Traffic flow intelligent sensing scheduling method and system under special meteorological conditions

Abstract

The invention provides a traffic flow intelligent sensing scheduling method and system under a special meteorological condition, and relates to the technical field of traffic management, and the method comprises the steps: carrying out the conjoint analysis of a historical traffic data set and a historical special meteorological data set, and constructing a meteorological-traffic correlation model; connecting a meteorological system to obtain real-time special meteorological data; inputting the real-time special meteorological data into the meteorological-traffic association model to obtain predicted traffic data of the next time period, and generating an initial traffic flow scheduling scheme; constructing a traffic accident thermodynamic diagram through traffic simulation; and optimizing the initial traffic flow scheduling scheme according to the traffic accident thermodynamic diagram to obtain a traffic flow scheduling optimization scheme. Through the traffic flow scheduling method and device, the technical problem of poor traffic flow scheduling efficiency caused by incomplete risk identification under the special meteorological condition in the prior art can be solved, the traffic risk under the special meteorological condition is comprehensively identified through joint analysis of the traffic flow data and the special meteorological data, and the traffic flow scheduling efficiency is improved.

Description

Technical Field

[0001] This application relates to the technical field of traffic management, and particularly to an intelligent traffic flow perception and scheduling method and system under special meteorological conditions. Background Art

[0002] With the development of intelligent transportation systems, more and more traffic management methods have introduced devices such as sensors, cameras, radars, and traffic flow perception modules, and combined with technologies such as big data analysis and artificial intelligence algorithms to form an automated and intelligent traffic flow perception and scheduling system. Existing traffic flow perception and scheduling methods mainly rely on real-time or historical traffic flow data for traffic flow prediction and scheduling, and optimize traffic management measures such as traffic signal control and speed limits through real-time data analysis. However, existing traffic flow perception and scheduling methods have significant deficiencies under special meteorological conditions. Usually based on conventional traffic flow patterns and static traffic management requirements, they cannot flexibly respond to sudden changes under special meteorological conditions. Under special meteorological conditions, the impact of meteorological factors on traffic flow is extremely complex. Existing methods have not effectively deeply integrated meteorological data with traffic data, and cannot comprehensively identify potential risks brought about by changes in meteorological conditions, resulting in inaccurate traffic flow prediction during sudden weather events, inability to timely adjust scheduling strategies, increasing the risk of traffic accidents, and often leading to traffic congestion.

[0003] In summary, there is a technical problem in the prior art that the traffic flow scheduling efficiency is poor due to incomplete risk identification under special meteorological conditions. Summary of the Invention

[0004] The purpose of this application is to provide an intelligent traffic flow perception and scheduling method and system under special meteorological conditions to solve the technical problem in the prior art that the traffic flow scheduling efficiency is poor due to incomplete risk identification under special meteorological conditions.

[0005] In view of the above problems, this application provides an intelligent traffic flow perception and scheduling method and system under special meteorological conditions.

[0006] In a first aspect, the present application provides a method for intelligent perception and scheduling of traffic flow under special meteorological conditions. The method for intelligent perception and scheduling of traffic flow under special meteorological conditions is implemented through an intelligent perception and scheduling system for traffic flow under special meteorological conditions. Among them, the method for intelligent perception and scheduling of traffic flow under special meteorological conditions includes: extracting a historical traffic data set within a preset time period from a traffic flow perception module, and at the same time obtaining a historical special meteorological data set for the same time period, jointly analyzing the historical traffic data set and the historical special meteorological data set, and constructing a meteorological-traffic correlation model; connecting to a meteorological system to obtain real-time special meteorological data; inputting the real-time special meteorological data into the meteorological-traffic correlation model to obtain predicted traffic data for the next time period, and generating an initial traffic flow scheduling plan; collecting real-time traffic flow data through the traffic flow perception module, combining with the real-time special meteorological data for traffic simulation, and constructing a traffic accident heat map; performing hierarchical analysis on the traffic accident heat map to obtain multiple regional risk levels, and optimizing the initial traffic flow scheduling plan to obtain an optimized traffic flow scheduling plan.

[0007] Optionally, based on the geographical environment and historical traffic flow of the target area, the target area is divided into multiple sub-areas; according to the historical traffic flow, multiple sensor layout plans for the multiple sub-areas are determined; multiple sensor arrays are arranged in the multiple sub-areas according to the multiple sensor layout plans to complete the construction of the traffic flow perception module.

[0008] Optionally, the historical traffic data set and the historical special meteorological data set are subjected to spatio-temporal alignment processing to obtain a spatio-temporal correlation data set; feature extraction is performed based on the historical traffic data set to obtain multiple traffic indicators; the multiple traffic indicators are traversed, and correlation analysis is performed in the spatio-temporal correlation data set to obtain multiple correlation coefficients, and a correlation coefficient matrix is determined; according to the correlation coefficient matrix, the meteorological-traffic correlation model is constructed.

[0009] Optionally, step a: based on the multiple traffic indicators, obtain a first traffic indicator, and extract multiple first meteorological variables with a correlation coefficient greater than a preset threshold in the correlation coefficient matrix; step b: based on the first traffic indicator and the multiple first meteorological variables, fit and construct a linear regression model; step c: perform least squares method calculation on the linear regression model to obtain multiple regression coefficients, construct a first correlation model, and add it to the meteorological-traffic correlation model; for the multiple traffic indicators, repeat steps a to c to complete the construction of the meteorological-traffic correlation model.

[0010] Optionally, obtain historical traffic accident data and establish a traffic flow simulation model; according to the multiple sub-regions, integrate the real-time traffic flow data and the real-time special meteorological data into multiple regional data corresponding to the multiple sub-regions; input the multiple regional data into the traffic flow simulation model for traffic simulation, evaluate the traffic simulation results, and obtain multiple traffic risk indices for the multiple sub-regions; generate the traffic accident heat map based on the multiple traffic risk indices.

[0011] Optionally, extract the vehicle driving information in the historical traffic accident data and establish a first simulation model using the ABM model; extract the average vehicle speed, traffic density, and traffic flow in the historical traffic accident data and establish a second simulation model using the LWR model; extract the regional traffic information and road network information in the historical traffic accident data and establish a third simulation model using the DTA model; integrate the first simulation model, the second simulation model, and the third simulation model to obtain an initial simulation model; optimize the initial simulation model based on the real-time traffic flow data to obtain the traffic flow simulation model.

[0012] Optionally, sort the multiple traffic risk indices in the traffic accident heat map in ascending order to obtain a traffic risk sequence; perform hierarchical processing on the traffic risk sequence through cluster analysis to obtain multiple risk level thresholds; determine the multiple regional risk levels corresponding to the multiple traffic risk indices based on the multiple risk level thresholds; determine the regional optimization target according to the multiple regional risk levels; mutate the initial traffic flow scheduling plan based on the regional optimization target to obtain a traffic flow scheduling optimization space; perform traffic simulation on the traffic flow scheduling optimization space through the traffic flow simulation model, and perform optimization analysis on the simulation results to obtain the traffic flow scheduling optimization plan.

[0013] In a second aspect, the present application also provides an intelligent traffic flow perception and scheduling system under special meteorological conditions for implementing the intelligent traffic flow perception and scheduling method under special meteorological conditions as described in the first aspect. The intelligent traffic flow perception and scheduling system under special meteorological conditions includes: an association model construction module for extracting a historical traffic data set within a preset time period from a traffic flow perception module, and simultaneously obtaining a historical special meteorological data set for the same time period, and performing joint analysis on the historical traffic data set and the historical special meteorological data set to construct a meteorology-traffic association model; a meteorological data acquisition module for connecting to a meteorological system to obtain real-time special meteorological data; an initial plan generation module for inputting the real-time special meteorological data into the meteorology-traffic association model to obtain predicted traffic data for the next time period and generating an initial traffic flow scheduling plan; a traffic simulation module for collecting real-time traffic flow data through the traffic flow perception module, performing traffic simulation in combination with the real-time special meteorological data, and constructing a traffic accident heat map; and a scheduling plan determination module for performing hierarchical analysis on the traffic accident heat map to obtain multiple regional risk levels, and optimizing the initial traffic flow scheduling plan to obtain an optimized traffic flow scheduling plan.

[0014] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0015] By extracting a historical traffic data set within a preset time period from a traffic flow perception module, and simultaneously obtaining a historical special meteorological data set for the same time period, performing joint analysis on the historical traffic data set and the historical special meteorological data set to construct a meteorology-traffic association model; connecting to a meteorological system to obtain real-time special meteorological data; inputting the real-time special meteorological data into the meteorology-traffic association model to obtain predicted traffic data for the next time period and generating an initial traffic flow scheduling plan; collecting real-time traffic flow data through the traffic flow perception module, performing traffic simulation in combination with the real-time special meteorological data, and constructing a traffic accident heat map; performing hierarchical analysis on the traffic accident heat map to obtain multiple regional risk levels, and optimizing the initial traffic flow scheduling plan to obtain an optimized traffic flow scheduling plan. That is to say, by jointly analyzing traffic flow data and special meteorological data, the relationship between meteorology and traffic flow is deeply identified, the traffic flow for the next time period is predicted according to the association model, and an initial traffic flow scheduling plan is generated accordingly. Traffic simulation is performed in combination with real-time traffic flow data to generate a traffic accident heat map, the risk areas where traffic accidents may occur under adverse meteorological conditions are predicted in advance, and the initial traffic flow scheduling plan is dynamically adjusted, improving the accuracy and efficiency of traffic flow scheduling, and thus improving traffic safety.

[0016] The above description is only an overview of the technical solution of this application. In order to understand the technical means of this application more clearly, it is implemented in accordance with the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the following specific embodiments of this application are specifically given. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of this application, nor is it used to limit the scope of this application. Other features of this application will become easily understandable through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in this application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only exemplary, and for those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0018] Figure 1 It is a schematic flow chart of the intelligent perception and scheduling method for traffic flow under special meteorological conditions of this application;

[0019] Figure 2 It is a schematic structural diagram of the intelligent perception and scheduling system for traffic flow under special meteorological conditions of this application.

[0020] Description of reference numerals: The associated model construction module 11, the meteorological data acquisition module 12, the initial scheme generation module 13, the traffic simulation module 14, the scheduling scheme determination module 15. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] By providing an intelligent perception and scheduling method and system for traffic flow under special meteorological conditions, this application solves the technical problem in the prior art that due to the incomplete risk identification under special meteorological conditions, the traffic flow scheduling efficiency is relatively poor. By jointly analyzing traffic flow data and special meteorological data, the relationship between meteorology and traffic flow is deeply identified, the traffic flow in the next time period is predicted according to the associated model, and an initial traffic flow scheduling scheme is generated accordingly. Combining real-time traffic flow data for traffic simulation, a traffic accident heat map is generated to predict in advance the risk areas where traffic accidents may occur under adverse meteorological conditions, and the initial traffic flow scheduling scheme is dynamically adjusted, improving the accuracy and efficiency of traffic flow scheduling, thereby improving traffic safety.

[0022] Next, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited by the example embodiments described herein. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application. Additionally, it should be noted that for the sake of description, only the parts related to the present application rather than all are shown in the accompanying drawings.

[0023] Embodiment 1. Please refer to the attached Figure 1 , the present application provides a traffic flow intelligent perception and scheduling method under special meteorological conditions. Among them, the traffic flow intelligent perception and scheduling method under special meteorological conditions is applied to a traffic flow intelligent perception and scheduling system under special meteorological conditions. The traffic flow intelligent perception and scheduling method under special meteorological conditions specifically includes the following steps:

[0024] S100: Extract the historical traffic data set within a preset time period from the traffic flow perception module, and at the same time obtain the historical special meteorological data set for the same time period. Jointly analyze the historical traffic data set and the historical special meteorological data set to construct a meteorology-traffic correlation model.

[0025] Specifically, monitor the target area through the traffic flow perception module to extract the historical traffic data set within a preset time period. The preset time period is the time length determined according to the needs and current characteristics of the target area, usually set relatively long to ensure comprehensive sample data, thereby improving the accuracy of subsequent predictions. If there are relatively large changes in the target area at a certain moment, such as updates to the road and its surroundings, such as adding the number of lanes, improving road traffic signs, strengthening traffic control, etc. Since these changes will significantly affect the traffic flow pattern, the old data may no longer be applicable to the new traffic environment, then extract the historical data starting from the time point of the change. If the data samples are few, perform data augmentation on the sample data to expand the existing data set. New time period data can be synthesized from the data at multiple time points to simulate the traffic flow changes at different time points. Fill the data gaps through interpolation methods (such as linear interpolation, spline interpolation, etc.) to generate new sample data.

[0026] Meanwhile, obtain special meteorological condition data within the same time period, such as precipitation, wind speed, temperature, visibility, etc. Special meteorological conditions usually refer to those abnormal meteorological conditions that have a significant impact on traffic flow and daily activities, including but not limited to strong winds, heavy rain, heavy snow, fog, hail, sandstorms, extreme high temperatures, extreme low temperatures, lightning, tornadoes, etc. The historical special meteorological data set usually includes various meteorological factors related to special meteorological conditions, such as precipitation, wind speed, temperature, visibility, etc. Perform spatio-temporal alignment processing on the historical traffic data set and the historical special meteorological data set, align data from different sources (such as traffic data and meteorological data) to the same time and space scale, so that they can be effectively compared and analyzed. Spatio-temporal alignment processing ensures that the timestamps and spatial positions of different data sources are consistent, thus obtaining a unified data format and structure.

[0027] Conduct a joint analysis of the historical traffic data set and the historical special meteorological data set with the aim of finding the relationship between meteorological factors and traffic flow. Combine traffic flow data and meteorological data, and consider the variables in both data sets simultaneously to analyze their mutual influence and dependence. Specifically, determine the meteorological variables that are strongly correlated with each traffic indicator, construct a meteorology-traffic correlation model, and predict the changes in traffic flow under different meteorological conditions. By constructing a meteorology-traffic correlation model, it is possible to predict the changes in traffic flow based on meteorological data, which not only improves the prediction accuracy of traffic flow but also effectively supports the decision-making of the traffic management system under different meteorological conditions.

[0028] S200: Connect to the meteorological system to obtain real-time special meteorological data.

[0029] Specifically, establish a connection between the traffic dispatching system and the meteorological system through the network or other technical means, and achieve it through a data interface or API (Application Programming Interface). The meteorological system is a system responsible for collecting, monitoring, and providing meteorological data, usually including weather stations, satellite observations, radars, meteorological models, etc. Meteorological service providers (such as OpenWeather) usually provide API interfaces that allow users to obtain real-time meteorological data through HTTP requests. Some meteorological systems may provide real-time meteorological data to the outside through proprietary protocols or data flows. In this way, the traffic dispatching system can directly access and obtain real-time meteorological data.

[0030] Through the connection with the meteorological system, obtain real-time special meteorological data, including precipitation, air temperature, wind speed, temperature, and other special meteorological factors, etc., to reflect the current meteorological conditions. By obtaining meteorological data in real time, the traffic dispatching system can quickly respond to weather changes and timely adjust traffic management strategies, which helps to improve the accuracy of traffic prediction.

[0031] S300: Input the real-time special meteorological data into the meteorology-traffic correlation model to obtain the predicted traffic data for the next time period and generate an initial traffic flow scheduling plan.

[0032] Specifically, the real-time special meteorological data can reflect the current climate conditions. Inputting the real-time special meteorological data into the meteorology-traffic correlation model can predict the traffic conditions in the future for a period of time, including traffic flow, vehicle speed, traffic density, etc. The model will predict the change in traffic flow in the future time period. For example, if a large amount of precipitation is predicted, the model will predict a decrease in traffic flow, especially in traffic-intensive areas or on highways.

[0033] Based on the predicted traffic data, traffic management measures are proposed, such as adjusting traffic lights, implementing traffic control or guidance, adjusting speed limits, adjusting traffic routes, etc., to generate an initial traffic flow scheduling plan. For example, when the meteorological system detects that the current wind speed on a certain highway section is relatively high and will continue to increase, the meteorology-traffic correlation model predicts that there will be a strong wind phenomenon, resulting in a decrease in vehicle speed. An initial scheduling plan is generated, suggesting implementing traffic control in this area, arranging the vehicles on the highway section to get off the highway nearby, and at the same time preventing other vehicles from entering this section to reduce the risk of traffic accidents caused by the influence of wind speed.

[0034] Through the real-time input of the meteorology-traffic correlation model, the changing trend of traffic flow under different meteorological conditions is predicted, especially the prediction effect is more prominent in special meteorological situations, improving the intelligent level of urban traffic management.

[0035] S400: Collect real-time traffic flow data through the traffic flow perception module, and combine the real-time special meteorological data for traffic simulation to construct a traffic accident heat map.

[0036] Furthermore, S400 of this application includes:

[0037] Obtain historical traffic accident data and establish a traffic flow simulation model; according to the multiple sub-regions, integrate the real-time traffic flow data and the real-time special meteorological data into multiple region data corresponding to the multiple sub-regions; input the multiple region data into the traffic flow simulation model for traffic simulation, evaluate the traffic simulation results, and obtain multiple traffic risk indices for the multiple sub-regions; based on the multiple traffic risk indices, generate the traffic accident heat map.

[0038] Specifically, obtain historical traffic accident data, that is, data related to traffic accidents that occurred in the past, including the time and location of the accident, the type of accident, vehicle information involved (such as vehicle speed, lane, vehicle type, etc.), accident results (such as injury conditions, vehicle damage degree, etc.), and weather conditions. Based on the collected data, construct a traffic flow simulation model. The traffic flow simulation model needs to be able to simulate the behavior of traffic flow in different regions and on different road sections, as well as how the movement of a single vehicle affects the overall traffic flow. And fine-tune the simulation model through real-time traffic flow data to make it better adapt to the current traffic conditions.

[0039] Integrate and process the traffic flow and meteorological data in the target area (such as a certain area of the city, a part of the road network) to form a comprehensive data set reflecting the traffic and meteorological conditions in this area. According to the geographical information of the multiple sub-regions after division, respectively integrate the corresponding real-time traffic flow data and meteorological data of this area to obtain multiple regional data corresponding to the multiple sub-regions, that is, a comprehensive data set reflecting the traffic and meteorological conditions in this area.

[0040] Use the traffic flow simulation model, input the integrated regional data, conduct traffic flow prediction, and simulate the changes in traffic flow in each sub-region. Based on the results of the simulation model, combined with historical accident data and changes in traffic flow, evaluate the traffic safety risks of each region. Input the integrated regional data into the traffic flow simulation model, simulate the changes in traffic flow under different weather conditions, and obtain the traffic conditions in each sub-region for a period of time in the future. Evaluate the traffic simulation results and calculate the traffic risk index of each sub-region.

[0041] Evaluate the traffic risks of each region according to multiple factors, including vehicle speed, traffic density, real-time meteorological factors, historical accident occurrence frequency, etc., to obtain multiple traffic risk indexes for multiple sub-regions. Specifically, according to the importance of each evaluation index to traffic risk, assign corresponding weights, and perform weighted calculations on each index to obtain the traffic risk index.

[0042] Visualize the traffic risk indices of multiple sub-regions in a visual form, and use a heat map to display the regions with different risk levels. The risk level of different regions is represented by the depth of color. The darker the color, the higher the risk. Usually, red represents high-risk areas, and green represents low-risk areas. Through the traffic accident heat map, the risk distribution of different regions can be intuitively displayed, helping traffic management personnel quickly identify potential high-risk road sections. The traffic accident heat map is a visualization tool for showing the traffic risk levels of different regions. Through the form of the heat map, the traffic accident risk situation of each region can be intuitively seen. Through real-time traffic risk assessment, quickly identify the traffic risks of different regions, especially under special meteorological conditions, accurately predict the potential risks of traffic flow changes and accidents, provide the dynamic changes of traffic flow, and provide a more scientific and accurate basis for traffic management and dispatching.

[0043] Furthermore, the present application further includes the following steps:

[0044] Extract the vehicle driving information from the historical traffic accident data and establish a first simulation model using the ABM model; extract the average vehicle speed, traffic density, and traffic flow from the historical traffic accident data and establish a second simulation model using the LWR model; extract the regional traffic information and road network information from the historical traffic accident data and establish a third simulation model using the DTA model; integrate the first simulation model, the second simulation model, and the third simulation model to obtain an initial simulation model; optimize the initial simulation model based on the real-time traffic flow data to obtain the traffic flow simulation model.

[0045] Specifically, obtain historical traffic accident data, including the time, location, vehicle types involved, traffic conditions, etc. of the accident. Extract the vehicle driving information from the historical traffic accident data, including the driving speed, vehicle spacing, lane selection, acceleration and deceleration behaviors, etc. of each vehicle, which reflects the microscopic behaviors of each vehicle in the traffic flow and its interaction with surrounding vehicles. The ABM model is an individual-based modeling method that studies the dynamic changes of the overall system by simulating the behaviors and interactions of individuals (agents). Through the ABM model, simulate the driving behaviors and interactions of each vehicle (agent). In the model, each vehicle has attributes such as speed and acceleration, as well as behavioral rules for obeying traffic rules. Through simulation, find out which factors (such as traffic signal settings, traffic volume, etc.) will cause congestion and propose optimization solutions. Establish a first simulation model by setting the behavioral rules of vehicles (such as maintaining a certain vehicle spacing, decelerating when encountering congestion, etc.) to simulate the mutual influence between vehicles. Through this simulation, potential traffic bottlenecks and accident-prone points can be identified.

[0046] Extract the average vehicle speed, traffic density, and traffic flow from historical traffic accident data. The vehicle speed represents the average driving speed of vehicles. Traffic density is the number of vehicles on a unit length of road. Traffic flow is the number of vehicles passing through a certain section within a unit time. The LWR model is a classic macroscopic traffic flow model that describes the relationship between vehicle flow, density, and speed, focusing on the characteristics of the overall traffic flow rather than the behavior of individual vehicles. It is commonly used for macroscopic simulation and prediction of traffic flow. By deriving the vehicle speed from the traffic density and further calculating the flow, it is very suitable for describing the overall traffic behavior on large-scale sections. Extract data such as average vehicle speed, traffic density, and traffic flow from historical traffic accident data, and then use these data to establish a second simulation model to describe the average behavior of traffic flow.

[0047] Extract the regional traffic information and road network information from historical traffic accident data. Regional traffic information refers to the traffic conditions within a specific region, such as traffic flow, average vehicle speed, accident rate, etc. of a section. Road network information refers to information such as sections, intersections, traffic lights, etc. in the traffic network. The structure of the road network will directly affect the distribution and change of traffic flow. The DTA model (Dynamic Traffic Assignment model) simulates the flow changes of vehicles on different sections and regions at different time periods, analyzes the dynamic distribution of traffic flow on the network, taking into account the topological structure of the road network, traffic signals, and changes in traffic flow. Extract regional traffic information (such as traffic flow, vehicle speed, etc. in different regions) and road network information (such as section length, number of intersections, signal control, etc.) from historical traffic accident data. Use the DTA model to simulate the flow of traffic flow between different regions and different sections. The DTA model is usually based on traffic assignment algorithms (such as the shortest path algorithm, traffic assignment algorithm, etc.) to simulate the traffic flow state at different time points and different regions. The DTA model can consider the structure of the traffic network, the mutual connection between sections, and the time-varying characteristics of traffic flow.

[0048] Integrate the first simulation model, the second simulation model, and the third simulation model to obtain a complete initial simulation model. Through this integrated simulation model, it is possible to simultaneously consider the influence of individual vehicle behavior, macroscopic traffic flow characteristics, and road network structure on traffic flow, and comprehensively reflect the dynamic changes of traffic flow. The initial simulation model covers the microscopic, macroscopic, and network-level information of traffic flow and can perform comprehensive traffic flow simulation.

[0049] Input the obtained real-time traffic flow data into the initial simulation model, and continuously optimize the model through a feedback mechanism to make the model more in line with the current traffic conditions and improve its prediction and scheduling capabilities. That is to say, according to the real-time traffic flow data, adjust the parameters of the simulation model so that the output results of the model are more consistent with the changes in the actual traffic flow. The purpose of optimization is to improve the prediction accuracy and response speed of the model. The real-time traffic flow data dynamically adjusts the model parameters through comparison with the model output to improve the adaptability and prediction ability of the model. Compare the real-time traffic flow data with the prediction results of the initial simulation model, and analyze the differences between the model output and the actual traffic flow. According to the comparison results, use appropriate optimization algorithms (such as the least squares method, particle swarm optimization, etc.) to adjust the parameters of the initial model. For example, it may be necessary to adjust the flow-speed relationship in the LWR model or optimize the traffic flow distribution method of the road network in the DTA model.

[0050] By integrating microscopic, macroscopic, and network-level simulation models, it is possible to comprehensively simulate traffic behaviors at different levels. The input of real-time traffic flow data enables the simulation model to be dynamically optimized, respond in a timely manner to changes in different traffic flow conditions, and thus formulate traffic flow prediction and management measures to reduce congestion, lower the accident rate, and improve traffic safety and smoothness.

[0051] Furthermore, step S100 of this application includes:

[0052] Based on the geographical environment and historical traffic flow of the target area, divide the target area into multiple sub-areas; according to the historical traffic flow, determine multiple sensor layout schemes for the multiple sub-areas; deploy multiple sensor arrays in the multiple sub-areas according to the multiple sensor layout schemes to complete the construction of the traffic flow perception module.

[0053] Specifically, determine the target area, that is, the specific area where traffic flow scheduling is required, usually a city, road network, or specific road section, and it is necessary to monitor and optimize traffic flow, traffic safety, etc. Obtain the geographical environment of the target area, that is, the geographical features of the target area, including road types (such as highways, urban roads, ring roads), road widths, the number of intersections, traffic signs, traffic flow directions, and high-incidence areas of traffic accidents, etc. Analyze the road distribution, intersections, traffic characteristics, etc. of the target area through map data and satellite images. If the target area is a city, different road types (such as main roads, branch roads, ring roads, etc.) and traffic flow fluctuations are usually considered. For example, during peak hours, congestion may occur at some intersections, so these areas may need to be divided into independent sub-areas for monitoring.

[0054] Meanwhile, obtain the historical traffic flow of the target area, that is, the traffic data over a past period, including the number of vehicles, vehicle speed, road congestion conditions, traffic density, etc. The historical traffic flow can reflect the common patterns and seasonal fluctuations of traffic flow, which helps to predict the future change trend of traffic flow. By analyzing the past traffic flow data of the target area, identify the high-flow and low-flow periods or areas. For example, during the morning rush hour, the central area of the city may have a higher traffic flow, while the suburbs may have a lower flow. Based on these data, the target area can be divided into multiple sub-areas, and each sub-area may have different traffic flow characteristics, which is convenient for targeted management and monitoring.

[0055] According to the geographical characteristics, different distributions of traffic flow, and traffic management requirements, divide the target area into multiple sub-areas, and there are significant differences in the traffic conditions of different sub-areas. By analyzing the historical traffic flow data, determine the traffic flow density of each sub-area, so as to determine the sensor layout plan for each sub-area, including the specific layout positions, quantities, and types of sensors, to ensure that the sensors can monitor important traffic data in appropriate places and completely cover the entire target area. For example, in busy traffic areas, more sensors may need to be installed to better monitor traffic flow and vehicle speed. For areas with less traffic flow, the number of sensors can be reduced. According to the traffic characteristics of each sub-area, select appropriate types of sensors, such as geomagnetic sensors, radar sensors, video surveillance, infrared sensors, etc. In areas with large traffic flow, geomagnetic sensors or radar sensors may need to be installed to perceive the passing state of vehicles in real time, while at multiple intersections, video surveillance may be combined to identify the traffic flow situation. Based on the characteristics of the sub-area and the main distribution area of traffic flow, reasonably determine the layout positions of sensors.

[0056] According to the sensor layout plan of each sub-area, deploy a sensor array in each sub-area. Each sensor array is composed of multiple sensors combined in a specific spatial layout to ensure that traffic flow, vehicle speed, lane occupancy, etc. data of the sub-area can be comprehensively collected. Install the sensors at the planned positions and conduct wiring for power and data transmission according to the design requirements of the traffic flow perception module. Each sensor array needs to be docked with the data acquisition and transmission module in the traffic flow perception module to ensure that the data can be collected in real time and transmitted to the central processing system for analysis. By analyzing the geographical environment and historical traffic flow of the target area, dividing sub-areas and designing sensor layout plans based on this information, and finally deploying sensor arrays in each sub-area, the monitoring accuracy and scheduling efficiency of traffic flow are significantly improved.

[0057] Furthermore, the present application further includes the following steps:

[0058] The historical traffic data set and the historical special meteorological data set are subjected to spatiotemporal alignment processing to obtain a spatiotemporal correlation data set; feature extraction is performed based on the historical traffic data set to obtain multiple traffic indicators; the multiple traffic indicators are traversed, correlation analysis is performed in the spatiotemporal correlation data set to obtain multiple correlation coefficients, and a correlation coefficient matrix is ​​determined; and the meteorological-traffic correlation model is constructed according to the correlation coefficient matrix.

[0059] Specifically, the historical traffic data set and the historical special meteorological data set are spatiotemporally aligned to obtain a spatiotemporally correlated data set. The core of the spatiotemporal alignment process is to ensure that the traffic data and meteorological data match in time and space, that is, to ensure that the two data sets are in the same time period and correspond to the same geographical location.

[0060] The time frequencies of historical traffic data and historical meteorological data may be inconsistent, and they need to be adjusted to a unified time scale. Extract the acquisition timestamps of the historical traffic data set and the historical special meteorological data set to determine the time granularity. If the time granularity is different, standardize the two to the same time granularity, usually choose to align them at a smaller time granularity. Sort the two data sets by the acquisition timestamp to ensure the consistency of the time series. If data is missing at a certain time point, interpolate the previous and next data to fill in the data at that time point to ensure that the timestamps of the traffic flow data and the meteorological data are consistent.

[0061] Spatial alignment is to ensure that data at different spatial locations can correspond. In traffic flow analysis, traffic sensing devices (such as road sensors, cameras, etc.) will collect traffic data in certain areas, while meteorological data may be total data within the area or measurement data from a meteorological station. Match the geographic coordinates of traffic data with the geographic coordinates of meteorological data. For example, traffic flow data is usually recorded based on a specific road section or intersection, while meteorological data may be recorded based on a larger area (such as the geographical location of a meteorological station). Therefore, it is necessary to correspond the specific location of traffic data with the area where the meteorological data is located. Meteorological data is collected from a specific meteorological station, and it may be necessary to match the spatial distribution of traffic data based on the geographical location of the meteorological station. For larger cities or regions, meteorological data of the surrounding area can be assigned to the corresponding road section based on the coverage of traffic flow sensing devices. The spatial range of meteorological data is large and may not accurately cover each traffic sensing point. Meteorological data is extrapolated to the specific location of the sensor through spatial interpolation methods. In addition, it is also possible to consider dividing the entire area into gridded areas and then aligning the traffic data and meteorological data in each grid. By assigning the regional meteorological data to these grids, the traffic flow data can be spatially aligned according to the location of the grids.

[0062] Merge the time-aligned and space-aligned datasets to form a spatio-temporal correlation dataset. Each record will include traffic data (such as vehicle speed, traffic density, etc.) and meteorological data (such as temperature, precipitation, etc.) at a specific time point and specific geographical location. Extract features from the historical traffic dataset to determine multiple traffic indicators, such as average vehicle speed, traffic flow, traffic density, congestion index, etc. Feature extraction refers to the process of extracting useful information or indicators from the original data, which can represent data indicators of traffic flow conditions.

[0063] Traverse multiple traffic indicators, perform correlation analysis in the spatio-temporal correlation dataset to obtain multiple correlation coefficients, and determine the correlation coefficient matrix. Measure the linear correlation between traffic flow and meteorological conditions (such as precipitation, temperature, wind speed, etc.) through the Pearson correlation coefficient. The value range of the Pearson correlation coefficient is from -1 to 1, where 1 represents a perfect positive correlation, -1 represents a perfect negative correlation, and 0 represents no correlation. Calculate the correlation between each traffic indicator and meteorological conditions through the formula of the Pearson correlation coefficient to obtain multiple correlation coefficients.

[0064] Based on multiple correlation coefficients, construct a correlation coefficient matrix. The rows in the matrix represent traffic indicators, the columns represent meteorological variables, and each element in the matrix represents the correlation coefficient between different traffic indicators and meteorological conditions. According to the determined correlation coefficient matrix, screen out meteorological features with significant correlations to obtain the set of meteorological variables corresponding to each traffic indicator. Based on this, perform linear fitting to construct multiple linear regression models corresponding to multiple traffic indicators.

[0065] Furthermore, this application also includes the following steps:

[0066] Step a: Based on the multiple traffic indicators, obtain the first traffic indicator, and extract multiple first meteorological variables with correlation coefficients greater than a preset threshold in the correlation coefficient matrix; Step b: Fit based on the first traffic indicator and the multiple first meteorological variables to construct a linear regression model; Step c: Perform least squares calculation on the linear regression model to obtain multiple regression coefficients, construct the first association model, and add it to the meteorological-traffic association model; For the multiple traffic indicators, repeat steps a to c to complete the construction of the meteorological-traffic association model.

[0067] Specifically, one traffic indicator is randomly selected from multiple traffic indicators. According to the first traffic indicator, screening is performed in the correlation coefficient matrix, and all meteorological variables with correlation coefficients greater than a preset threshold are extracted to obtain multiple first meteorological variables. The multiple first meteorological variables refer to meteorological variables that have a significant correlation with the first traffic indicator. The preset threshold is a predefined value set according to specific application scenarios and analysis objectives, used to screen meteorological variables with a strong correlation coefficient with the first traffic indicator, and determines which meteorological variables are considered to have a highly relevant correlation with the traffic indicator.

[0068] According to the first traffic indicator and multiple first meteorological variables, fitting is performed to construct a linear regression model. For example, assuming that the traffic flow is the first traffic indicator, precipitation and wind speed are multiple first meteorological variables, and the constructed linear regression model is Y = a + a1X1 + a2X2 + c, where Y is the traffic flow, X1 is the precipitation, X2 is the wind speed, a is the intercept term, a1 and a2 are regression coefficients, and c is the error term. The least squares method is used to calculate the linear regression model to obtain multiple regression coefficients. Using the least squares method to calculate the regression coefficients determines the influence degree (i.e., the regression coefficients) of each meteorological variable on the traffic flow. The goal of the least squares method is to minimize the sum of squared errors and calculate the optimal regression coefficients. Using multiple regression coefficients to construct the first association model and adding it to the meteorological-traffic association model. For example, through the least squares method, a = 350, which is the intercept term, indicating the value of the traffic flow in the absence of precipitation and wind speed; a1 = -10, indicating how much the traffic flow will decrease when the precipitation increases by 1 millimeter per hour; a2 = -3, indicating how much the traffic flow will decrease when the wind speed increases by 1 kilometer per hour; the error term is 5, ensuring a closer approximation to the actual situation in the case of traffic emergencies or other external factors.

[0069] The above steps are performed for each traffic indicator among multiple traffic indicators, extracting meteorological variables with strong correlations, constructing linear regression models, and calculating regression coefficients to obtain multiple association models corresponding to multiple traffic indicators. All the association models are aggregated into a complete meteorological-traffic association model, which can predict changes in different traffic indicators based on real-time meteorological data. By establishing the relationship between multiple traffic indicators and multiple meteorological variables to construct a meteorological-traffic association model helps to improve the accuracy of traffic flow prediction, especially under special meteorological conditions. Through historical traffic data and meteorological data, analyzing the impact of meteorological changes on traffic flow and constructing a meteorological-traffic association model can accurately predict changes in traffic indicators such as traffic flow and vehicle speed under different meteorological conditions, thereby achieving precise prediction of traffic flow.

[0070] S500: Perform hierarchical analysis on the traffic accident heat map to obtain multiple regional risk levels, and optimize the initial traffic flow scheduling plan to obtain an optimized traffic flow scheduling plan.

[0071] Further, step S500 of the present application includes:

[0072] Sort the multiple traffic risk indices in the traffic accident heat map in ascending order to obtain a traffic risk sequence; perform hierarchical processing on the traffic risk sequence through cluster analysis to obtain multiple risk level thresholds; based on the multiple risk level thresholds, determine the multiple regional risk levels corresponding to the multiple traffic risk indices; according to the multiple regional risk levels, determine the regional optimization objectives; based on the regional optimization objectives, mutate the initial traffic flow scheduling plan to obtain a traffic flow scheduling search space; through the traffic flow simulation model, perform traffic simulation on the traffic flow scheduling search space, and perform optimization analysis on the simulation results to obtain the optimized traffic flow scheduling plan.

[0073] Specifically, layer the different risk regions of the traffic accident heat map, and determine different risk levels (such as high risk, medium risk, low risk, etc.) according to different traffic risk indices. By analyzing the traffic accident heat map, identify the traffic accident risks in different regions, and divide these regions into multiple levels according to the degree of risk. The purpose of hierarchical analysis is to quantify the traffic accident risks and perform regional management.

[0074] Sort the multiple traffic risk indices in the traffic accident heat map in ascending order, arrange the traffic risk indices of different regions in ascending order to obtain a traffic risk sequence. Through a clustering method (such as K-means), divide the sorted traffic risk sequence into several groups, and each group represents a specific risk level. Use a clustering algorithm (such as K-means) to divide the traffic risk sequence into several groups. Each group represents a specific risk level. For example, high-risk regions may be clustered into one group, and low-risk regions may be clustered into another group. According to the clustering results, obtain multiple risk level thresholds. The range of each category in the clustering results is the risk level threshold of that category. The risk level threshold refers to the standard value determined by clustering or other analysis methods on the traffic risk sequence for dividing different risk levels. Through these thresholds, the traffic risk indices can be divided into multiple levels, such as high risk, medium risk, low risk, etc.

[0075] Regions are divided into different risk levels according to the intervals in which the traffic risk index falls, such as high risk, medium risk, low risk, etc. According to the risk level of each region, traffic flow scheduling optimization goals corresponding to each region are set. For example, the optimization goal of a high-risk region may be to reduce the accident rate or improve traffic, the optimization goal of a medium-risk region may be to reduce traffic congestion, optimize the signal light cycle, etc., while the low-risk region may focus on improving road traffic efficiency.

[0076] According to the optimization goal of each region, the initial traffic flow scheduling plan is mutated. By adjusting multiple parameters of the initial traffic flow scheduling plan, a set of different scheduling plans is generated. For example, changing the signal light cycle, adjusting the lane configuration, increasing or decreasing traffic control, etc. Each adjustment generates a new scheduling plan, and these plans together constitute the traffic flow scheduling optimization space. The traffic flow scheduling optimization space is a space composed of multiple possible scheduling plans generated by mutating the initial traffic flow scheduling plan, and the optimal plan is selected from it. Each scheduling plan represents a possible traffic flow state, and the optimization analysis is the process of finding the optimal plan from these plans.

[0077] The different traffic flow scheduling plans in the traffic flow scheduling optimization space are input into the traffic flow simulation model to simulate their impact on the traffic flow, and the simulation results are obtained, including indicators such as vehicle speed, traffic density, accident frequency, etc. According to the simulation results, each plan is evaluated, the effects of different traffic flow scheduling plans are analyzed, and their impact on indicators such as traffic flow, vehicle speed, and congestion is evaluated. According to each indicator, the fitness value of each scheduling plan is determined, and the scheduling plan corresponding to the maximum fitness value is selected as the traffic flow scheduling optimization plan to perform traffic flow scheduling on the target region. The optimization analysis refers to the process of determining the optimal plan by comparing the results of different traffic flow scheduling plans (such as vehicle speed, traffic density, accident frequency, etc.), including analyzing the performance of different plans through the model and selecting the best plan through the optimization algorithm.

[0078] The traffic flow simulation model will simulate the actual traffic flow state according to the input traffic scheduling plan, including parameters such as vehicle speed, traffic density, traffic accidents, and traffic congestion. Through the evaluation of these parameters, that is, by assigning weights according to the importance of each evaluation index and performing weighted calculations on each index, the advantages and disadvantages of each traffic flow scheduling plan are evaluated. Based on the optimization analysis, considering various factors such as vehicle speed, traffic density, accident rate, and congestion, a plan that best meets the goal is selected. Specifically, the fitness of each traffic flow scheduling plan is evaluated, the plans with a fitness less than the preset threshold are screened out, and the plan is fine-tuned, thereby updating the traffic flow scheduling optimization space. The fitness of each scheduling plan in the traffic flow scheduling optimization space is evaluated, including the updated plans, to obtain the fitness value, and the plan with the maximum fitness value is extracted as the traffic flow scheduling optimization plan.

[0079] The traffic flow scheduling optimization plan is the optimal traffic flow scheduling plan selected based on the simulation results and optimization analysis, and is usually the optimal plan comprehensively obtained based on factors such as traffic flow stability, accident reduction, and efficiency improvement. Through the simulation and optimization analysis of the traffic flow simulation model, the best traffic scheduling plan can be effectively identified and selected, improving the road traffic efficiency and reducing traffic congestion.

[0080] In summary, the traffic flow intelligent perception and scheduling method under special meteorological conditions provided by this application has the following technical effects:

[0081] By extracting the historical traffic data set within a preset time period from the traffic flow perception module, and simultaneously obtaining the historical special meteorological data set in the same time period, jointly analyzing the historical traffic data set and the historical special meteorological data set to construct a meteorology-traffic correlation model; connecting to the meteorological system to obtain real-time special meteorological data; inputting the real-time special meteorological data into the meteorology-traffic correlation model to obtain the predicted traffic data for the next time period and generate an initial traffic flow scheduling plan; collecting real-time traffic flow data through the traffic flow perception module, combining with the real-time special meteorological data for traffic simulation to construct a traffic accident heat map; hierarchically analyzing the traffic accident heat map to obtain multiple regional risk levels, and optimizing the initial traffic flow scheduling plan to obtain a traffic flow scheduling optimization plan. That is to say, by jointly analyzing traffic flow data and special meteorological data, deeply identifying the relationship between meteorology and traffic flow, predicting the traffic flow for the next time period according to the correlation model, generating an initial traffic flow scheduling plan based on this, conducting traffic simulation in combination with real-time traffic flow data, generating a traffic accident heat map, predicting in advance the risk areas where traffic accidents may occur under adverse meteorological conditions, and dynamically adjusting the initial traffic flow scheduling plan, improving the accuracy and efficiency of traffic flow scheduling, and thus improving traffic safety.

[0082] Embodiment 2, based on the same inventive concept as the traffic flow intelligent perception and scheduling method under special meteorological conditions in the foregoing Embodiment 1, this application also provides a traffic flow intelligent perception and scheduling system under special meteorological conditions. Please refer to the attached Figure 2 , the traffic flow intelligent perception and scheduling system under special meteorological conditions includes:

[0083] The correlation model construction module 11 is used to extract the historical traffic data set within a preset time period from the traffic flow perception module, and at the same time obtain the historical special meteorological data set in the same time period, perform joint analysis on the historical traffic data set and the historical special meteorological data set, and construct a meteorology-traffic correlation model; the meteorological data acquisition module 12 is used to connect to the meteorological system to obtain real-time special meteorological data; the initial scheme generation module 13 is used to input the real-time special meteorological data into the meteorology-traffic correlation model, obtain the predicted traffic data for the next time period, and generate an initial traffic flow scheduling scheme; the traffic simulation module 14 is used to collect real-time traffic flow data through the traffic flow perception module, combine the real-time special meteorological data for traffic simulation, and construct a traffic accident heat map; the scheduling scheme determination module 15 is used to perform hierarchical analysis on the traffic accident heat map to obtain multiple regional risk levels, and optimize the initial traffic flow scheduling scheme to obtain an optimized traffic flow scheduling scheme.

[0084] Further, the correlation model construction module 11 in the intelligent traffic flow perception and scheduling system under special meteorological conditions is further used for:

[0085] Based on the geographical environment and historical traffic flow of the target area, divide the target area into multiple sub-areas; according to the historical traffic flow, determine multiple sensor layout schemes for the multiple sub-areas; deploy multiple sensor arrays in the multiple sub-areas according to the multiple sensor layout schemes to complete the construction of the traffic flow perception module.

[0086] Further, the correlation model construction module 11 in the intelligent traffic flow perception and scheduling system under special meteorological conditions is further used for:

[0087] Perform spatio-temporal alignment processing on the historical traffic data set and the historical special meteorological data set to obtain a spatio-temporal correlation data set; perform feature extraction based on the historical traffic data set to obtain multiple traffic indicators; traverse the multiple traffic indicators, perform correlation analysis in the spatio-temporal correlation data set to obtain multiple correlation coefficients, and determine a correlation coefficient matrix; construct the meteorology-traffic correlation model according to the correlation coefficient matrix.

[0088] Further, the correlation model construction module 11 in the intelligent traffic flow perception and scheduling system under special meteorological conditions is further used for:

[0089] Step a: Based on the multiple traffic indicators, obtain the first traffic indicator, and extract multiple first meteorological variables in the correlation coefficient matrix whose correlation coefficients are greater than a preset threshold; Step b: Fit based on the first traffic indicator and the multiple first meteorological variables to construct a linear regression model; Step c: Perform least squares method calculation on the linear regression model to obtain multiple regression coefficients, construct a first association model, and add it to the meteorological-traffic association model; For the multiple traffic indicators, repeat Steps a to c to complete the construction of the meteorological-traffic association model.

[0090] Further, the traffic simulation module 14 in the intelligent traffic flow perception and scheduling system under special meteorological conditions is further configured to:

[0091] Obtain historical traffic accident data, and establish a traffic flow simulation model; According to the multiple sub-regions, integrate the real-time traffic flow data and the real-time special meteorological data into multiple regional data corresponding to the multiple sub-regions; Input the multiple regional data into the traffic flow simulation model for traffic simulation, evaluate the traffic simulation results, and obtain multiple traffic risk indexes for the multiple sub-regions; Based on the multiple traffic risk indexes, generate the traffic accident heat map.

[0092] Further, the traffic simulation module 14 in the intelligent traffic flow perception and scheduling system under special meteorological conditions is further configured to:

[0093] Extract the vehicle driving information in the historical traffic accident data, and establish a first simulation model using the ABM model; Extract the average vehicle speed, traffic density, and traffic flow in the historical traffic accident data, and establish a second simulation model using the LWR model; Extract the regional traffic information and road network information in the historical traffic accident data, and establish a third simulation model using the DTA model; Integrate the first simulation model, the second simulation model, and the third simulation model to obtain an initial simulation model; Based on the real-time traffic flow data, optimize the initial simulation model to obtain the traffic flow simulation model.

[0094] Further, the scheduling plan determination module 15 in the intelligent traffic flow perception and scheduling system under special meteorological conditions is further configured to:

[0095] Ascendingly sort the multiple traffic risk indices in the traffic accident heat map to obtain a traffic risk sequence; perform hierarchical processing on the traffic risk sequence through cluster analysis to obtain multiple risk level thresholds; based on the multiple risk level thresholds, determine the multiple regional risk levels corresponding to the multiple traffic risk indices; according to the multiple regional risk levels, determine a regional optimization goal; based on the regional optimization goal, mutate the initial traffic flow scheduling plan to obtain a traffic flow scheduling optimization search space; through the traffic flow simulation model, perform traffic simulation on the traffic flow scheduling optimization search space, and perform optimization analysis on the simulation results to obtain the traffic flow scheduling optimization plan.

[0096] The various embodiments in this specification are described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. The foregoing Figure 1 The traffic flow intelligent perception scheduling method and specific examples under special meteorological conditions in the first embodiment are equally applicable to the traffic flow intelligent perception scheduling system under special meteorological conditions in this embodiment. Through the foregoing detailed description of the traffic flow intelligent perception scheduling method under special meteorological conditions, those skilled in the art can clearly know the traffic flow intelligent perception scheduling system under special meteorological conditions in this embodiment. Therefore, for the sake of simplicity of the specification, it will not be elaborated here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0097] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0098] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is also intended to include these changes and modifications.

Claims

1. The intelligent perception and dispatching method of traffic flow under special meteorological conditions is characterized by: include: Extracting a historical traffic data set within a preset period from the traffic flow perception module, and simultaneously obtaining a historical special meteorological data set for the same period, jointly analyzing the historical traffic data set and the historical special meteorological data set, and constructing a meteorological-traffic correlation model; Connect to the meteorological system to obtain real-time special meteorological data; Inputting the real-time special meteorological data into the meteorological-traffic correlation model to obtain the predicted traffic data for the next period and generate an initial traffic flow scheduling plan; The real-time traffic flow data is collected through the traffic flow perception module, and the traffic simulation is performed in combination with the real-time special meteorological data to construct a traffic accident heat map; The traffic accident heat map is analyzed in layers to obtain multiple regional risk levels, and the initial traffic flow scheduling plan is optimized to obtain a traffic flow scheduling optimization plan.

2. The intelligent perception and dispatching method for traffic flow under special meteorological conditions according to claim 1 is characterized in that: The steps of constructing the traffic flow perception module include: Dividing the target area into a plurality of sub-areas based on the geographical environment and historical traffic flow of the target area; Determining, based on the historical traffic flow, a plurality of sensor deployment schemes for the plurality of sub-areas; According to the multiple sensor deployment schemes, multiple sensor arrays are deployed in the multiple sub-areas to complete the construction of the traffic flow perception module.

3. The intelligent perception and dispatching method for traffic flow under special meteorological conditions according to claim 1 is characterized in that: Extracting a historical traffic data set within a preset period from the traffic flow perception module, and simultaneously obtaining a historical special meteorological data set for the same period, jointly analyzing the historical traffic data set and the historical special meteorological data set, and constructing a meteorological-traffic correlation model, including: Performing spatiotemporal alignment processing on the historical traffic dataset and the historical special meteorological dataset to obtain a spatiotemporal correlation dataset; Extracting features based on the historical traffic data set to obtain multiple traffic indicators; Traversing the plurality of traffic indicators, performing correlation analysis in the spatiotemporal correlation data set, obtaining a plurality of correlation coefficients, and determining a correlation coefficient matrix; The weather-traffic correlation model is constructed according to the correlation coefficient matrix.

4. The intelligent perception and dispatching method for traffic flow under special meteorological conditions according to claim 3 is characterized in that: According to the correlation coefficient matrix, the weather-traffic correlation model is constructed, including: Step a: based on the multiple traffic indicators, obtaining a first traffic indicator, and extracting a plurality of first meteorological variables having correlation coefficients greater than a preset threshold in the correlation coefficient matrix; Step b: constructing a linear regression model based on fitting the first traffic index and the plurality of first meteorological variables; Step c: performing least squares calculation on the linear regression model to obtain multiple regression coefficients, constructing a first correlation model, and adding the first correlation model to the weather-traffic correlation model; For the multiple traffic indicators, repeat steps a to c to complete the construction of the weather-traffic correlation model.

5. The intelligent perception and dispatching method for traffic flow under special meteorological conditions according to claim 2 is characterized in that: The traffic flow sensing module collects real-time traffic flow data, combines the real-time special meteorological data to perform traffic simulation, and constructs a traffic accident heat map, including: Obtain historical traffic accident data and establish a traffic flow simulation model; According to the multiple sub-regions, the real-time traffic flow data and the real-time special meteorological data are integrated into multiple regional data corresponding to the multiple sub-regions; Inputting the plurality of regional data into the traffic flow simulation model to perform traffic simulation, evaluating the traffic simulation results, and obtaining a plurality of traffic risk indexes for the plurality of sub-regions; Based on the multiple traffic risk indices, the traffic accident heat map is generated.

6. The intelligent perception and dispatching method for traffic flow under special meteorological conditions according to claim 5 is characterized in that: Obtain historical traffic accident data and establish a traffic flow simulation model, including: Extracting vehicle driving information from the historical traffic accident data, and establishing a first simulation model using an ABM model; Extracting the average vehicle speed, traffic density and traffic flow in the historical traffic accident data, and establishing a second simulation model using the LWR model; Extracting regional traffic information and road network information from the historical traffic accident data, and using the DTA model to establish a third simulation model; Integrating the first simulation model, the second simulation model and the third simulation model to obtain an initial simulation model; Based on the real-time traffic flow data, the initial simulation model is optimized to obtain the traffic flow simulation model.

7. The intelligent perception and dispatching method for traffic flow under special meteorological conditions according to claim 5 is characterized in that: The traffic accident heat map is analyzed in layers to obtain multiple regional risk levels, and the initial traffic flow scheduling scheme is optimized to obtain a traffic flow scheduling optimization scheme, including: Sorting the plurality of traffic risk indexes in the traffic accident heat map in ascending order to obtain a traffic risk sequence; The traffic risk sequence is hierarchically processed by cluster analysis to obtain multiple risk level thresholds; Based on the multiple risk level thresholds, determining the multiple regional risk levels corresponding to the multiple traffic risk indexes; Determining regional optimization targets according to the multiple regional risk levels; Based on the regional optimization goal, the initial traffic flow scheduling scheme is mutated to obtain a traffic flow scheduling optimization space; Through the traffic flow simulation model, traffic simulation is performed on the traffic flow scheduling optimization space, and optimization analysis is performed on the simulation results to obtain the traffic flow scheduling optimization plan.

8. The intelligent perception and dispatching system of traffic flow under special meteorological conditions is characterized by: The method for implementing the intelligent perception and dispatching method of traffic flow under special meteorological conditions according to any one of claims 1 to 7, wherein the intelligent perception and dispatching system of traffic flow under special meteorological conditions comprises: A correlation model building module is used to extract a historical traffic data set within a preset period from the traffic flow perception module, and simultaneously obtain a historical special meteorological data set for the same period, and jointly analyze the historical traffic data set and the historical special meteorological data set to build a meteorological-traffic correlation model; Meteorological data acquisition module, used to connect to the meteorological system to obtain real-time special meteorological data; An initial plan generation module, used to input the real-time special meteorological data into the meteorological-traffic correlation model, obtain the predicted traffic data for the next period, and generate an initial traffic flow scheduling plan; A traffic simulation module is used to collect real-time traffic flow data through a traffic flow perception module, perform traffic simulation in combination with the real-time special meteorological data, and construct a traffic accident heat map; The scheduling scheme determination module is used to perform hierarchical analysis on the traffic accident heat map to obtain multiple regional risk levels, optimize the initial traffic flow scheduling scheme, and obtain a traffic flow scheduling optimization scheme.

Images