Low-altitude three-dimensional traffic road network division method and system

Abstract

The invention provides a low-altitude three-dimensional traffic road network division method and system. The method comprises the following steps: S1, data acquisition and processing: performing real-time data acquisition and processing on low-altitude three-dimensional traffic road network division areas; s2, dynamically dividing the processed data to obtain divided logic data, and integrating low-altitude air traffic data and low-altitude road surface data to form three-dimensional traffic data; s3, performing structured processing on the dynamically divided logic data and the multi-modal resource integrated three-dimensional traffic data; obtaining a structured result of the low-altitude three-dimensional traffic road network; and S4, calculating a corresponding action specification based on a structured result. Efficient utilization and intelligent management of low-altitude traffic resources are realized through the steps of dynamic division of a low-altitude three-dimensional area, multi-mode resource integration, structured management, intelligent operation standardization and the like.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent traffic network management and low-altitude airspace planning and operation technology for unmanned equipment, and relates to a low-altitude three-dimensional traffic network division method and system. Background Art

[0002] In the existing technology, the division and management technology of low-altitude traffic network mainly relies on: Regular airspace division, through fixed large-scale airspace hierarchical division and static route design, to meet the basic operation needs of low-altitude aircraft. Aircraft route planning, based on geographic information system, combined with basic flight altitude and speed restrictions, to design UAV flight channels. Dynamic airspace resource management, within a limited range, adopts an airspace adjustment method based on task priority. Real-time monitoring technology, tracking aircraft through radar or optical equipment to ensure flight safety.

[0003] With the continuous development of the low-altitude economy, more and more low-altitude economic industries are being explored. Drones alone can no longer meet all industry scenarios. With the continuous emergence of other unmanned equipment, as well as the increasing types and numbers, existing technologies have gradually exposed some problems and shortcomings, as follows: 1. Static planning leads to waste of resources. Existing airspace division and route planning are mostly static, lacking the ability to respond to real-time mission requirements, which easily leads to waste of airspace resources and route congestion.

[0004] 2. Lack of dynamic coordination capabilities. In high-density drone operation scenarios, the ability to dynamically adjust routes is insufficient and cannot effectively avoid flight conflicts and congestion.

[0005] 3. Insufficient data integration makes it difficult to integrate multiple source data such as weather, terrain, and real-time traffic, resulting in insufficient accuracy and applicability of route planning and airspace division.

[0006] 4. Lack of intelligent management methods. Existing technologies for drone airspace management are mostly semi-automatic or manual operations, with a low level of intelligence and unable to fully utilize intelligent technology for optimization.

[0007] 5. Cross-system coordination is difficult. Existing technologies lack the ability to coordinate with ground transportation, port logistics and other systems, making it difficult to achieve low-altitude, three-dimensional integrated traffic management.

[0008] 6. The management equipment is single. The existing technology only divides and manages drone equipment. There is no management technology and specifications for other unmanned equipment such as unmanned vehicles and unmanned ships.

[0009] In summary, solving the above problems is the primary goal of the present invention, while integrating low-altitude multimodal resources to provide a low-altitude three-dimensional transportation network division technology. Summary of the invention

[0010] In response to the above technical problems, the present invention discloses a method and system for dividing a low-altitude three-dimensional transportation road network. By collecting data of the divided areas and utilizing technologies such as dynamic division, multimodal resource integration, structured management and intelligent operation specifications of the low-altitude three-dimensional transportation road network, the problems of the prior art in terms of resource waste, insufficient dynamic coordination and intelligent management are solved, the efficiency and safety of the low-altitude transportation system are improved, and an intelligent low-altitude three-dimensional transportation road network planning technology is formed.

[0011] In order to achieve the above object, the technical solution adopted by the present invention is: A method for dividing a low-altitude three-dimensional traffic road network, characterized in that the method comprises: S1, data collection and processing: Real-time data collection of low-altitude three-dimensional traffic road network division areas is carried out through unmanned equipment and multi-source sensing equipment, and data is processed in real time through computing equipment; during data processing, data from different sources are cleaned, errors and redundant information are eliminated, and standardized processing is performed to ensure data integrity and consistency. Road flow, accident information, and equipment status data are integrated to form a comprehensive three-dimensional traffic data, providing an accurate basis for subsequent analysis.

[0012] By collecting and updating the collected data in real time, the traffic situation can be monitored in real time. Data can be continuously updated according to factors such as road traffic flow, weather changes, and emergencies to ensure the timeliness and accuracy of the data. Real-time data such as road flow, vehicle speed, and traffic congestion can be used to dynamically update and adjust the current data of the traffic network.

[0013] S2, dynamically divide the processed data to obtain the divided logical data, integrate the low-altitude air traffic data and low-altitude road data to form three-dimensional traffic data; S3, structural processing is performed on the dynamically divided logical data and the three-dimensional traffic data integrated with multi-modal resources to obtain the structured results of the low-altitude three-dimensional traffic network; S4, calculate the corresponding action specifications based on the structured results.

[0014] Furthermore, step S1 specifically includes: S11, confirming the area scope of the low-altitude three-dimensional transportation road network, summarizing the unmanned equipment and task information in the area, and designing the data content to be collected based on the task information; S12, using computing equipment to perform data cleaning, format standardization, data conversion, and classified storage on the collected data; S13, evaluate whether the processed data meets the requirements for the division of the three-dimensional transportation network. The data that meets the requirements will be subjected to subsequent dynamic division and multimodal resource integration. The data that does not meet the requirements will be re-processed through the steps S11-S12 until the processed data meets the requirements.

[0015] Furthermore, the data collected in step S11 includes: airspace, terrain, road surface, water body, building, environment, weather, traffic, and unmanned equipment flow in low altitude.

[0016] Furthermore, in step S2, the processed data is organized into a real-time status model according to the dynamic division type to reflect the current traffic status of each node or area; the dynamic division types include: layering, diversion, node, grading, classification, and traffic light.

[0017] Furthermore, in step S2, the three-dimensional traffic data also includes low-altitude water body data. The three-dimensional traffic data includes the following features: S21, three-dimensional traffic road network stratification; according to the height, the low-altitude area is divided from high to low into a long-distance layer, a medium-distance layer, a short-distance layer, a ground layer, and an underwater layer; S22, three-dimensional traffic network diversion; diversion is carried out according to the purpose of the channel, and is divided into operation channel, dispatching channel, task channel and emergency channel; the operation channel is the channel where the unmanned equipment is located when it is operating normally between nodes, the dispatching channel is the channel where the unmanned equipment is located when there is no task to transfer, the task channel is the channel where the unmanned equipment is located when there is a task to execute, and the emergency channel is the fast channel for unmanned equipment in special emergency situations; S23, three-dimensional traffic network nodes; nodes can be divided into normal nodes and temporary nodes, normal nodes are divided into task nodes and exchange nodes, temporary nodes are divided into activity nodes and exchange nodes; wherein, task nodes are small areas temporarily occupied by unmanned equipment when performing normal task items, and activity nodes are small areas temporarily occupied by unmanned equipment when performing temporary activities, and the task node and activity node areas are exchanged through the exchange nodes; S24, three-dimensional traffic road network classification; the classification can be divided into P, C, S, and G according to the priority level, among which P has the lowest priority and G has the highest priority. Equipment running in high-priority road networks has priority to pass. P can only run non-commercial unmanned equipment, C can only run enterprise-level unmanned equipment, S can only run government-level unmanned equipment, and G is for control-level unmanned equipment; S25, classification of three-dimensional traffic network; the classification includes vector class, plane class, three-dimensional class, 4DT class, and special class; the vector class is a point-to-point one-way operation direction; the plane class is an operation area with small upper and lower restrictions; the three-dimensional class is an operation area with large upper and lower restrictions; the 4DT class adds time restrictions to the plane class and three-dimensional class, and can only be operated within the specified time; the special class is any type of vector class, plane class, three-dimensional class, and 4DT class, but can only operate bound unmanned equipment; S26, traffic lights in the three-dimensional traffic network. The traffic lights are divided into red and green lights. Red light means stop and green light means go.

[0018] Furthermore, step S3 specifically includes: Convert 3D traffic data into key-value processable structured data to facilitate subsequent analysis, query and application; Label the data according to the partition logic of the processed data to distinguish different types of data streams; Represent low-altitude three-dimensional transportation network in the form of graph data; Establish a multidimensional data model to dynamically incorporate time and space data into structured results, making it easier to query three-dimensional transportation network data; Align data from different modalities; Use standardized data formats to represent spatial information; Database storage: The database architecture adopts a hybrid database system, which supports the storage of three-dimensional traffic data and dynamically divided data.

[0019] Furthermore, the action specifications of step S4 include: Operation specification: The conditions under which the road section is allowed to operate include that the traffic volume does not reach the maximum threshold and the node connectivity is normal; W 运行 =a1*f 通行能力 +a2*f 流量占比 −a3*f 拥堵指数 ; Among them, a1, a2, and a3 are weight coefficients, which depend on the actual environment; Stop specification: Determine the stop condition based on dynamic data status and path interruption data; stop conditions include: road section interruption, node failure or stop signal, traffic dynamic exceeding threshold, setting safe stop area and backup detour plan based on time dynamic prediction; Waiting rules: The waiting rules are based on the signal light time and node capacity, and the waiting time is dynamically adjusted; T 等待 =(C 当前节点容量 / Q 到达流量 )×F 信号灯时间 ; Switching node entry and exit specifications: P 交换 =F 流量 *F 节点距离 −F 节点负载 , adjust the node entry and exit directions according to the priority to avoid congestion of switching nodes; Scheduling specifications: dynamically adjust the scheduling frequency; F 调度 =Q 总流量 / T 预测时间窗口 ; Emergency specifications: Based on abnormal detection data, set emergency levels and define emergency actions in different situations; Generate detour paths based on dynamic information; Emergency response time: T 应急 =D 目标距离 / V 应急速度 , emergency response plan and route planning.

[0020] Furthermore, the action specifications of step S4 include: Operation specifications include pre-startup inspection, operation conditions, operation monitoring, and fault-tolerant mechanisms for unmanned equipment. Pre-startup inspection should check that the equipment hardware is intact, the sensor is working properly, the battery power is sufficient, the communication module is normal, the self-test program has passed, and the system has no faults; the operation conditions should confirm whether the environment and road network meet the operation requirements; operation monitoring should monitor the equipment posture, speed, remaining power, and communication signal strength in real time; the fault-tolerant mechanism includes standby stop, fault detection, and safety mode; Stop specifications include manual stop, automatic stop, and safe parking. The manual stop means that the remote control terminal sends a stop command, and the equipment responds immediately and stops safely. The automatic stop means that the task is completed and the equipment stops automatically. The safe parking includes return after stopping, parking in a specific area, and battery charging and discharging management. Waiting for the specification, unmanned equipment standby state to low power standby in the temporary parking area of ​​the switching node; Exchange node entry and exit specifications, implement first-in-first-out, sequential exchange principle; Scheduling specifications: Scheduling based on task allocation, conflict avoidance, and coordinated operation, with real-time feedback of equipment location information during the scheduling process; Emergency specifications, including equipment emergency response and emergency tasks, equipment equipped with fault detection and emergency recovery, and the emergency tasks are strictly carried out according to task requirements.

[0021] Furthermore, a digital twin model is created: a real-time mapping digital twin system is constructed using low-altitude three-dimensional traffic network data to realize a three-dimensional traffic network digital twin model; Understand traffic status through real-time collected data flowing into the digital twin model; Use historical data and real-time data to predict traffic changes, node pressure and emergencies, test adjustment strategies in the digital twin model, and obtain simulation results; Path planning optimization: Plan flight paths based on Dijkstra, A* or reinforcement learning path optimization algorithms; adjust route priorities based on collected data; Traffic scheduling: Use traffic prediction algorithms to predict traffic peaks, adjust the airspace stratification and diversion strategy, and balance the traffic pressure in each airspace; Anomaly detection and response: Detect abnormal events and generate emergency response strategies based on the detection results; Real-time visualization of traffic status: Dynamically present low-altitude traffic conditions on the digital twin model.

[0022] The present invention also discloses a low-altitude three-dimensional traffic road network division system, which runs the above-mentioned low-altitude three-dimensional traffic road network division method, including: Unmanned equipment and multi-source sensing equipment are used to collect data on the low-altitude three-dimensional traffic network division area; Computing equipment for real-time processing of collected data; A dynamic partitioning module is used to dynamically partition the processed data to obtain the partitioned logical data; Multimodal resource integration module, used to combine low-altitude air traffic data, low-altitude road data, and low-altitude waterway data to obtain three-dimensional traffic data; Structured management module: used to carry out structured design and management of dynamically divided logical data and three-dimensional traffic data integrated with multi-modal resources; obtain the structured results of low-altitude three-dimensional traffic network; Standardization management module; used to calculate corresponding action specifications based on structured results.

[0023] The present invention adopts the above technical solution to obtain the following beneficial effects: 1. Improve the efficiency of low-altitude resource utilization.

[0024] Through the dynamic three-dimensional transportation network division technology, the low-altitude three-dimensional transportation network can be dynamically adjusted according to real-time demand and traffic flow, reducing the waste of airspace resources and giving full play to the low-altitude economic potential.

[0025] 2. Enhance low-altitude traffic safety.

[0026] Combined with intelligent action specifications, through conflict warning, dynamic adjustment and intelligent scheduling, the occurrence of low-altitude equipment traffic accidents can be reduced and the overall safety of low-altitude traffic operations can be improved.

[0027] 3. Realize efficient integration of multimodal resources.

[0028] Integrate low-altitude unmanned equipment and other low-altitude resources to form a unified low-altitude three-dimensional transportation network, provide support for multiple tasks and multiple scenarios, and solve the problems of resource isolation and coordination.

[0029] 4. Improve the intelligence level of the system.

[0030] With the help of algorithms and digital twin technology, intelligent management of traffic planning and operation can be achieved, providing support for rapid response to emergencies and optimizing the operating efficiency of low-altitude traffic systems.

[0031] (1) Operation optimization through digital twins: Create a digital twin model and use multi-source data (such as low-altitude equipment, geographic information, and traffic flow) to build a real-time mapping digital twin system to realize a three-dimensional transportation network digital twin model.

[0032] The technologies involved in building a real-time mapped digital twin system include: (1.1.1) Select a platform to build a real-time mapping digital twin system. In order to improve real-time performance and data processing efficiency, use an edge computing platform to assist in real-time data preprocessing near the data source.

[0033] (1.1.2) Data collection and processing: Real-time data is obtained through low-altitude equipment (such as drones, unmanned ships, payload equipment, etc.), geographic information systems (GIS), traffic flow sensors, meteorological data, remote sensing data, etc., and after cleaning, de-noising and normalization, it is used for subsequent model construction.

[0034] (1.1.3) The collected visible light data is imported into 3D modeling software or GIS tools to create a 3D model of the digital twin of the traffic network. The support vector machine and neural network algorithms of machine learning are used to learn and construct traffic patterns from the data.

[0035] (1.1.4) Use real-time data stream processing algorithms and deep learning algorithms to process real-time data and predict optimization.

[0036] (1.1.5) Design application models based on actual implementation scenarios, import real-time processed data for training, evaluate model performance through cross-validation to prevent overfitting, and verify on real-time data to ensure its accuracy in actual application.

[0037] (1.1.6) The trained models and data are integrated into the existing visualization system through the API interface and displayed together with the digital twin model to facilitate user understanding and interaction.

[0038] (1.1.7) Real-time data synchronization is performed through the message subscription system, and the application model is updated through regular training.

[0039] (1.1.8) Establish a continuous feedback mechanism to continuously improve models and algorithms through active user interaction and system log analysis to achieve optimization of the three-dimensional transportation network.

[0040] Real-time status monitoring: Real-time data flows into the twin model to understand traffic status, including aircraft location, speed, congestion areas and emergencies.

[0041] Scenario prediction: Use historical and real-time data to simulate possible traffic changes, node pressures, and emergencies. Pre-test multiple operation plans (such as detour path selection and traffic distribution strategies) in a virtual environment to find the optimal strategy. Dynamically adjust the strategy, input real-time traffic data into the prediction model, and update the prediction results; dynamically adjust the three-dimensional traffic network management strategy based on the latest prediction results to ensure that the system is always in the optimal state. Convert simulation results into real-time operation instructions to optimize low-altitude traffic distribution and scheduling decisions.

[0042] The test run scenarios include: (1.2.1) Detour route selection: shortest route detour, least congested route detour, route detour based on time and distance; (1.2.2) Traffic distribution strategy: evenly distribute traffic to all available paths, dynamically allocate traffic based on real-time traffic conditions, and prioritize traffic on high-capacity paths; (1.2.3) Emergency response: Immediately close the affected nodes and redistribute traffic; gradually reduce the traffic on the affected nodes and gradually guide it to alternative paths; keep the affected nodes open but restrict traffic entering.

[0043] The test steps include: (1.3.1) Data preparation: historical data and real-time data. Historical data includes generated traffic, node pressure, emergencies, etc.; real-time data includes data obtained by sensor devices in real time.

[0044] (1.3.2) Build a simulation environment: Use the digital twin system to build a virtual three-dimensional transportation network, input data, simulate the operation status of the real-world three-dimensional transportation network, and dynamically change the environment by setting different traffic parameters.

[0045] (1.3.3) Run the program test: Program 1, apply the shortest path detour strategy in the virtual environment, record the flow changes and node pressure. Program 2, apply the least congested path detour strategy, record the relevant data. Program 3, apply the path detour strategy that comprehensively considers time and distance, and record the relevant data. Repeat the above steps to test all traffic distribution strategies and emergency response plans.

[0046] Methods for determining the optimal strategy include: (1.4.1) Performance indicators: Traffic flow: total flow, average flow, peak flow.

[0047] Node pressure: pressure index, maximum pressure, and average pressure of each node.

[0048] Emergency response time: the time from the occurrence of an incident to the system response.

[0049] User satisfaction: Obtain user satisfaction with traffic conditions through surveys or feedback mechanisms.

[0050] (1.4.2) Data analysis: Comparative analysis: Compare the performance indicators under different solutions and find the optimal solution.

[0051] Sensitivity analysis: Analyze the impact of different parameter changes on solution performance to ensure the robustness of the solution.

[0052] (1.4.3) Decision support system: Multi-objective optimization: Consider multiple performance indicators and use a multi-objective optimization algorithm (such as NSGA-II) to find the optimal strategy.

[0053] Expert review: Invite traffic management experts to review the test results and provide professional opinions.

[0054] (2) Auxiliary optimization.

[0055] Path planning optimization: Based on path optimization algorithms such as Dijkstra, A* or reinforcement learning, combined with real-time geographic data, dynamically plan the optimal flight path. Consider external factors such as weather, traffic, obstacles, etc., adjust route priorities, and avoid wasting resources.

[0056] Traffic scheduling: Use traffic prediction algorithms (such as LSTM or Transformer models) to predict traffic peaks in advance and dynamically allocate resources. For specific time periods, adjust the airspace stratification and diversion strategy to balance the traffic pressure in each airspace.

[0057] Abnormal detection and response: Detect abnormal events such as weather changes, equipment failures, or congestion. Generate corresponding emergency response strategies based on the detection results to optimize aircraft paths and resource allocation.

[0058] (2.1) Use geographic information systems (GIS) to obtain geographic data such as terrain, buildings, roads, etc., and combine it with real-time weather, traffic flow, obstacles and other information obtained through sensors, satellites, drones, etc., and perform data cleaning, fusion, and standardization pre-processing before use.

[0059] Weather factors: Real-time weather data (such as wind speed, rainfall, etc.). Assess the impact of weather on the flight path and adjust the path priority. Dynamically adjust the flight path according to weather changes.

[0060] Traffic flow factors: Real-time traffic flow data. Evaluate the impact of traffic flow on route selection to avoid congestion. Dynamically adjust route priorities based on traffic changes.

[0061] Obstacle factors: Real-time obstacle data (such as buildings, trees, and other unmanned equipment). Assess the impact of obstacles on the path and plan detours. Dynamically adjust the operation path based on changes in obstacles.

[0062] System Integration: Integrate path optimization algorithms such as Dijkstra, A* or reinforcement learning with existing systems through RESTful API interfaces. Use tools such as WebGL and Three.js to visualize path planning results.

[0063] Performance optimization: Use multi-threading, GPU acceleration and other technologies to improve algorithm calculation efficiency. Cache common path planning results to reduce repeated calculations.

[0064] Continuous improvement: Build user feedback and system logs to continuously improve algorithms and models. Update models regularly to adapt to new geographic data and external factors.

[0065] (2.2) The core of anomaly detection and response is to quickly identify abnormal events and generate effective emergency response strategies. Specific methods for generating emergency response strategies include: Anomaly Detection: (2.2.1) Data collection: Obtain real-time weather data (such as wind speed, rainfall, etc.) through meteorological satellites, weather stations, etc. Obtain real-time status data of the aircraft and its equipment (such as battery power, equipment status, load data) through sensors and monitoring systems.

[0066] (2.2.2) Abnormal categories include: Abnormal weather: such as storms, heavy fog, etc. Equipment failure: such as low battery, equipment failure, abnormal load data, etc.

[0067] Specific content of adjusting paths and resources through emergency response strategies includes: (2.3.1) Path adjustment: Detour path, generate detour paths to avoid abnormal areas. Alternative path, pre-plan multiple alternative paths, and select the optimal path based on abnormal conditions. Dynamic adjustment, dynamically adjust the path based on real-time data to ensure that the aircraft is always on the optimal path.

[0068] (2.3.2) Resource allocation: Spare equipment: configure spare equipment at key nodes to ensure rapid replacement when equipment fails. Resource scheduling: dynamically schedule resources according to abnormal conditions (such as increasing the number of aircraft and re-adjusting aircraft tasks). Resource optimization: use optimization algorithms (such as linear programming and integer programming) to optimize resource allocation and ensure maximum resource utilization.

[0069] (2.3.3) Priority adjustment: Task priority, adjust task priority according to abnormal impact (such as giving priority to urgent tasks and delaying non-urgent tasks). Path priority, adjust path priority according to path conditions (such as giving priority to safe paths and avoiding high-risk paths). Resource priority, adjust resource priority according to resource conditions (such as allocating resources to high-priority tasks first).

[0070] (3) Incident response mechanism Real-time response to emergencies: Through the digital twin model, the impact of emergencies on the transportation system (such as node interruption or traffic surge) can be simulated in real time. According to the event and the corresponding emergency regulations, the emergency regulations are triggered automatically (such as flight restriction airspace, reallocation of flight paths) (4) Intelligent operation management Real-time visualization of traffic status: Dynamically present low-altitude traffic conditions on the digital twin model to intuitively identify key issues.

[0071] Multimodal fusion control: Use the rule engine to dynamically adjust operating specifications, such as airspace priority and aircraft speed.

[0072] Resource integration: Intelligently allocate low-altitude resources (such as infrastructure, equipment, and aircraft) to avoid idleness or overload.

[0073] 5. Adapt to complex task requirements.

[0074] Through structured management and multi-level planning, it can flexibly adapt to different types of mission scenarios such as commercial missions, public safety, emergency response, etc., to meet the diverse needs of low-altitude industry equipment transportation.

[0075] 6. Promote the development of low-altitude economy.

[0076] Provide efficient and standardized technical support to various enterprises in the low-altitude economic field, improve the overall industrialization level of the low-altitude field, and accelerate the construction and popularization of low-altitude three-dimensional transportation networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 It is a flow chart of a method for dividing a low-altitude three-dimensional transportation network according to an embodiment of the present invention.

[0078] Figure 2It is a topological diagram of the low-altitude three-dimensional traffic road network division method according to an embodiment of the present invention.

[0079] Figure 3 It is a logical architecture diagram of dynamic partitioning according to an embodiment of the present invention.

[0080] Figure 4 It is a logical architecture diagram of the layered three-dimensional transportation network of an embodiment of the present invention.

[0081] Figure 5 It is a logical architecture diagram of the three-dimensional traffic network diversion according to an embodiment of the present invention.

[0082] Figure 6 It is a logical architecture diagram of a three-dimensional transportation network node according to an embodiment of the present invention.

[0083] Figure 7 It is a logical architecture diagram of the three-dimensional transportation road network hierarchy according to an embodiment of the present invention.

[0084] Figure 8 It is a logical architecture diagram of the three-dimensional transportation network classification according to an embodiment of the present invention.

[0085] Fig. 9 It is a logical architecture diagram of a three-dimensional traffic network traffic light according to an embodiment of the present invention.

[0086] Fig.10 It is a logical architecture diagram of the action specification of an embodiment of the present invention. DETAILED DESCRIPTION

[0087] To facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with embodiments and drawings.

[0088] See also Figure 1 , Figure 2 This embodiment provides a low-altitude three-dimensional transportation road network division method and system, which constructs a low-altitude three-dimensional transportation road network and its unmanned equipment intelligent operation specifications by implementing a series of low-altitude three-dimensional transportation road network division contents and combining low-altitude multimodal resources.

[0089] Figure 1 : is a flow chart of a method for dividing a low-altitude three-dimensional traffic network according to an embodiment of the present invention, which shows the specific process of the method according to this embodiment. Figure 2 The topological diagram of the method for dividing the low-altitude three-dimensional traffic network in the embodiment of the present invention is a star-shaped structure. To complete the division of the three-dimensional traffic network, five basic elements are required: data collection, dynamic division, multi-modal data integration, structured management, and standardized management. The method of this embodiment includes: S1, regional data collection and data processing, the specific process includes: S11, confirm the area scope of the low-altitude three-dimensional transportation road network, summarize the unmanned equipment and mission information in the area, and design the data content that needs to be collected based on the mission information. The data content includes road surface information for road inspection, terrain information for forest inspection, water body information for water conservancy inspection, building information for urban management inspection, and environmental information for environmental protection inspection.

[0090] S12, using computing equipment to clean, standardize, convert, and classify the collected data.

[0091] S13, evaluate whether the processed data meets the requirements for the division of the three-dimensional transportation network. The data that meets the requirements will be subjected to subsequent dynamic division and multimodal resource integration. The data that does not meet the requirements will be re-processed through the steps S11-S12 until the processed data meets the requirements.

[0092] S2, dynamically divide the processed data to obtain the divided logical data; at the same time, integrate multi-modal resources, integrate low-altitude air traffic data, low-altitude road data, and low-altitude water data to form three-dimensional traffic data; Dynamic partitioning: 1. Dynamic data collection and update Real-time data collection technology is used to collect and update data generated by unmanned equipment, including sensors, drones, and ground equipment, to monitor traffic conditions in real time. Data is continuously updated based on factors such as road traffic flow, weather changes, and emergencies to ensure the timeliness and accuracy of the data. Real-time data such as road flow, vehicle speed, and traffic congestion are used to dynamically update and adjust the current data of the traffic network.

[0093] 2. Data cleaning Clean the data from different sources, remove errors and redundant information, and perform standardization to ensure the integrity and consistency of the data. Use big data technology to integrate various data such as road traffic, accident information, and equipment status in multiple dimensions to form a comprehensive three-dimensional traffic data, providing an accurate basis for subsequent analysis.

[0094] 3. Real-time road network status modeling The data is organized into six types of divisions (layering, diversion, node, grading, classification, and traffic light) into real-time status models (6 types) to reflect the current traffic status of each node or area.

[0095] The update frequency of each model can be dynamically adjusted according to traffic conditions, such as increasing the update frequency during rush hour.

[0096] 4. Dynamic data support Provide 6 types of data support and feedback for unmanned equipment.

[0097] 5. Dynamic Tuning Analyze dynamic data patterns and optimize the road network division scheme. Introduce self-learning mechanism to continuously improve response and accuracy.

[0098] 6. Dynamic data display and interaction Build a real-time interactive interface to dynamically present key indicators such as traffic status, diversion results, node pressure, etc., to assist decision makers in responding quickly.

[0099] See also Figure 3 , the contents of dynamic division include the following: The three-dimensional transportation network is layered, such as Figure 4 As shown, preferably, the 1000-meter low-altitude area is divided into a long-range layer (100 meters to 1000 meters), a medium-range layer (50 meters to 100 meters), a short-range layer (10 meters to 50 meters), a ground layer (below 10 meters to the surface of the terrain), and an underwater layer (no limit on altitude, with the water surface of the terrain as the boundary).

[0100] Three-dimensional traffic network diversion, such as Figure 5 As shown in the figure, according to the channel usage, it is divided into operation channel, scheduling channel, task channel and emergency channel. The operation channel is the channel where the unmanned equipment is located when it is operating normally between nodes, the scheduling channel is the channel where the unmanned equipment is located when there is no task transfer, the task channel is the channel where the unmanned equipment is located when there is a task to execute, and the emergency channel is the fast channel for unmanned equipment in special emergency situations. (The channels are divided into four parts: up, down, left, and right. The left and right are the positive and negative operation directions, and the up and down are the geographical directions of east-west and north-south).

[0101] Three-dimensional traffic network nodes, such as Figure 6 As shown, nodes can be divided into normal nodes and temporary nodes. Normal nodes are subdivided into task nodes and exchange nodes. Temporary nodes are also subdivided into active nodes and exchange nodes. Task nodes are small areas temporarily occupied by unmanned equipment when performing normal tasks. Active nodes are small areas temporarily occupied by unmanned equipment when performing temporary activities. The task nodes and active node areas are interchanged through exchange nodes.

[0102] Three-dimensional transportation network classification, such as Figure 7 As shown, they can be divided into P, C, S and G classes in order of priority, among which P class has the lowest priority and G class has the highest priority. Equipment running in high-priority road networks have priority. P class can only run non-commercial unmanned equipment, C class can only run enterprise-level unmanned equipment, S class can only run government-level unmanned equipment, and G class is control-level unmanned equipment.

[0103] Classification of three-dimensional transportation network, such as Figure 8As shown, the classification includes vector class, plane class, solid class, 4DT class and special class, among which the vector class is a point-to-point one-way straight line running direction; the plane class is a free running area with a small upper and lower range (within 10 meters); the solid class is a free running area with a large upper and lower range (more than 10 meters); the 4DT class adds time restrictions to the plane class and solid class and can only run within the specified time; the special class can be any of the first 4 types, but can only be operated through bound unmanned equipment.

[0104] Traffic lights in three-dimensional traffic network, such as Fig. 9 As shown, traffic lights are divided into red lights and green lights. Red lights mean stop and green lights mean go.

[0105] S3, Structuring: Structural processing is performed on the dynamically divided logical data and the three-dimensional traffic data integrated with multimodal resources to obtain the structured results of the low-altitude three-dimensional traffic road network.

[0106] 1. Structured processing of dynamically divided data Objective: To convert the collected low-altitude three-dimensional traffic network related data (such as equipment, location, roads, etc.) into a processable structured data format based on key-value to facilitate subsequent analysis, query and application.

[0107] 2. Data Standardization Unify data from different sources (such as traffic cameras, drone images, sensors) into a standard format, such as unified timestamps, spatial coordinate systems, and data units.

[0108] 3. Hierarchical classification According to the dynamic division logic of the low-altitude road network (layering, diversion, nodes, grading, classification, traffic lights), the data is labeled to distinguish different types of data streams.

[0109] 4. Data Modeling The low-altitude three-dimensional transportation road network is represented in the form of graph data. For example, the layers are represented as height a to height b, the nodes are represented as transportation hubs or the starting and ending points of road sections, etc., and weights are assigned (such as current traffic flow, travelable time, etc.).

[0110] A multidimensional data model is established to dynamically incorporate time and space data into structured results, making it easier to query three-dimensional transportation network data in a certain area at a certain time.

[0111] 5. Multimodal resource integration Multimodal resources include videos, images, unmanned equipment, and sensor data. Targets are extracted from videos and images (such as vehicles, roads, bridges, and traffic lights). Data from different modalities are aligned, such as aligning video frames and sensor streams based on timestamps, and integrating aerial images and unmanned equipment data based on geographic coordinates.

[0112] 6. Generation and management of structured results Use standardized data formats (such as GeoJSON, Shapefile) to represent spatial information, combined with tabular data to represent dynamic indicators.

[0113] Database storage: The database architecture adopts a hybrid database system that supports both graph data (storing traffic network structure) and relational data (storing dynamic attributes).

[0114] S4, Standardization: Calculate the corresponding action norms based on the structured results. Fig.10 As shown, Fig.10 This is a logical architecture diagram of the action specification of an embodiment of the present invention, showing that the key components of the action specification include: Operation specifications: Develop operation specifications, including but not limited to pre-startup inspection, operation conditions, operation monitoring, and fault-tolerant mechanisms for unmanned equipment. Before starting, check that the equipment hardware is intact, the sensor is working properly, the battery is sufficient, the communication module is normal, the self-test program has passed, and the system has no faults. Operation conditions confirm whether the environment, road network and other conditions meet the operation requirements. Operation monitoring monitors the equipment's attitude, speed, remaining power, communication signal strength, etc. in real time. Fault-tolerant mechanisms include standby stop, fault detection, and safe mode.

[0115] Stop specification: formulate stop specifications, including but not limited to manual stop, automatic stop, and safe parking. Manual stop means the remote control terminal sends a stop command, and the equipment responds immediately and stops safely. Automatic stop means automatic stop after task execution. Safe parking includes return after stopping, parking in a specific area, and battery charge and discharge management.

[0116] Waiting specification: From unmanned equipment standby state to low-power standby in the temporary parking area of ​​the switching node.

[0117] Exchange specification: that is, the exchange node entry and exit specifications, implementing the first-in-first-out and sequential exchange principles.

[0118] Scheduling specifications: Scheduling is based on the principles of task allocation, conflict avoidance, coordinated operation, etc., and real-time feedback of equipment location information during the scheduling process.

[0119] Emergency specifications: Emergency specifications are formulated, including the equipment's own emergency response and emergency tasks. The equipment is equipped with fault detection and emergency recovery (such as emergency stop and automatic return, etc.). Emergency tasks are strictly carried out according to task requirements.

[0120] The calculation ideas and formulas for action norms are as follows: 1. Operation Specifications Calculation idea: The conditions for the road section to be allowed to operate include that the traffic volume does not reach the maximum threshold and the node connectivity is normal. Reference formula: W 运行 =a1*f 通行能力 +a2*f 流量占比 −a3*f 拥堵指数 , where a1, a2, and a3 are weight coefficients that depend on the actual environment. The priority path for the road section and the dynamic node entry and exit plan are obtained.

[0121] 2. Stop standardization Calculation ideas: Based on the dynamically divided data status (abnormality, failure) and path interruption data, the stopping conditions include: road section interruption (such as abnormal weather, obstacle detection); node failure or stop signal (such as equipment failure); traffic dynamic exceeding the threshold (such as exceeding the safe carrying capacity); setting up safe stopping areas and backup detour plans based on time dynamic prediction; dynamic prompting of stopping areas and detour plans.

[0122] 3. Waiting for Standardization Calculation idea: The waiting rule is based on the traffic light time and node capacity, and the waiting time is dynamically adjusted.

[0123] Reference formula: T 等待 =(C 当前节点容量 / Q 到达流量 )×F 信号灯时间 , where C is the node carrying capacity and Q is the current traffic.

[0124] 4. Switching node entry and exit specifications Calculation ideas: road network node connectivity and node capacity data.

[0125] Exchange priority reference formula: P 交换 =F 流量 *F 节点距离 −F 节点负载 , adjust the node entry and exit directions according to the priority to avoid congestion of switching nodes.

[0126] 5. Scheduling specifications Calculation ideas: global traffic distribution, node load rate.

[0127] Dynamic adjustment of scheduling frequency reference formula: F 调度 =Q 总流量 / T 预测时间窗口 , and obtain the optimal scheduling path and frequency.

[0128] 6. Emergency regulations Computational idea: Anomaly detection data (weather, obstacles, equipment failure).

[0129] Set emergency levels (such as low, medium, and high) and define emergency actions in different situations.

[0130] Generate a detour path based on dynamic information.

[0131] Emergency response time reference formula: T 应急 =D 目标距离 / V 应急速度 , emergency response plan and route planning.

[0132] S5, Operation: Based on the above structured and standardized results, guide the overall operation of low-altitude unmanned equipment.

[0133] This embodiment also discloses a low-altitude three-dimensional traffic road network division system, including: Unmanned equipment and multi-source sensing equipment are used to collect data on the low-altitude three-dimensional traffic network division area; Computing equipment for real-time processing of collected data; A dynamic partitioning module is used to dynamically partition the processed data to obtain the partitioned logical data; Multimodal resource integration module, used to combine low-altitude air traffic data, low-altitude road data, and low-altitude waterway data to obtain three-dimensional traffic data; Structured management module: used to carry out structured design and management of dynamically divided logical data and three-dimensional traffic data integrated with multi-modal resources; obtain the structured results of low-altitude three-dimensional traffic network; Standardization management module; used to calculate corresponding action specifications based on structured results.

[0134] The above embodiments are only for illustrating the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for dividing a low-altitude three-dimensional traffic network, characterized in that: The method includes: S1, data collection and processing: real-time data collection of low-altitude three-dimensional traffic network division areas through unmanned equipment and multi-source sensing equipment, and real-time data processing through computing equipment; data from different sources are cleaned and unified into a standard format; S2, dynamically divide the processed data to obtain the divided logical data, integrate the low-altitude air traffic data and low-altitude road data to form three-dimensional traffic data; S3, structural processing is performed on the dynamically divided logical data and the three-dimensional traffic data integrated with multi-modal resources to obtain the structured results of the low-altitude three-dimensional traffic network; S4, calculate the corresponding action specifications based on the structured results.

2. The method for dividing a low-altitude three-dimensional traffic network according to claim 1 is characterized in that: The step S1 specifically includes: S11, confirming the area scope of the low-altitude three-dimensional transportation road network, summarizing the unmanned equipment and task information in the area, and designing the data content to be collected based on the task information; S12, using computing equipment to perform data cleaning, format standardization, data conversion, and classified storage on the collected data; S13, evaluate whether the processed data meets the requirements for the division of the three-dimensional transportation network. The data that meets the requirements will be subjected to subsequent dynamic division and multimodal resource integration. The data that does not meet the requirements will be re-processed through the steps S11-S12 until the processed data meets the requirements.

3. The method for dividing a low-altitude three-dimensional traffic network according to claim 2 is characterized in that: The data collected in step S11 includes: airspace, terrain, road surface, water body, building, environment, weather, traffic, and unmanned equipment flow in low altitude.

4. The method for dividing a low-altitude three-dimensional traffic network according to claim 1 is characterized in that: In step S2, the dynamic division types include: layering, diversion, node, grading, classification, and traffic light.

5. The method for dividing a low-altitude three-dimensional traffic network according to claim 4 is characterized in that: In step S2, the three-dimensional traffic data also includes low-altitude water body data, specifically including: Layering: According to the height, the low-altitude area is divided from high to low into long-distance layer, medium-distance layer, short-distance layer, ground layer, and underwater layer; Diversion: Diversion is carried out according to the purpose of the channel, which is divided into operation channel, scheduling channel, task channel and emergency channel; the operation channel is the channel where the unmanned equipment is located when it is operating normally between nodes, the scheduling channel is the channel where the unmanned equipment is located when there is no task to transfer, the task channel is the channel where the unmanned equipment is located when there is a task to execute, and the emergency channel is the fast channel for unmanned equipment in special emergency situations; Nodes: Nodes are divided into normal nodes and temporary nodes. Normal nodes are divided into task nodes and exchange nodes. Temporary nodes are divided into active nodes and exchange nodes. Among them, task nodes are small areas temporarily occupied by unmanned equipment when performing normal tasks, and active nodes are small areas temporarily occupied by unmanned equipment when performing temporary activities. The task nodes and active node areas are exchanged through the exchange nodes. Classification: Classification is divided into P, C, S and G according to the priority. P has the lowest priority and G has the highest priority. Unmanned equipment running in a road network with a high priority has the right of way. P can only run non-commercial unmanned equipment, C can only run enterprise-level unmanned equipment, S can only run government-level unmanned equipment, and G is a regulatory-level unmanned equipment. Classification: Classification includes vector class, plane class, stereo class, 4DT class, and special class; vector class is point-to-point one-way operation direction; plane class is an operation area with small upper and lower restrictions; stereo class is an operation area with large upper and lower restrictions; 4DT class adds time restrictions to plane class and stereo class, and can only be operated within the specified time; special class is any type of vector class, plane class, stereo class, and 4DT class, but can only operate bound unmanned equipment; Traffic lights: Traffic lights are divided into red lights and green lights. Red lights mean stop and green lights mean go.

6. The method for dividing a low-altitude three-dimensional traffic network according to claim 1 or 4, characterized in that: The step S3 specifically includes: Convert the three-dimensional traffic data into a key-value processable structured data format; Label the data according to the dynamic partitioning logic to distinguish different types of data streams; Represent low-altitude three-dimensional transportation network in the form of graph data; Establish a multidimensional data model to dynamically incorporate time and space data into structured results, making it easier to query three-dimensional transportation network data; Align data from different modalities; Use standardized data formats to represent spatial information; Database storage: The database architecture adopts a hybrid database system, which supports the storage of three-dimensional traffic data and dynamically divided data.

7. The method for dividing a low-altitude three-dimensional traffic network according to claim 1 is characterized in that: The action specifications of step S4 include: Operation specification: The conditions under which the road section is allowed to operate include that the traffic volume does not reach the maximum threshold and the node connectivity is normal; <h2 style=";text-align:left;direction:ltr">W<h2 style=";text-align:left;direction:ltr"> 运行 <h2 style=";text-align:left;direction:ltr"> =a1*f<h2 style=";text-align:left;direction:ltr"> 通行能力 <h2 style=";text-align:left;direction:ltr"> +a2*f<h2 style=";text-align:left;direction:ltr"> 流量占比 <h2 style=";text-align:left;direction:ltr"> −a3*f<h2 style=";text-align:left;direction:ltr"> 拥堵指数 <h2 style=";text-align:left;direction:ltr"> ; Among them, a1, a2, and a3 are weight coefficients; Stopping rules: Stopping conditions include road section interruption, node failure or stop signal, dynamic exceeding of traffic flow threshold, setting of safe stopping area and alternative detour plan based on time dynamic prediction; Waiting rules: The waiting rules are based on the signal light time and node capacity, and the waiting time is dynamically adjusted; T 等待 =(C 当前节点容量 / Q 到达流量 )×F 信号灯时间 ; Switching node entry and exit specifications: P 交换 =F 流量 *F 节点距离 −F 节点负载 , adjust the node entry and exit directions according to the priority to avoid congestion of switching nodes; Scheduling specifications: dynamically adjust the scheduling frequency; F 调度 =Q 总流量 / T 预测时间窗口 ; Emergency specifications: Based on abnormal detection data, set emergency levels and define emergency actions in different situations; Generate detour paths based on dynamic information; Emergency response time: T 应急 =D 目标距离 / V 应急速度 , emergency response plan and route planning.

8. The method for dividing a low-altitude three-dimensional traffic network according to any one of claims 1, 6 and 7, characterized in that: The action specifications of step S4 include: Operation specifications include pre-startup inspection, operation conditions, operation monitoring, and fault-tolerant mechanisms for unmanned equipment. Pre-startup inspection should check that the equipment hardware is intact, the sensor is working properly, the battery power is sufficient, the communication module is normal, the self-test program has passed, and the system has no faults; the operation conditions should confirm whether the environment and road network meet the operation requirements; operation monitoring should monitor the equipment posture, speed, remaining power, and communication signal strength in real time; the fault-tolerant mechanism includes standby stop, fault detection, and safety mode; Stop specifications include manual stop, automatic stop, and safe parking. The manual stop means that the remote control terminal sends a stop command, and the equipment responds immediately and stops safely. The automatic stop means that the equipment stops automatically after the task is completed. The safe parking includes return after stopping, parking in a specific area, and battery charging and discharging management. Waiting for the specification, unmanned equipment standby state to low power standby in the temporary parking area of ​​the switching node; Exchange node entry and exit specifications, implement first-in-first-out, sequential exchange principle; Scheduling specifications: Scheduling based on task allocation, conflict avoidance, and coordinated operation, with real-time feedback of equipment location information during the scheduling process; Emergency specifications, including equipment emergency and emergency tasks, equipment equipped with fault detection and emergency recovery, and the emergency tasks are performed according to task requirements.

9. The method for dividing a low-altitude three-dimensional traffic network according to claim 1 is characterized in that: Create a digital twin model: Use low-altitude three-dimensional traffic network data to build a real-time mapping digital twin system to realize a three-dimensional traffic network digital twin model; Understand traffic status through real-time collected data flowing into the digital twin model; Use historical data and real-time data to predict traffic changes, node pressure and emergencies, test adjustment strategies in the digital twin model, and obtain simulation results; Path planning optimization: Plan flight paths based on Dijkstra, A* or reinforcement learning path optimization algorithms; adjust route priorities based on collected data; Traffic scheduling: Use traffic prediction algorithms to predict traffic peaks, adjust the airspace stratification and diversion strategy, and balance the traffic pressure in each airspace; Anomaly detection and response: Detect abnormal events and generate emergency response strategies based on the detection results; Real-time visualization of traffic status: Dynamically present low-altitude traffic conditions on the digital twin model.

10. A low-altitude three-dimensional traffic road network division system, characterized in that: Running the low-altitude three-dimensional traffic road network division method according to any one of claims 1 to 9 comprises: Unmanned equipment and multi-source sensing equipment are used to collect data on the low-altitude three-dimensional traffic network division area; Computing equipment for real-time processing of collected data; A dynamic partitioning module is used to dynamically partition the processed data to obtain the partitioned logical data; Multimodal resource integration module, used to integrate low-altitude air traffic data, low-altitude road data, and low-altitude waterway data to obtain three-dimensional traffic data; Structured management module: used to carry out structured design and management of dynamically divided logical data and three-dimensional traffic data integrated with multi-modal resources; obtain the structured results of low-altitude three-dimensional traffic network; Standardization management module; used to calculate corresponding action specifications based on structured results.

Images