Multi-level real-time traffic flow monitoring and early warning method of three-dimensional traffic management and control system

Abstract

The invention discloses a multi-level real-time traffic flow monitoring and early warning method of a three-dimensional traffic management and control system, and relates to the technical field of traffic management and control, and the method comprises the steps: determining a multi-level traffic structure of the three-dimensional traffic management and control system; establishing a traffic distribution network of each layer; traffic flow abnormal probability identification is carried out, and flow monitoring key points of each layer are established; performing flow monitoring, and establishing a flow monitoring data set of each layer; and carrying out abnormal flow identification based on the flow monitoring data set of each layer, and sending an abnormal early warning signal. The technical problems that in the prior art, in the aspect of three-dimensional traffic flow monitoring, accuracy is insufficient, early warning lags behind, and the comprehensive control capacity of a multi-level traffic structure is poor are solved, and the purposes that the multi-level traffic structure is comprehensively monitored and analyzed, flow monitoring key points are accurately positioned, and the traffic flow monitoring accuracy is improved are achieved. And a data set is established to efficiently identify the abnormal flow and perform early warning in time, so that the accuracy and timeliness of three-dimensional traffic control are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic control, and in particular to a multi-level real-time traffic flow monitoring and early warning method for a three-dimensional traffic control system. Background Art

[0002] With the acceleration of urbanization and the continuous growth of traffic demand, three-dimensional transportation systems have been widely used as a key means to alleviate traffic congestion and improve transportation efficiency. However, existing traffic monitoring technologies have gradually exposed many defects in the face of complex three-dimensional traffic environments. Traditional traffic monitoring methods often focus only on monitoring traffic flow on a single plane, such as ground roads, and lack comprehensive consideration of multi-level traffic structures such as elevated, underground and high-altitude, resulting in the inability to fully and accurately grasp the operating status of the entire three-dimensional transportation system. In terms of traffic flow anomaly monitoring, existing technical means usually rely on simple threshold settings or empirical judgments, which are difficult to adapt to complex and changeable traffic conditions, prone to false alarms and missed reports, and unable to timely and effectively identify potential traffic congestion and safety hazards. Moreover, due to the mutual correlation and mutual influence between the various layers of traffic structures, local traffic flow anomalies may trigger a chain reaction in the entire three-dimensional transportation network, but the existing monitoring system lacks the ability to deeply analyze and predict such systemic impacts, making it difficult for traffic management departments to take targeted control measures in a timely manner, thereby affecting the overall operating efficiency and safety of the three-dimensional transportation system.

[0003] Existing technologies have technical problems in three-dimensional traffic flow monitoring, such as insufficient accuracy, delayed warning, and lack of comprehensive control over multi-level traffic structures. Summary of the invention

[0004] The present application provides a multi-level real-time traffic flow monitoring and early warning method for a three-dimensional traffic control system, which is used to solve the technical problems of insufficient accuracy, delayed early warning, and lack of comprehensive control over multi-level traffic structures in the prior art of three-dimensional traffic flow monitoring.

[0005] In view of the above problems, the present application provides a multi-level real-time traffic flow monitoring and early warning method for a three-dimensional traffic control system, the method comprising:

[0006] Determine the multi-level traffic structure of the three-dimensional traffic control system; determine the distribution information of each level of the traffic structure in the multi-level traffic structure within the preset monitoring area, and establish a traffic distribution network for each level; identify the probability of abnormal traffic flow based on the traffic distribution network for each level, and establish key points for traffic flow monitoring at each level; monitor the traffic flow of each level of the traffic structure based on the key points for traffic flow monitoring at each level, and establish a data set for traffic flow monitoring at each level; identify abnormal traffic flow based on the data set for traffic flow monitoring at each level, and issue an abnormal warning signal.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] Determine the multi-level traffic structure of the three-dimensional traffic control system; determine the distribution information of each level of traffic structure in the multi-level traffic structure within the preset monitoring area, and establish the traffic distribution network of each level; identify the probability of abnormal traffic flow based on the traffic distribution network of each level, and establish the key points of traffic flow monitoring at each level; monitor the traffic flow of each level of traffic structure based on the key points of traffic flow monitoring at each level, and establish the traffic flow monitoring data set of each level; identify abnormal traffic flow based on the traffic flow monitoring data set of each level, and issue abnormal warning signals. The technical effect of comprehensively monitoring and analyzing the multi-level traffic structure, accurately locating the key points of traffic flow monitoring, establishing a data set to efficiently identify abnormal traffic flow and issue timely warnings, and improving the accuracy and timeliness of three-dimensional traffic control is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 A flow chart of a multi-level real-time traffic flow monitoring and early warning method for a three-dimensional traffic control system provided in an embodiment of the present application;

[0011] Figure 2 A schematic diagram of the process of establishing key points for traffic flow monitoring at each layer in the multi-level real-time traffic flow monitoring and early warning method for the three-dimensional traffic control system provided in the embodiment of the present application. DETAILED DESCRIPTION

[0012] The present application provides a multi-level real-time traffic flow monitoring and early warning method for a three-dimensional traffic control system, which is used to solve the technical problems of insufficient accuracy, delayed early warning and lack of comprehensive control over multi-level traffic structures in the prior art in three-dimensional traffic flow monitoring.

[0013] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0014] Examples, such as Figure 1As shown, the present application provides a multi-level real-time traffic flow monitoring and early warning method for a three-dimensional traffic control system, the method comprising:

[0015] Step S100: Determine the multi-level traffic structure of the three-dimensional traffic control system.

[0016] Specifically, determining the multi-level traffic structure of the three-dimensional traffic control system is the basic link of the entire monitoring process. For ground traffic, it is necessary to conduct a detailed survey of the layout of the urban road network, including the direction and connection relationship of the main roads, secondary roads and branches, clarify the types of each intersection (such as crossroads, T-junctions, etc.) and the setting of traffic lights, and master the distribution of bus lanes, non-motorized lanes and sidewalks. In terms of elevated traffic, accurately determine the specific location and number of ramps and the connection method with the ground road, count the number of lanes and traffic directions of different sections, and consider the distribution of emergency parking strips and entrances and exits on the elevated road. For underground traffic, find out the starting and ending points, length and internal lane settings of the tunnel, master the entrance and exit locations of the underground parking lot and the connectivity with surrounding buildings, and clarify the direction and station layout of the subway line underground. For high-altitude transportation, it is necessary to clarify the track lines, station distribution and train operation schedules of the urban light rail. If it involves specific airspace routes, it is necessary to determine its starting and ending airports, route direction and flight altitude layer and other information. By comprehensively and carefully sorting out these elements, the traffic structure of the four levels of ground, elevated, underground and high altitude in the three-dimensional traffic control system can be clearly determined, providing key infrastructure information for subsequent monitoring work.

[0017] Step S200: determining the distribution information of each layer of the multi-layer traffic structure within a preset monitoring area, and establishing a traffic distribution network for each layer.

[0018] Specifically, based on the multi-level traffic structure of the defined three-dimensional traffic control system, the preset monitoring area is precisely focused. For ground traffic structures, with the help of high-precision geographic information system (GIS) technology and field survey data, the starting and ending coordinates, road width, number of lanes and division methods of each road in the monitoring area are determined in detail, and the specific location information of key facilities such as traffic signs, markings, bus stops, and parking lot entrances and exits are marked, and these data are converted into digital node and section information to build a basic network framework for ground traffic. For elevated traffic structures, through laser radar scanning and aerial image analysis, the coordinates of the ramp access point and exit point projected on the ground are accurately determined, and the direction, span and connection relationship between the upper and lower lanes of the elevated bridge body in the monitoring area are clarified. This information is integrated into the network model so that it is interconnected with the ground traffic network and has clear layers. For underground traffic structures, underground detection radar and existing engineering drawings are used to determine the entrance and exit locations of tunnels, internal contour shapes, lane layouts, and the locations of connecting passages with surrounding underground buildings (such as shopping mall basements, underground parking lots, etc.), and these data are integrated into the overall traffic distribution network in a three-dimensional form to reflect its vertical spatial relationship with ground and elevated traffic. For high-altitude traffic structures, combined with radar data and flight plan information provided by the air traffic control department, the projection path of aircraft routes over the monitoring area, the range of flight altitude layers, and the flight trajectories of inbound and outbound flights are determined, and these dynamic air traffic information is added to the traffic distribution network in specific symbols and data formats. Finally, a comprehensive, three-dimensional traffic distribution network covering the detailed distribution information of ground, elevated, underground, and high-altitude traffic structures in the preset monitoring area is established, providing an accurate and intuitive basic data platform for subsequent traffic flow monitoring and analysis.

[0019] Step S300: Identify the probability of traffic flow anomalies based on the traffic distribution network at each layer, and establish key points for traffic flow monitoring at each layer.

[0020] Specifically, using big data collection technology, long-term traffic data is extracted from the historical traffic flow database corresponding to each layer of the traffic distribution network, including multi-dimensional information such as vehicle flow, speed, and vehicle type. For ground transportation, through in-depth mining of traffic data in different sections and intersections in various scenarios such as weekdays, holidays, peak hours, and non-peak hours, statistical analysis methods are used to calculate the mean, variance, standard deviation and other statistics of traffic, and time series analysis models (such as ARIMA models) are used to predict the trend range of normal traffic. The actual observed traffic is compared with the predicted traffic. When the deviation exceeds a certain threshold, the possibility of abnormal traffic at that location is marked. At the same time, combined with the traffic congestion propagation theory, key intersections and sections that are prone to cause large-scale congestion or sudden changes in traffic are determined as key points for ground transportation traffic monitoring. In terms of elevated traffic, the traffic flow variation patterns of different sections are analyzed based on the connection relationship of each section and the flow inflow and outflow characteristics of the ramp. Considering the characteristics of relatively fast vehicle speed and relatively stable traffic flow on the elevated road, the cluster analysis method is used to group sections with similar flow patterns to find sections with large flow fluctuations and significant impact on the overall traffic capacity, such as areas near the on- and off-ramps where merging conflicts and diversion bottlenecks are prone to occur, and locations where vehicles are backlogged due to changes in the number of lanes in the main lane. These areas are set as key points for flow monitoring of elevated traffic. For underground traffic, due to its relatively closed environment and limited space, the focus is on the traffic flow at the entrance and exit of the tunnel, internal curves, and connections with other underground facilities. By jointly analyzing the air quality data fed back by the ventilation system in the tunnel (indirectly reflecting the size of the traffic flow) and the vehicle count data at the entrance and exit, the Bayesian network model is used to infer the probability of flow anomalies, and key locations that are prone to traffic congestion due to vehicle failures, accidents or special events (such as fire drills, maintenance operations, etc.) are identified as key points for flow monitoring of underground traffic. In high-altitude traffic, combined with radar trajectory data and flight takeoff and landing records provided by the air traffic control department, analyze indicators such as aircraft flow density and flight interval time over different routes in the monitoring area. Use the Markov chain model to predict the normal state of air traffic flow. When abnormal situations such as flight delays, alternate landings, and air traffic control cause aircraft flow to deviate from the normal range, identify the route intersections that are more affected, the airspace near the entrance and exit of the airport runway, and busy flight sections as key points for flow monitoring of high-altitude traffic. Through the above analysis methods for each layer of traffic structure, the key points for flow monitoring at each layer are comprehensively established. These key points will become the core focus of subsequent accurate monitoring of traffic flow abnormalities, providing strong support for timely detection and response to traffic congestion and safety hazards.

[0021] Step S400: performing flow monitoring on each layer of the traffic structure based on the flow monitoring key points of each layer, and establishing a flow monitoring data set for each layer.

[0022] Specifically, various advanced monitoring equipment are arranged according to the key points of traffic monitoring at each level that have been determined, so as to implement accurate traffic monitoring actions for each level of traffic structure. For example, for ground traffic, high-definition cameras and geomagnetic sensors are installed at the same time at key intersections and sections prone to congestion to form a monitoring system. Among them, the camera can capture the driving trajectory of vehicles in real time, accurately measure the length of the vehicle queue, and clearly identify visual information such as the vehicle model; while the geomagnetic sensor can accurately sense the number of vehicles passing, driving speed, and stay time. Afterwards, the data collected from the key points of ground traffic will be transmitted to the data processing center with the help of wireless transmission technology.

[0023] By continuously and efficiently monitoring the key points of traffic flow monitoring in each layer of the traffic structure, the huge amount of data collected from diverse sources and in different formats is cleaned, sorted, and stored in a classified manner. In accordance with a unified data format and time series, a traffic flow monitoring data set for each layer is successfully established, providing a comprehensive, accurate and reliable data basis for the subsequent identification and analysis of abnormal traffic, and ultimately achieving accurate grasp and efficient management of the operating status of the three-dimensional transportation system.

[0024] Step S500: Identify abnormal traffic based on the traffic monitoring data sets of each layer and issue an abnormal warning signal.

[0025] Specifically, intelligent data analysis algorithms are used to conduct in-depth mining and processing of traffic monitoring data sets at all levels. For ground traffic data sets, clustering algorithms in machine learning are used to cluster parameters such as vehicle speed, vehicle volume, and vehicle density under normal traffic patterns into one category. When new data points deviate from the normal clustering range, the rule-based reasoning system is combined to determine whether abnormal situations such as a sudden drop or surge in vehicle volume, an abnormal decrease in vehicle speed, etc. are caused by traffic accidents, road construction, or large-scale events. Once an abnormality is confirmed, an alarm device connected to the traffic control system and the communication network, such as an electronic display screen and an intelligent roadside unit, immediately sends an abnormal warning signal to surrounding vehicles and traffic management departments. The signal content contains key information such as the specific location of the abnormality, the possible scope of impact, and the expected duration. In terms of elevated traffic, the deep learning model is used to train the long-term accumulated traffic monitoring data so that it can automatically identify the spatiotemporal characteristics of normal traffic flow. When the speed change rate, lane occupancy rate and other indicators in the real-time monitoring data exceed the normal threshold predicted by the model, the topological structure of the elevated road and the real-time road condition information are combined to determine whether the abnormal traffic is caused by ramp congestion, main line accidents or bad weather. Then, the wireless communication module sends warning information to the vehicles traveling on the elevated road, and notifies the traffic command center to take corresponding traffic control measures, such as flow restriction and diversion, to alleviate the pressure caused by abnormal traffic. For underground traffic, based on the environmental monitoring data in the tunnel (such as ventilation volume, humidity and other parameters that indirectly reflect the traffic flow status) and vehicle flow data, the time series anomaly detection algorithm, such as the Holt-Winters method combined with wavelet transform, is used to detect the mutation point and trend anomaly of the traffic flow. At the same time, considering the closedness and safety of underground traffic, once an abnormality is found, the emergency lighting, ventilation system adjustment and other plans are quickly activated, and an early warning is issued to the variable information board at the entrance of the tunnel to inform the vehicles about to enter the tunnel of the abnormal situation ahead to avoid the vehicles falling into danger. For high-altitude traffic, the aircraft flow, position, speed and other data monitored by the air traffic control radar are compared in real time with the preset air traffic flow management model. When it is found that the aircraft density in a certain airspace exceeds the safety threshold, the flight interval is too short, or there is an abnormal situation such as deviation from the scheduled route, a voice warning signal is immediately sent to the pilot through the aviation communication frequency. At the same time, the abnormal area is highlighted on the display screen of the air traffic control command center, and the surrounding airports and airspace resources are coordinated to re-dispatch and command flights to ensure the safety and smoothness of high-altitude traffic. In this way, accurate identification and timely warning of abnormal traffic can be achieved based on the traffic monitoring data sets at all levels, ensuring the stable operation of the entire three-dimensional transportation system.

[0026] In a possible implementation, step S100 further includes:

[0027] Step S110: The multi-level transportation structure at least includes a ground transportation structure, an elevated transportation structure, an underground transportation structure and an aerial transportation structure.

[0028] Specifically, the multi-level traffic structure in the three-dimensional traffic control system is crucial, which at least covers the ground traffic structure, including the urban road network and various traffic facilities; the elevated traffic structure, which is composed of viaducts, ramps and lanes; the underground traffic structure, which has tunnels, parking lots and subway lines; and the high-altitude traffic structure, such as urban light rail and specific airspace routes. The various layers of structure cooperate with each other, from the ground to the sky, to build a complex and orderly three-dimensional traffic system in all directions, jointly undertake the transportation tasks of urban personnel and materials, meet the diverse traffic needs in different scenarios, improve the overall efficiency and capacity of urban traffic, promote the efficient operation and development of the city, and play an indispensable and key role in the traffic operation of modern cities.

[0029] In one possible implementation, Figure 2 As shown, step S300 also includes:

[0030] Step S310: Collecting historical traffic monitoring data for each layer of the traffic distribution network to generate historical traffic monitoring data sets for each layer.

[0031] Step S320: Calculate the traffic anomaly probability at different locations based on the historical traffic monitoring data sets of each layer, and generate a set of traffic anomaly probabilities of each layer.

[0032] Step S330: locating the key monitoring points of the traffic distribution network at each layer based on the abnormal probability set of the traffic flow at each layer, and generating the key monitoring points of the traffic flow at each layer.

[0033] Specifically, for the traffic distribution networks at all levels that have been constructed, various data collection methods are used to collect historical traffic monitoring data. For the ground traffic network, geomagnetic sensors, high-definition cameras and electronic police equipment deployed at key locations on the road are used to obtain information such as traffic volume, speed, and vehicle type at different sections and intersections in the past period of time, and these data are sorted and stored according to time series and spatial positions to form a historical traffic monitoring data set for the ground layer. At the elevated traffic level, microwave radars and intelligent counters installed at the entrances and exits of ramps and above the main lanes are used to record data such as the number of vehicles entering and exiting, driving speed, and driving trajectory. After data cleaning and preprocessing, a historical traffic monitoring data set for the elevated layer is generated. For underground traffic, such as laser sensors and video surveillance equipment in tunnels, data such as the time interval between vehicle passages, speed changes, and environmental parameters in the tunnel (indirectly reflecting traffic conditions) are collected to construct a historical traffic monitoring data set for the underground layer. High-altitude traffic relies on radar monitoring data from the air traffic control department, including information such as aircraft take-off and landing times, flight altitudes, route trajectories, and flight traffic, which are summarized as a historical traffic monitoring data set for the high-altitude layer.

[0034] Based on the historical traffic monitoring data sets of each layer mentioned above, the probability of traffic anomaly at different locations is calculated using statistical analysis methods. In ground traffic, the mean, standard deviation and change trend of traffic flow in different time periods (weekday peak, non-peak, weekend, etc.) are analyzed with each intersection and road section as the unit. By constructing a probability distribution model, such as a normal distribution model, the probability of actual traffic flow deviating from the normal range is calculated, and these probability values ​​are associated with the corresponding location information to form a set of ground-level traffic anomaly probabilities. For elevated traffic, considering the continuity and directionality of its traffic flow, the Markov chain model is used to evaluate the possibility of traffic anomaly in each section based on the transfer probability of vehicles between different sections and the speed change, thereby generating a set of elevated-level traffic anomaly probabilities. Due to its relatively closed environment and special traffic operation rules, underground traffic combines the correlation between environmental data such as ventilation volume and humidity in the tunnel and traffic flow, and uses the ARIMA model in time series analysis to predict the normal traffic range, calculate the probability of deviation between the actual traffic flow and the predicted value, and obtain a set of underground-level traffic anomaly probabilities. In terms of high-altitude traffic, based on historical flight traffic data and the busyness of the airspace, different airspaces are divided into different traffic patterns through cluster analysis, and the probability of abnormal situations such as flight delays and congestion in each pattern is calculated to generate a high-altitude layer traffic abnormality probability set.

[0035] The key monitoring points of the traffic distribution network at each layer are located based on the set of flow anomaly probabilities at each layer. At the ground level, the intersections and sections where the flow anomaly probability is higher than the set threshold are screened out. These locations are often areas where traffic accidents are frequent, traffic congestion is easy to form, or traffic flow changes are sensitive, such as the main intersections in the city's central business district, roads connecting important transportation hubs, etc. These locations are identified as the key points for flow monitoring at the ground level. At the elevated level, based on the flow anomaly probability set, the key sections that have a greater impact on the overall traffic flow are identified, such as the merging area of ​​the ramp and the main line, the bottleneck section where the number of lanes changes, and the elevated bridge section that is easily affected by bad weather, etc., as the key points for flow monitoring at the elevated level. For underground traffic, locations with a high probability of flow anomalies, such as tunnel entrances and exits, internal curves, and connection points with other underground facilities, are set as key points for underground flow monitoring. Once flow anomalies occur at these locations, they may cause serious traffic congestion and safety hazards. At high altitudes, airspace areas with dense flight traffic, frequent route intersections and easy to be disturbed by weather are identified as key points for high altitude traffic monitoring, such as the approach airspace of large airports, intersection points on busy routes, etc. By focusing on monitoring these key monitoring points, abnormal traffic flow can be discovered in a timely manner, providing strong support for subsequent traffic control and early warning, thereby generating a complete set of key points for traffic monitoring at each layer, and improving the monitoring efficiency and decision-making accuracy of the three-dimensional traffic control system.

[0036] In a possible implementation, step S320 further includes:

[0037] Step S321: Calculate and count the traffic abnormality time ratio coefficients at different locations in the traffic distribution network at each layer based on the historical traffic monitoring data sets at each layer.

[0038] Step S322: Generate the traffic anomaly probability set of each layer according to the traffic anomaly time proportion coefficient.

[0039] Specifically, for each layer of historical traffic monitoring data set, it is necessary to deeply explore the information contained therein to determine the abnormal time ratio coefficient of traffic flow at different locations. Taking ground transportation as an example, the traffic data of each section and intersection at different time periods (such as morning and evening peak hours, off-peak hours on weekdays, all day on weekends, etc.) are extracted from the data set, and the abnormality of traffic flow is defined by setting a reasonable flow threshold range. The duration of traffic flow in an abnormal state within a certain time period is counted, and it is divided by the total monitoring time to obtain the abnormal time ratio coefficient of traffic flow at that location. Similarly, in the elevated traffic layer, based on the historical traffic data of ramp entrances and exits and main lanes, considering the relatively stable and fast characteristics of elevated traffic flow, the adaptive speed and traffic flow thresholds are set as the standard for judging abnormalities, and the abnormal time ratio coefficient of traffic flow at each section is calculated. Due to the relatively closed environment of underground transportation, based on the traffic flow data of tunnel entrances and exits and key internal locations, combined with the indirect traffic information fed back by ventilation, lighting and other systems, the abnormal traffic flow is comprehensively judged and its time ratio coefficient is calculated. For high-altitude traffic, the abnormal traffic time coefficient for each airspace location is determined based on data such as aircraft take-off and landing times, the number of aircraft on routes, etc. provided by the air traffic control department, combined with the normal traffic capacity standards of the airspace.

[0040] The above-calculated flow anomaly time ratio coefficients at different locations in the traffic distribution network of each layer are used to generate the flow anomaly probability set of each layer. On the ground layer, the flow anomaly time ratio coefficients of each intersection and section are normalized to convert these coefficients into values ​​between 0 and 1. These values ​​represent the relative probability of flow anomaly at each location, thereby constructing the flow anomaly probability set of the ground layer. On the elevated layer, similar normalization operations are performed on the flow anomaly time ratio coefficients of each section. At the same time, considering the connectivity and mutual influence between different elevated sections, the probability values ​​are appropriately weighted and adjusted to form the flow anomaly probability set of the elevated layer, so as to more accurately reflect the possibility of flow anomaly at each section. For underground transportation, the probability is corrected according to the flow anomaly time ratio coefficients of each location in the tunnel, combined with structural factors such as the length and slope of the tunnel, to generate the flow anomaly probability set of the underground layer. For high-altitude transportation, the high-altitude flow anomaly probability set is determined based on the flow anomaly time ratio coefficients of each airspace location, combined with the degree of influence of external factors such as weather conditions and air traffic control. In this way, the coefficient of traffic anomaly time proportion is converted into a set of traffic anomaly probabilities at each layer with practical application value, providing a key basis for the subsequent precise positioning of key monitoring points, and improving the three-dimensional traffic control system's ability to predict and respond to traffic anomalies.

[0041] In a possible implementation, step S330 further includes:

[0042] Step S331: Determine a preset critical monitoring flow threshold.

[0043] Step S332: Based on the flow anomaly probability set of each layer, the positions where the flow anomaly probability is greater than or equal to the preset key monitoring flow threshold are screened as key monitoring points to generate the flow monitoring key points of each layer.

[0044] Specifically, determining the preset key monitoring flow threshold is a key link, which requires comprehensive consideration of many factors. First, based on the historical flow data characteristics of each layer of traffic, analyze the flow change patterns and fluctuation ranges in different time periods and regions. For ground transportation, study the differences in traffic volume between weekdays and holidays, peak and off-peak periods, and obtain the frequency of abnormal traffic and the corresponding flow value range under various circumstances through statistical analysis. Combined with the definition standards of ground traffic congestion and smoothness by the traffic management department, determine a flow threshold that can effectively capture potential abnormal traffic conditions and will not generate too many false alarms due to oversensitivity. In terms of elevated transportation, considering the characteristics of high speed and relatively stable traffic flow on the elevated road, a key monitoring flow threshold that is adapted to the operation characteristics of elevated transportation is set based on the degree of influence of ramp inflow and outflow flow on main line traffic, as well as historical data of traffic congestion or accidents caused by abnormal flow in the past. Due to its closedness and special environmental conditions such as ventilation and lighting, underground transportation is limited by the maximum traffic volume for safe passage in the tunnel and the flow restrictions under the requirements of fire protection and emergency evacuation, and the key monitoring flow threshold suitable for underground transportation is determined. For high-altitude traffic, the key monitoring flow threshold that can promptly detect abnormal high-altitude traffic flow is determined based on the airspace busyness level standards set by the air traffic control department, the normal take-off and landing interval requirements for flights, and the air traffic flow restrictions under different weather conditions.

[0045] The key monitoring points are screened based on the determined set of flow anomaly probabilities at each layer and the preset key monitoring flow thresholds. At the ground level, the flow anomaly probabilities of each intersection and road section are compared with the preset thresholds one by one. Locations where the flow anomaly probability is greater than or equal to the threshold are often high-incidence areas of traffic congestion, accident-prone areas, or key nodes that have a greater impact on the overall traffic flow, such as major intersections in the city center, road sections connecting important commercial areas and residential areas, etc. These locations are screened as key monitoring points at the ground level, thereby generating a set of key points for flow monitoring at the ground level. For elevated traffic, along the sections of the elevated bridge, according to the comparison results of their flow anomaly probabilities with the thresholds, sections that are prone to merging conflicts near ramps, sections that are affected by factors such as changes in the number of lanes and sudden changes in vehicle speed, and sections that are greatly affected by factors such as surrounding construction are determined as key monitoring points, forming a set of key points for flow monitoring at the elevated level. In underground transportation, if the probability of abnormal flow reaches or exceeds the preset threshold at locations such as tunnel entrances and exits, internal curves, and connecting passages with other underground facilities, they are selected as key monitoring points in the underground layer, forming a set of key points for monitoring flow in the underground layer, given that abnormal flow at these locations can easily cause serious safety problems and traffic congestion. For high-altitude transportation, based on the comparison of abnormal flow probability and threshold, the airspace near the entrance and exit of the airport runway, the intersection of busy routes, and the airspace range significantly affected by weather changes are identified as key monitoring points in different airspace areas, thereby generating a set of key points for monitoring flow in the high-altitude layer. Through such a screening process, the key monitoring points in each layer of the traffic structure are accurately determined, providing an important basic support for the subsequent implementation of efficient and accurate traffic flow monitoring and abnormal situation warning, which helps to improve the overall operating efficiency and safety of the three-dimensional traffic control system.

[0046] In a possible implementation, step S500 further includes:

[0047] Step S510: configuring a traffic anomaly monitoring model for each layer of traffic monitoring key points, and generating a traffic anomaly monitoring model for each layer.

[0048] Step S520: inputting the traffic monitoring data set of each layer into the traffic anomaly monitoring model of each layer to perform anomaly identification, and generating anomaly identification results of each layer.

[0049] Step S530: issuing the abnormal warning signal based on the abnormal identification results of each layer.

[0050] Specifically, configuring a traffic anomaly monitoring model for key traffic monitoring points at each layer is a key step to achieve accurate monitoring. For key traffic monitoring points of ground traffic, such as intersections in urban centers and congested sections, a support vector machine (SVM) model based on machine learning can be used. Combined with multi-dimensional features such as vehicle volume, speed, and vehicle type in historical traffic flow data, the model can be trained to identify normal traffic flow patterns and various abnormal patterns, such as vehicle backlogs caused by traffic accidents and temporary traffic peaks caused by large-scale activities. By optimizing the model parameters, a traffic anomaly monitoring model suitable for the ground layer is generated, which can quickly and accurately judge the traffic anomaly at key points of ground traffic. In terms of elevated traffic, considering the continuity and relative stability of its traffic flow, the long short-term memory network (LSTM) model in deep learning is used to train the model using information such as the speed change trend of different periods in historical data, the flow inflow and outflow of ramps, etc., so that the model can learn the spatiotemporal characteristics and potential abnormal change laws of elevated traffic flow, so as to generate an elevated layer flow anomaly monitoring model that can predict and identify flow anomalies in advance for key points of flow monitoring on the elevated layer, such as the connection area between the ramp and the main line, the section where the number of lanes changes, etc., and effectively capture traffic anomalies caused by vehicle merging conflicts and bad weather. Due to the special environment of underground traffic, based on the flow data characteristics of key monitoring points such as tunnel entrances and exits and internal curves, a flow anomaly monitoring model based on Bayesian networks is adopted. Combined with the correlation between environmental parameters in the tunnel (such as ventilation volume, humidity, etc.) and vehicle flow, probabilistic reasoning is used to determine whether abnormal flow conditions occur, such as traffic congestion caused by vehicle failure or fire hazards. After structural learning and parameter estimation of the model, an underground layer flow anomaly monitoring model is generated to ensure that safety hazards of underground traffic can be detected in time. For key monitoring points such as the airspace near the entrances and exits of airport runways and intersections of busy routes for high-altitude traffic, a rule-based expert system model is constructed based on data such as aircraft flow, speed, altitude, and flight plans provided by the air traffic control department. Combined with professional knowledge and empirical rules in the aviation field, such as aircraft spacing standards and airspace capacity restrictions, abnormal high-altitude traffic flow is monitored and diagnosed, and a high-altitude layer flow anomaly monitoring model is generated to ensure the safety and smoothness of high-altitude flight.

[0051] The traffic monitoring data sets of each layer are input into the corresponding traffic anomaly monitoring models of each layer for anomaly identification. After the monitoring data sets of the ground layer enter the SVM model, the model classifies and judges the real-time data according to the learned normal and abnormal traffic pattern characteristics, and outputs the abnormal identification results of the ground layer, such as judging whether a certain intersection is blocked due to an accident, the traffic volume drops sharply and the speed approaches zero, and other abnormal conditions. After receiving the real-time traffic monitoring data, the LSTM model of the elevated layer uses its powerful processing ability for time series data to analyze the dynamic change trend of vehicle speed and traffic volume, compares it with the normal mode, and outputs the abnormal identification results of the elevated layer, such as judging whether a certain section of the road has abnormal conditions such as slow vehicle driving and backlog due to bad weather. The Bayesian network model of the underground layer combines the environment and traffic data in the tunnel to calculate the probability of various abnormal conditions. When the probability exceeds the set threshold, the abnormal identification results of the underground layer are determined, such as judging whether the traffic congestion and poor ventilation in the tunnel are caused by vehicle failures. The expert system model of the upper layer obtains abnormality identification results of the upper layer based on the input aircraft flow and flight status data, according to pre-set rules and logical reasoning. For example, it determines whether a certain airspace is in a dangerous situation due to excessive flight flow or control errors resulting in too close aircraft intervals.

[0052] Abnormal warning signals are issued based on the abnormal identification results of each layer. When the abnormal identification results of the ground layer show that a certain key monitoring point has abnormal conditions such as traffic congestion or accidents, the communication system of the traffic control center sends warning information to the surrounding traffic guidance screens to inform the driver of the abnormal conditions at the intersection ahead and recommend a reasonable detour route. At the same time, the accident alarm information is pushed to the terminal equipment of the traffic police department so that the police force can be deployed in time for on-site guidance. For abnormal conditions on the elevated layer, wireless communication technology is used to send voice and text warnings to vehicles traveling on the elevated, prompting the driver to pay attention to the speed and distance between vehicles, and the driver is informed of the traffic status of the ramp through the variable information board to prevent the main line traffic from being paralyzed due to ramp congestion. Once an abnormality occurs in underground traffic, the emergency lighting and ventilation systems in the tunnel are immediately activated, the abnormal conditions and emergency measures are broadcast to the occupants of the vehicle through the tunnel broadcast, and the prohibition or restriction information is issued on the display screen at the entrance of the tunnel to prevent more vehicles from entering the dangerous area. Abnormal warning signals for high-altitude traffic are sent to pilots through the air traffic control communication frequency, informing them of abnormal conditions in the airspace ahead and requiring them to take appropriate avoidance or waiting measures. At the same time, the abnormal airspace is highlighted on the display screen of the air traffic control command center to facilitate flight deployment and command by controllers, ensuring the safe and stable operation of the entire three-dimensional transportation system, responding to various traffic flow abnormalities in a timely and effective manner, and reducing the adverse effects on people's travel and urban operations.

[0053] In a possible implementation, step S510 further includes:

[0054] Step S511: Collect historical traffic anomaly monitoring samples of the key traffic monitoring points at each layer.

[0055] Step S512: training the traffic anomaly monitoring models of each layer based on the historical traffic anomaly monitoring samples.

[0056] Specifically, in order to train accurate and effective traffic anomaly monitoring models for each layer, it is necessary to collect historical traffic anomaly monitoring samples of key traffic monitoring points at each layer. For key monitoring points of ground transportation, such as busy intersections on urban main roads, video data from traffic cameras and vehicle traffic data recorded by geomagnetic sensors are used to trace back past traffic accidents, traffic flow data fragments during large-scale events, and traffic flow data under severe weather conditions, and extract detailed feature information such as vehicle flow, speed change curves, and vehicle queue length under these abnormal conditions as historical traffic anomaly monitoring samples for the ground layer. In terms of elevated transportation, through the intelligent counters at the ramp entrances and the microwave radar equipment on the main lanes, traffic flow data from previous traffic congestion and slow traffic periods caused by ramp construction, vehicle breakdowns, or foggy weather are collected, including speed distribution on different sections, sudden changes in ramp inflow flow, etc., and compiled into historical traffic anomaly monitoring samples for the elevated layer. For underground transportation, laser sensors and video surveillance systems in tunnels are used to obtain data on vehicle flow, ventilation volume, humidity changes, and other abnormal events such as fire hazards caused by vehicle failures and water accumulation in tunnels, and to construct historical traffic anomaly monitoring samples for the underground layer. For high-altitude transportation, the radar data of the air traffic control department and the flight operation records of the airport are used to screen out aircraft flow, flight trajectories, altitude and speed changes, and other data on abnormal situations such as flight delays, alternate landings, and air waiting caused by air traffic control errors, temporary runway closures, and extreme weather, to form historical traffic anomaly monitoring samples for the high-altitude layer.

[0057] Based on the above-mentioned historical traffic anomaly monitoring samples, targeted training is performed on the traffic anomaly monitoring models at each layer. The support vector machine (SVM) model at the ground layer takes the collected various types of ground traffic anomaly samples as input, standardizes the feature vectors such as vehicle flow, vehicle speed, and vehicle type in the samples, and substitutes them into the SVM training algorithm. By continuously adjusting the model's hyperparameters, such as the kernel function type, the model can accurately distinguish between normal traffic flow and various abnormal traffic modes, and improve the recognition accuracy of ground traffic flow anomalies. The long short-term memory network (LSTM) model at the elevated layer receives historical traffic anomaly samples of elevated traffic, inputs feature information such as vehicle speed change trends and ramp flow data into the model in the order of time series, and uses the back propagation algorithm to continuously optimize the weight matrix in the model, so that the model can learn the spatiotemporal evolution of elevated traffic flow under abnormal conditions, and enhance the ability to predict and judge elevated traffic flow anomalies. The Bayesian network model of the underground layer builds the probability relationship between network nodes based on the abnormal samples of underground traffic, and learns and updates the conditional probability table in the model through methods such as maximum likelihood estimation, so that the model can accurately infer the probability of abnormal situations based on the environment and flow data in the tunnel, and improve the monitoring accuracy of underground traffic safety hazards. The rule-based expert system model of the high-altitude layer matches and verifies the aircraft flow, flight status and other data collected in the abnormal samples of high-altitude traffic with the pre-set aviation field rules and empirical knowledge, and continuously improves the rule base and reasoning mechanism in the expert system, so that it can more accurately diagnose and monitor abnormal high-altitude traffic flow, and ensure that the abnormal flow monitoring models of each layer can efficiently and accurately identify and monitor abnormal real-time traffic flow after sample training, providing strong technical support for the three-dimensional traffic control system.

[0058] In a possible implementation, step S530 further includes:

[0059] Step S531: extracting real-time abnormal traffic distribution information based on the abnormal warning signal.

[0060] Step S532: Input the real-time abnormal traffic distribution information into the abnormal handling decision module to optimize the traffic control decision and generate the target control decision, wherein the abnormal handling decision module is constructed based on the historical traffic abnormal distribution and historical abnormal handling decision of the key points of traffic monitoring at each layer.

[0061] Step S533: Recommending the target control decision to the management terminal of the three-dimensional traffic control system.

[0062] Specifically, once the abnormal warning signal is triggered, the real-time abnormal traffic flow distribution information is quickly extracted based on the signal. For ground transportation, the specific location, affected range, and detailed data such as vehicle flow and speed of the abnormal traffic area can be quickly obtained through monitoring equipment distributed at key intersections and sections, such as high-definition cameras and geomagnetic sensors, to accurately determine the traffic congestion of a certain section of road due to a traffic accident, and the traffic flow convergence of its upstream and downstream intersections, so as to form real-time abnormal traffic flow distribution information for ground transportation. In terms of elevated transportation, with the help of monitoring devices on ramps and main lanes, the congestion degree at the ramp entrance, the slow-moving section of the main lane, and the direction of vehicle backlog are clearly defined, and integrated into real-time abnormal traffic flow distribution information for elevated transportation. Underground transportation relies on various sensors in the tunnel to grasp the starting position, development trend, and impact of abnormal traffic flow on ventilation, lighting and other systems in the tunnel, and generate real-time abnormal traffic flow distribution information for underground transportation. Based on air traffic control radar and airport surface monitoring system, high-altitude transportation determines the airspace range, number of flights and flight status affected by abnormal conditions, and summarizes real-time abnormal traffic flow distribution information for high-altitude transportation.

[0063] These real-time abnormal traffic distribution information are input into the abnormal handling decision module, which is an intelligent system built based on the historical abnormal traffic distribution and historical abnormal handling decisions of key points of traffic monitoring at each layer. For ground traffic, it stores the traffic control measures taken in the past when abnormal traffic was caused by similar traffic accidents and large-scale activities, such as intersection traffic control plans, vehicle diversion routes, etc., and uses intelligent algorithms to optimize traffic control decisions in combination with the current real-time abnormal traffic distribution information. By simulating the impact of different control strategies on traffic flow, such as the changes in traffic flow after implementing temporary road closures, signal light timing adjustments, and inducing vehicles to detour, the optimal solution is selected to generate the target control decision for ground traffic. The abnormal handling decision module for elevated traffic is based on historical experience such as ramp congestion and elevated driving safety in bad weather, combined with real-time abnormal traffic conditions, and considers measures such as restricting ramp vehicle entry and adjusting the main line speed limit. Through the optimization algorithm, the control decision that is most conducive to alleviating elevated traffic congestion and ensuring safety is determined, and the elevated traffic target control decision is generated. The decision-making module for underground transportation is based on past cases of handling emergencies in tunnels, combined with abnormal conditions in the current tunnels, such as fire hazards caused by vehicle failures, to develop an optimized combination of measures including starting emergency ventilation, organizing vehicle evacuation, and implementing traffic control, and generate underground transportation target control decisions. The abnormal handling decision-making module for high-altitude transportation refers to the historical handling of flight delays and air traffic congestion caused by aviation control errors, extreme weather, etc., combined with the current affected airspace scope and flight operation status, and through coordination with airlines, airports and other relevant departments, determine target control decisions such as adjusting the order of flight take-offs and landings, optimizing route planning, and implementing traffic control to ensure the safety and smoothness of high-altitude transportation.

[0064] The generated target control decisions are recommended to the management terminal of the three-dimensional traffic control system. The staff of the management terminal, such as the traffic police at the traffic police command center and the airport air traffic control personnel, can receive these optimized control decision information in a timely manner. The control decisions of ground traffic are presented in the form of intuitive map annotations, signal light control instructions, and traffic guidance screen information release, which facilitates the traffic police to quickly implement traffic control and guide vehicle traffic. The target control decisions of elevated traffic are transmitted to the elevated traffic management platform through wireless communication, providing management personnel with operational suggestions such as ramp control and vehicle speed adjustment to ensure the orderly operation of elevated traffic. The decision information of underground traffic is transmitted to the tunnel management department to guide it in emergency response and traffic diversion to ensure the safety of underground traffic. The target control decisions of high-altitude traffic are directly pushed to the controllers of the air traffic control command center. The controllers issue flight instructions to pilots based on these decision information, coordinate airlines to adjust flight plans, and realize effective management of high-altitude traffic. Through this series of steps, the complete process from abnormal warning to accurate decision-making to effective implementation is realized, and the ability and efficiency of the three-dimensional traffic system to deal with abnormal traffic conditions is improved, ensuring the safe and stable operation of the entire traffic network.

[0065] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0066] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0067] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the 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 equivalents, the present application intends to include these modifications and variations.

Claims

1. A multi-level real-time traffic flow monitoring and early warning method for a three-dimensional traffic control system, characterized in that: include: Determine the multi-level traffic structure of the three-dimensional traffic control system; Determine the distribution information of each layer of the multi-layer traffic structure in the preset monitoring area, and establish a traffic distribution network for each layer; Based on the traffic distribution network of each layer, the probability of traffic flow anomaly is identified, and key points for traffic flow monitoring at each layer are established; Based on the key points of flow monitoring at each layer, flow monitoring is performed on the traffic structure at each layer, and a flow monitoring data set at each layer is established; Abnormal traffic is identified based on the traffic monitoring data sets of each layer, and abnormal warning signals are issued.

2. The multi-level real-time traffic flow monitoring and early warning method of the three-dimensional traffic control system as claimed in claim 1 is characterized in that: The multi-level traffic structure at least includes a ground traffic structure, an elevated traffic structure, an underground traffic structure and an aerial traffic structure.

3. The multi-level real-time traffic flow monitoring and early warning method of the three-dimensional traffic control system as claimed in claim 1 is characterized in that: Based on the traffic distribution network of each layer, the abnormal probability of traffic flow is identified, and the key points of traffic flow monitoring at each layer are established, including: Collecting historical traffic monitoring data for each layer of the traffic distribution network to generate historical traffic monitoring data sets for each layer; Calculating the traffic anomaly probability at different locations based on the historical traffic monitoring data sets of each layer, and generating a set of traffic anomaly probabilities at each layer; Based on the abnormal probability sets of the traffic flow at each layer, key monitoring points of the traffic distribution network at each layer are located to generate key monitoring points of the traffic flow at each layer.

4. The multi-level real-time traffic flow monitoring and early warning method of the three-dimensional traffic control system as claimed in claim 3 is characterized in that: Based on the historical traffic monitoring data sets of each layer, the traffic anomaly probabilities at different locations are calculated to generate a set of traffic anomaly probabilities at each layer, including: Calculate and count the traffic abnormality time ratio coefficients at different locations in the traffic distribution network at each layer based on the historical traffic monitoring data sets at each layer; The flow anomaly probability set of each layer is generated according to the flow anomaly time proportion coefficient.

5. The multi-level real-time traffic flow monitoring and early warning method of the three-dimensional traffic control system as claimed in claim 4 is characterized in that: Locating the key monitoring points of the traffic distribution network at each layer based on the abnormal probability set of each layer of traffic flow, and generating the key monitoring points of the traffic flow at each layer, including: Determine preset critical monitoring flow thresholds; Based on the flow anomaly probability set of each layer, the positions where the flow anomaly probability is greater than or equal to the preset key monitoring flow threshold are screened as key monitoring points to generate the flow monitoring key points of each layer.

6. The multi-level real-time traffic flow monitoring and early warning method of the three-dimensional traffic control system according to claim 1 is characterized in that: Based on the traffic monitoring data sets of each layer, abnormal traffic is identified and abnormal warning signals are issued, including: Configuring a traffic anomaly monitoring model for each layer of traffic monitoring key points to generate a traffic anomaly monitoring model for each layer; Inputting the traffic monitoring data sets of each layer into the traffic anomaly monitoring model of each layer to perform anomaly identification and generate anomaly identification results of each layer; The abnormality warning signal is issued based on the abnormality identification results of each layer.

7. The multi-level real-time traffic flow monitoring and early warning method of the three-dimensional traffic control system as claimed in claim 6 is characterized in that: Configuring a traffic anomaly monitoring model for each layer of traffic monitoring key points to generate a traffic anomaly monitoring model for each layer, including: Collect historical traffic anomaly monitoring samples of key traffic monitoring points at each layer; The traffic anomaly monitoring models of each layer are trained based on the historical traffic anomaly monitoring samples.

8. The multi-level real-time traffic flow monitoring and early warning method of the three-dimensional traffic control system as claimed in claim 1 is characterized in that: After the abnormal warning signal is issued, the method further includes: Extracting real-time abnormal traffic distribution information based on the abnormal warning signal; Input the real-time abnormal traffic flow distribution information into the abnormal handling decision module to optimize the traffic control decision and generate the target control decision, wherein the abnormal handling decision module is constructed based on the historical traffic abnormal distribution and historical abnormal handling decision of the key points of the traffic monitoring at each layer; The target control decision is recommended to the management terminal of the three-dimensional traffic control system.

Images