impact prediction results and historical data of traffic emergency plans, as well as implementation effect feedback data. It uses machine learning technology to optimize the path planning and speed limit calculation of the optimization algorithm in the early warning decision unit 30, and provides continuous improvement support for the optimization algorithm of the early warning decision unit 30.

[0072] Specifically, historical wind and sand concentration distribution data is received from the remote sensing monitoring unit 10, and actual traffic emergency plans, including generated detour routes and speed limit measures, are received from the early warning decision-making unit 30. Furthermore, historical emergency plans and feedback data on their implementation effectiveness, including traffic efficiency and safety evaluations, are obtained from traffic management departments or traffic monitoring equipment. This data is stored in the database of the data management unit 50 by time and region index. Furthermore, in some embodiments, a multi-level data index structure can be established to facilitate searching. For example, a two-level index structure can be constructed in the data management unit 50, with the first-level index being time period (e.g., hour, day) and the second-level index being region (e.g., city, road section).

[0073] The data management unit 50 stores historical emergency plans and their implementation results, including detour routes, speed limits, travel times, and safety. By comparing historical emergency plans with actual results, it extracts information on the road network capacity improvement rate, detour time increase rate, and the impact of wind and sand concentration thresholds on traffic flow, which all influence emergency efficiency. These indicators are quantitatively analyzed to train optimization algorithms for route planning and speed limit adjustment. The goal of this optimization is to ensure that the route planning and speed limit strategies of the early warning decision unit 30 improve overall traffic efficiency while maintaining the same or higher safety requirements.

[0074] Assume that the historical data stored by data management unit 50 for a particular road includes: sand concentration, wind speed, assessed road risk level, travel time for planned detour routes, average travel time for vehicles actually traveling along the detour routes, and changes in accident probability after the implementation of speed limits. When using machine learning optimization techniques, these data are labeled as input features and target variables for optimization, where the input features are sand concentration, wind speed, and road risk level, and the target variables are travel time and safety indicators (accident probability).

[0075] In some embodiments, the data management unit 50 uses the input features and target variables of the recorded historical data to train a deep learning model to predict the time cost and safety score of the detour route. The model output is the optimal strategy that minimizes the time cost of the detour route. Furthermore, the model needs to be trained to predict the impact of different speed limits on accident probability and travel efficiency (travel time). In some embodiments, the deep learning model can use regression trees to predict time cost and utilize random forests to analyze the relationship between speed limits and accident rates to optimize speed limit intervals. The trained deep learning model is embedded in the data management unit 50 to process monitored wind and sand concentration and traffic data, dynamically adjusting path planning and speed limit measures.

[0076] Specifically, by recording the system's actual performance, adding new data to the training set, and continuously iterating the model, for example, if it is detected that the detour recommended by the system's early warning decision unit 30 actually takes longer, this result is marked as negative feedback. If the new speed limit strategy significantly reduces the accident rate, it is marked as positive feedback. For example, if in the historical record of a sandstorm disaster, the speed limit range of "40-50 km / h" recommended by the system's early warning decision unit 30 successfully reduced the traffic accident rate in previous sandstorm disasters of the same level, this will be used as a positive example to optimize the above-mentioned deep learning model.

[0077] In summary, it can be understood that through the above process, the data management unit 50 completes the mining of historical data, model training and adjustment of the dynamic optimization algorithm, so that the system can continuously improve the accuracy and efficiency of path planning and speed limit calculation in actual operation. This can not only provide more reliable emergency plans for traffic management departments and drivers, but also significantly improve the practicality and safety of the system through long-term optimization.

[0078] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. The scope of patent protection of the present invention shall be based on the claims. Any equivalent structural changes made using the description and drawings of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A traffic emergency warning management system suitable for severe weather, characterized by: include: The remote sensing monitoring unit acquires remote sensing image data within the monitoring area in real time, extracts the wind-blown sand coverage area and its sand concentration distribution by calculating the spectral difference between the corresponding bands of sand and dust in the remote sensing image and the background objects; uses multi-time series remote sensing images to analyze the spatiotemporal variation characteristics of the sand concentration distribution, calculates the transfer rate and direction of the wind-blown sand, and combines it with a fluid dynamics model to simulate the movement trajectory of the wind-blown sand and its changing trend; The prediction unit integrates the dust concentration distribution data provided by the remote sensing monitoring unit with the terrain and traffic data of the monitoring area into a unified grid map, calculates the average wind-blown sand concentration and wind speed of the grid cells covered by each road based on the overlap ratio between road segments and each grid cell in the grid map, and assesses the impact of wind-blown sand on the traffic conditions and risk level of each road; An early warning decision unit generates a traffic emergency plan based on the wind and sand impact data evaluated by the prediction unit, aligning it with real-time traffic flow data in time and space. The plan selects a road route with the lowest detour time cost as the detour route plan, and a speed limit zone on a high-risk road section through an optimization algorithm. a data management unit that stores and analyzes the remote sensing monitoring data, sandstorm impact prediction results, and historical data on traffic emergency plans, and uses machine learning technology to optimize the path planning and speed limit calculation of the optimization algorithm in the early warning decision-making unit; The traffic guidance unit receives and publishes the traffic emergency plan generated by the early warning decision unit in real time, and pushes the detour route plan and speed limit zone information to the traffic management department and drivers in the monitoring area in real time.

2. A traffic emergency warning management system suitable for severe weather according to claim 1, characterized in that: The remote sensing monitoring unit is configured to acquire remote sensing images of the monitoring area at a frequency according to the period of frequent sandstorms, and to perform correction processing on the remote sensing data after acquiring the images; The corrected remote sensing images are subjected to band fusion, the spectral index of the selected specific band of wind and sand is calculated, and the dust coverage range is extracted by setting the threshold segmentation range.

3. A traffic emergency warning management system suitable for severe weather according to claim 2, characterized in that: The remote sensing monitoring unit is based on remote sensing image data of multiple time periods, and compares the position changes of the wind and sand area in different time periods to invert the movement trajectory and flow trend of the wind and sand; specifically: Perform geometric correction on remote sensing images at different times to ensure the consistency of pixel positions; Using a change detection algorithm to identify changes in dust areas at different times, the dust areas at different times are superimposed, the movement direction and speed are calculated based on the time intervals, and a dynamic trend map of wind and sand is obtained by analyzing the dust movement trajectory; The wind speed is calculated based on the movement rate of the wind and sand area, the concentration changes are calibrated using spectral characteristics and ground reference points, and a sand concentration inversion model is established to obtain the concentration distribution of the wind and sand.

4. A traffic emergency warning management system suitable for severe weather according to claim 1, characterized in that: The prediction unit receives the wind and sand coverage area, wind speed and movement trajectory data provided by the remote sensing monitoring unit, and obtains the terrain data and traffic basic data of the monitoring area, projects the remote sensing data, terrain data and traffic data into a grid map with unified coordinates, simulates the diffusion process of wind and sand particles through a fluid dynamics model, and combines the terrain conditions and wind speed and direction characteristics to predict the spatial distribution of wind and sand in different time periods.

5. A traffic emergency warning management system suitable for severe weather according to claim 4, characterized in that: The prediction unit superimposes the wind and sand diffusion range with the traffic network, determines the affected road sections through spatial intersection analysis, evaluates the impact on road traffic based on wind and sand concentration, wind speed and terrain conditions, including the location of the road section and the type of impact, and sets impact thresholds to divide the sections in the monitoring area into high-risk, medium-risk and low-risk routes.

6. A traffic emergency warning management system suitable for severe weather according to claim 5, characterized in that: The prediction unit integrates the wind and sand movement trajectory, the topography and traffic data of the monitoring area to provide a wind and sand concentration distribution forecast map for several time periods in the future, and outputs a traffic risk level distribution map, marking high-risk sections and alternative routes.

7. The traffic emergency warning management system suitable for severe weather according to claim 1, characterized in that: The early warning decision unit integrates the sandstorm impact assessment results provided by the prediction unit with real-time traffic flow data to provide basic data for dynamic planning; specifically: Align wind and sand impact data with real-time traffic flow data in time and space, quantify the data uniformly based on road risk level and capacity, and construct a comprehensive assessment matrix; An optimization algorithm is used to calculate the path with the minimum detour time cost in the wind and sand covered area as the detour path plan; wherein, a list of high-risk road sections that are impassable and have speed limits is extracted from the evaluation results of the prediction unit, and a set of candidate routes for bypassing the high-risk routes is generated based on the topological structure of the traffic road network in the monitoring area. The optimization algorithm is used to select the path with the minimum overall travel time cost from the candidate routes as the detour path plan, and the safe driving speed threshold is calculated based on the wind and sand concentration, wind speed and road section attributes, so as to set the speed limit interval.

8. A traffic emergency warning management system suitable for severe weather according to claim 7, characterized in that: The decision-making unit receiving the warning integrates the detour route, speed limit measures and real-time traffic flow information to generate a traffic emergency plan for the entire area, and sends the traffic emergency plan to the traffic guidance unit, which transmits it to the traffic management department and drivers traveling in the monitoring area.

9. A traffic emergency warning management system suitable for severe weather according to any one of claims 1 to 8, characterized in that: The data management unit receives and stores the remote sensing image data and wind and sand concentration distribution results of the remote sensing monitoring unit, the road risk assessment results of the prediction unit, and the emergency plan generated by the early warning decision unit. The data management unit indexes the data by time and region and builds a case library of historical emergency plans and implementation effects; The historical emergency plans were compared and analyzed with the actual implementation results to extract the impact indicators. Based on the case library, machine learning technology was used to optimize the path planning and speed limit algorithms in the early warning decision unit.

Images