A method and system for dividing a low-altitude three-dimensional traffic road network

Abstract

The present invention provides a method and system for dividing a low-altitude three-dimensional traffic road network. The method includes: S1, data collection and processing: real-time data collection and processing are carried out on the area for dividing the low-altitude three-dimensional traffic road network; S2, the processed data is dynamically divided to obtain the divided logical data, and the low-altitude air traffic data and low-altitude road surface data are integrated to form three-dimensional traffic data; S3, the logically divided data and the three-dimensional traffic data integrated with multimodal resources are subjected to structured processing to obtain the structured result of the low-altitude three-dimensional traffic road network; S4, corresponding operation specifications are calculated based on the structured result. Through steps such as dynamic division of the low-altitude three-dimensional area, multimodal resource integration, structured management, and intelligent operation specifications, the efficient utilization and intelligent management of low-altitude traffic resources are realized.

Description

Technical Field

[0001] The present invention belongs to the technical fields of intelligent transportation road network management and unmanned equipment low-altitude airspace planning and operation; it relates to a method and system for dividing a low-altitude three-dimensional transportation road network. Background Art

[0002] In the prior art, the division and management technologies of low-altitude transportation road networks mainly rely on: regular airspace division, through fixed large-scale airspace level division and static route design, to meet the basic operation needs of low-altitude aircraft. Aircraft route planning, based on geographic information systems, combined with basic flight altitude and speed limits, to design UAV flight channels. Dynamic airspace resource management, within a limited range, using an airspace adjustment method based on task priority. Real-time monitoring technology, tracking aircraft through radar or optical devices to ensure flight safety.

[0003] With the continuous development of the low-altitude economy, more and more low-altitude economic industries have been continuously explored. Only relying on UAVs can no longer meet all industry scenarios, and the continuous emergence of other unmanned equipment, as well as the continuous increase in types and quantities, have gradually exposed some problems and shortcomings in the existing technologies, as follows:

[0004] 1. Static planning leads to waste of resources. The existing airspace division and route planning are mostly static, lacking the ability to respond to real-time task requirements, and easily causing waste of airspace resources and route congestion.

[0005] 2. Lack of dynamic coordination ability. In high-density UAV operation scenarios, the ability to dynamically adjust routes is insufficient, and it is impossible to effectively avoid flight conflicts and congestion.

[0006] 3. Insufficient data integration. It is difficult to integrate multi-source data such as weather, terrain, and real-time traffic, resulting in insufficient accuracy and applicability of route planning and airspace division.

[0007] 4. Lack of intelligent management means. The existing technologies for UAV airspace management are mostly semi-automated or manual operations, with low intelligence, and cannot make full use of intelligent technologies for optimization.

[0008] 5. Difficulty in cross-system collaboration. The existing technologies lack the collaborative ability with systems such as ground transportation and port logistics, and it is difficult to achieve low-altitude three-dimensional comprehensive traffic management.

[0009] 6. Single management equipment. The existing technologies only divide and manage UAV equipment, and there are no management technologies and specifications for other unmanned equipment such as unmanned vehicles and unmanned boats.

[0010] In summary, solving the above problems is the primary goal of the present invention. At the same time, integrating low-altitude multimodal resources, a technology for dividing a low-altitude three-dimensional transportation road network is provided. Summary of the Invention

[0011] In view of the above technical problems, the present invention discloses a method and system for dividing a low-altitude three-dimensional traffic road network. By collecting data of the divided area and using technologies such as dynamic division of the low-altitude three-dimensional traffic road network, multi-modal resource integration, structured management, and intelligent operation specifications, the problems in the prior art such as resource waste, insufficient dynamic coordination, and intelligent management are solved, the efficiency and safety of the low-altitude traffic system are improved, and an intelligent low-altitude three-dimensional traffic road network planning technology is formed.

[0012] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0013] A method for dividing a low-altitude three-dimensional traffic road network, characterized in that the method includes:

[0014] S1, data collection and processing: Real-time data collection of the low-altitude three-dimensional traffic road network division area is carried out through unmanned devices and multi-source sensing devices, and the data is processed in real time by a computing device; in data processing, data from different sources is cleaned, error and redundant information is removed, and standardized processing is carried out to ensure the integrity and consistency of the data. The road traffic flow, accident information, and equipment status data are fused to form a comprehensive three-dimensional traffic data, providing an accurate basis for subsequent analysis.

[0015] By collecting and updating the collected data in real time, the traffic situation is monitored in real time. The data is 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 traffic flow, vehicle speed, and traffic congestion is used to dynamically update and adjust the current data of the traffic road network.

[0016] S2, the processed data is dynamically divided to obtain the divided logical data, and the low-altitude air traffic data and low-altitude road surface data are integrated to form three-dimensional traffic data;

[0017] S3, perform structured processing on the logically divided data and the three-dimensional traffic data integrated with multi-modal resources; obtain the structured result of the low-altitude three-dimensional traffic road network;

[0018] S4, calculate the corresponding action specifications based on the structured results.

[0019] Further, step S1 specifically includes:

[0020] S11, confirm the area range of the low-altitude three-dimensional traffic road network division, summarize the unmanned devices and task information within the area, and design the data content to be collected based on the task information;

[0021] S12, use a computing device to perform data cleaning, format standardization, data conversion, and classified storage on the collected data;

[0022] S13. Evaluate whether the processed data meets the requirements for the division of the three-dimensional transportation road network. The data that meets the requirements will undergo subsequent dynamic division and multimodal resource integration. The data that does not meet the requirements will be reprocessed through the data collection and processing procedures of steps S11 - S12 until the processed data meets the requirements.

[0023] Furthermore, the data content collected in step S11 includes: airspace, terrain, road surface, water body, buildings, environment, meteorology, traffic, and unmanned device flow in the low altitude.

[0024] Furthermore, in step S2, the processed data is regularized into a real-time state model according to the dynamic division types to reflect the current traffic state of each node or area. The dynamic division types include: layering, flow splitting, nodes, grading, classification, and traffic lights.

[0025] Furthermore, in step S2, the three-dimensional transportation data also includes low-altitude water body data. The three-dimensional transportation data has the following characteristics: S21. The three-dimensional transportation road network is layered; 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 according to height;

[0026] S22. The three-dimensional transportation road network is flow-split; it is split according to the channel usage into an operation channel, a scheduling channel, a task channel, and an emergency channel. Among them, the operation channel is the channel where unmanned devices normally operate between nodes, the scheduling channel is the channel where unmanned devices are located during non-task transfer, the task channel is the channel where unmanned devices are located when performing tasks, and the emergency channel is the fast channel for unmanned devices in special emergency situations;

[0027] S23. The three-dimensional transportation road network nodes; the nodes can be divided into normal nodes and temporary nodes. The normal nodes are divided into task nodes and exchange nodes, and the temporary nodes are divided into activity nodes and exchange nodes. Among them, the task node is a small area temporarily occupied by an unmanned device when performing a normal task item, and the activity node is a small area temporarily occupied by an unmanned device when performing a temporary activity. The areas of the task node and the activity node are interchanged through the exchange node;

[0028] S24. The three-dimensional transportation road network is graded; the grading can be divided into P level, C level, S level, and G level according to the priority from low to high. Among them, the P level has the lowest priority, and the G level has the highest priority. Devices operating in the high-priority road network have the right of way. Only non-commercial unmanned devices can operate at the P level, only enterprise-level unmanned devices can operate at the C level, only government-level unmanned devices can operate at the S level, and the G level is for controlled unmanned devices;

[0029] S25, Classification of three-dimensional transportation road networks; the classification includes vector type, planar type, three-dimensional type, 4DT type, and special type; the vector type has a one-way running direction from point to point; the planar type has a running area restricted in a small range up and down; the three-dimensional type has a running area restricted in a large range up and down; the 4DT type adds a time limit to the planar type and the three-dimensional type and can only run within the specified time; the special type is any type among the vector type, planar type, three-dimensional type, and 4DT type, but can only run the bound unmanned devices.

[0030] S26, Traffic lights for three-dimensional transportation road networks. The traffic lights are divided into red lights and green lights. Red lights mean stop, and green lights mean go.

[0031] Further, step S3 specifically includes:

[0032] Convert three-dimensional transportation data into a key-value based processable structured data form for subsequent analysis, query, and application;

[0033] Mark tags for the data according to the partitioning logic of the processed data to distinguish different types of data streams;

[0034] Represent the low-altitude three-dimensional transportation road network in the form of graph data;

[0035] Establish a multi-dimensional data model and dynamically incorporate time and space data into the structured result for easy query of three-dimensional transportation road network data;

[0036] Align data from different modalities;

[0037] Use a standardized data format to represent spatial information;

[0038] Database storage: The database architecture adopts a hybrid database system, which supports storing both three-dimensional transportation data and dynamically partitioned data.

[0039] Further, the operation specifications of step S4 include:

[0040] Operation specifications: The conditions for allowing running on a section of the road include that the traffic flow has not reached the maximum threshold and the node connectivity is normal;

[0041] W 运行 =a1*f 通行能力 +a2*f 流量占比 −a3*f 拥堵指数 ;

[0042] where a1, a2, and a3 are weight coefficients, depending on the actual environment;

[0043] Stop Specification: Determine the stop condition based on the dynamic data status and path interruption data; the stop conditions include: road section interruption, node failure or stop signal, traffic flow dynamics exceeding the threshold, dynamic prediction based on time, setting up a safety stop area and an alternative detour plan;

[0044] Waiting Specification: The waiting rule is based on the signal light time and node capacity, and dynamically adjusts the waiting time;

[0045] T 等待 = (C 当前节点容量 / Q 到达流量 ) × F 信号灯时间 ;

[0046] In-and-Out Specification of Switching Nodes:

[0047] P 交换 = F 流量 * F 节点距离 − F 节点负载 , adjust the in-and-out direction of the node according to the priority to avoid congestion at the switching node;

[0048] Scheduling Specification: Dynamically adjust the scheduling frequency;

[0049] F 调度 = Q 总流量 / T 预测时间窗口 ;

[0050] Emergency Specification: Based on the anomaly detection data, set the emergency level and define the emergency actions in different situations;

[0051] Generate a detour path based on the dynamic information;

[0052] Emergency Response Time: T 应急 = D 目标距离 / V 应急速度 , emergency response plan and path planning.

[0053] Furthermore, the action specification in step S4 includes:

[0054] Operation Specification, including pre-start inspection, operation conditions, operation monitoring, and fault tolerance mechanism of the unmanned device. The pre-start inspection should check that the device hardware is intact, the sensor is working properly, the battery power is sufficient, the communication module is normal, the self-check program passes, and the system is fault-free; the operation conditions should confirm whether the environment and road network meet the operation requirements; the operation monitoring should monitor the device attitude, speed, remaining battery power, and communication signal strength in real time; the fault tolerance mechanism includes standby stop, fault detection, and safety mode;

[0055] Stop specifications, including manual stop, automatic stop, and safe parking. The manual stop is that the remote control terminal sends a stop instruction, and the device immediately responds and shuts down safely. The automatic stop is to complete the automatic stop after the task execution is completed. The safe parking includes returning after stopping, parking in a specific area, and battery charge and discharge management.

[0056] Waiting specifications: The unmanned device is in a standby state and waits in a low-power state in the temporary parking area of the switching node.

[0057] Switching node access specifications: Implement the principle of first-in, first-out and sequence exchange.

[0058] Scheduling specifications: Perform scheduling based on task assignment, conflict avoidance, and coordinated operation, and provide real-time feedback on the device position information during the scheduling process.

[0059] Emergency specifications, including device self-emergency and emergency tasks. The device is equipped with fault detection and emergency recovery, and the emergency tasks are strictly operated according to the task requirements.

[0060] Furthermore, create a digital twin model: Use the low-altitude three-dimensional traffic road network data to construct a real-time mapped digital twin system to realize the digital twin model of the three-dimensional traffic road network.

[0061] Master the traffic status by flowing the real-time collected data into the digital twin model.

[0062] Use historical data and real-time collected data to predict traffic flow changes, node pressures, and emergencies, and test and adjust strategies in the digital twin model to obtain simulation results.

[0063] Path planning optimization: Based on path optimization algorithms such as Dijkstra, A*, or reinforcement learning, plan the flight path; adjust the route priority in combination with the collected data.

[0064] Traffic flow scheduling: Use traffic flow prediction algorithms to predict traffic flow peaks, adjust the hierarchical diversion strategy of airspace, and balance the traffic flow pressure in each airspace.

[0065] Anomaly detection and response: Detect abnormal events and generate emergency response strategies according to the detection results.

[0066] Real-time visualization of traffic flow status: Dynamically present the low-altitude traffic conditions on the digital twin model.

[0067] 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:

[0068] Unmanned devices and multi-source perception devices, used to collect data for the area divided by the low-altitude three-dimensional traffic road network.

[0069] A computing device for real-time processing of collected data;

[0070] A dynamic partitioning module for dynamically partitioning the processed data to obtain partitioned logical data;

[0071] A multi-modal resource integration module for integrating low-altitude air traffic data, low-altitude road surface data, and low-altitude waterway data to obtain three-dimensional traffic data;

[0072] A structured management module for performing structured design and management on the dynamically partitioned logical data and the three-dimensional traffic data integrated by multi-modal resources, and obtaining a structured result of the low-altitude three-dimensional traffic road network;

[0073] A normalization management module for calculating corresponding operation specifications based on the structured result.

[0074] By adopting the above technical solutions, the present invention can achieve the following beneficial effects:

[0075] 1. Improve the utilization efficiency of low-altitude resources.

[0076] Through the dynamic partitioning technology of the three-dimensional traffic road network, the low-altitude three-dimensional traffic road network is dynamically adjusted according to real-time requirements and traffic flow, reducing the waste of airspace resources and giving full play to the potential of the low-altitude economy.

[0077] 2. Enhance the safety of low-altitude traffic.

[0078] Combined with intelligent operation specifications, through conflict warning, dynamic adjustment, and intelligent scheduling, the occurrence of traffic accidents of low-altitude devices is reduced, and the overall safety of low-altitude traffic operation is improved.

[0079] 3. Achieve efficient integration of multi-modal resources.

[0080] Integrate unmanned devices and other low-altitude resources operating at low altitudes to form a unified low-altitude three-dimensional traffic road network, providing support for multi-tasks and multi-scenarios, and solving the problems of resource isolation and coordination.

[0081] 4. Improve the intelligence level of the system.

[0082] With the help of algorithms and digital twin technology, realize the intelligent management of traffic planning and operation, and provide support for quickly responding to emergencies and optimizing the operation efficiency of the low-altitude traffic system.

[0083] (1) Achieve operation optimization through digital twin: Create a digital twin model, and use multi-source data (such as low-altitude devices, geographic information, traffic flow) to construct a digital twin system with real-time mapping, and realize the digital twin model of the three-dimensional traffic road network.

[0084] The technologies involved in constructing a digital twin system with real-time mapping include:

[0085] (1.1.1) Select a platform for building a real-time mapping digital twin system. To improve real-time performance and data processing efficiency, an edge computing platform is paired to assist in real-time data preprocessing near the data source.

[0086] (1.1.2) Data collection and processing. Obtain real-time data through low-altitude devices (such as drones, unmanned boats, payload devices, etc.), geographic information systems (GIS), traffic flow sensors, meteorological data, remote sensing data, etc. After cleaning, denoising, and normalizing, it is used for subsequent model construction.

[0087] (1.1.3) Import the collected visible light data into 3D modeling software or GIS tools to create a 3D model of the digital twin within the traffic road network. Use the support vector machine and neural network algorithms of machine learning to learn and construct traffic patterns from the data.

[0088] (1.1.4) Use real-time data stream processing algorithms and deep learning algorithms to process real-time data and perform prediction optimization.

[0089] (1.1.5) Design an application model according to the actual implementation scenario. Import the data after real-time processing for training, and evaluate the model performance through cross-validation to prevent overfitting, and verify it on real-time data to ensure the accuracy of its actual application.

[0090] (1.1.6) Integrate the trained model and data into the existing visualization system in the form of an API interface and merge it with the digital twin model for display, facilitating user understanding and interaction.

[0091] (1.1.7) Perform real-time data synchronization through a message subscription system and update the application model in a regularly trained manner.

[0092] (1.1.8) Establish a continuous feedback mechanism. Through the active interaction of users and system log analysis, continuously improve the model and algorithm to achieve the optimization of the three-dimensional traffic road network.

[0093] Real-time status monitoring: By flowing real-time data into the twin model, master the traffic status, including the position, speed, congestion area, and emergencies of the aircraft.

[0094] Scenario Prediction: Utilize historical and real-time data to simulate possible traffic flow changes, node pressures, and emergencies. Pre-test various operation plans (such as detour path selection, traffic flow allocation strategies) in a virtual environment to find the optimal strategy. Dynamically adjust the strategy by inputting real-time traffic data into the prediction model and updating the prediction results; according to the latest prediction results, dynamically adjust the management strategy of the three-dimensional traffic road network to ensure that the system is always in an optimal state. Convert the simulation results into real-time operation instructions to optimize low-altitude traffic flow allocation and scheduling decisions.

[0095] The tested operation plans include:

[0096] (1.2.1), Detour Path Selection: Shortest Path Detour, Least Congested Path Detour, Path Detour Considering Both Time and Distance;

[0097] (1.2.2), Traffic Flow Allocation Strategy: Uniformly Allocate Traffic to All Available Paths, Dynamically Allocate Traffic According to Real-Time Traffic Conditions, Preferentially Allocate Traffic to High-Capacity Paths;

[0098] (1.2.3), Emergency Response: Immediately Close Affected Nodes and Reallocate Traffic; Gradually Reduce the Traffic of Affected Nodes and Gradually Guide It to Alternative Paths; Keep Affected Nodes Open but Restrict Traffic Entry.

[0099] The test steps include:

[0100] (1.3.1), Data Preparation: Historical data and real-time data, where historical data includes generated traffic flow, node pressures, emergencies, etc.; real-time data includes data obtained in real-time by sensor devices.

[0101] (1.3.2), Build a Simulation Environment: Use a digital twin system to construct a virtual three-dimensional traffic road network, input data, simulate the operation status of the real-world three-dimensional traffic road network, and dynamically change the environment by setting different traffic parameters.

[0102] (1.3.3), Test of Operation Plans: Plan 1, Apply the shortest path detour strategy in the virtual environment and record traffic flow changes and node pressures. Plan 2, Apply the least congested path detour strategy and record relevant data. Plan 3, Apply the path detour strategy considering both time and distance and record relevant data. Repeat the above steps to test all traffic flow allocation strategies and emergency response plans.

[0103] The methods for determining the optimal strategy include:

[0104] (1.4.1), Performance Metrics:

[0105] Traffic Flow: Total Flow, Average Flow, Peak Flow.

[0106] Node pressure: The pressure index, maximum pressure, and average pressure of each node.

[0107] Emergency response time: The time from the occurrence of an event to the system response.

[0108] User satisfaction: Obtain users' satisfaction with the traffic situation through surveys or feedback mechanisms.

[0109] (1.4.2), Data analysis:

[0110] Comparative analysis: Compare the performance indicators under different scenarios to find the optimal solution.

[0111] Sensitivity analysis: Analyze the impact of changes in different parameters on the performance of the scenario to ensure the robustness of the scenario.

[0112] (1.4.3), Decision support system:

[0113] Multi-objective optimization: Considering multiple performance indicators comprehensively, use multi-objective optimization algorithms (such as NSGA-II) to find the optimal strategy.

[0114] Expert review: Invite traffic management experts to review the test results and provide professional opinions.

[0115] (2) Auxiliary optimization.

[0116] Path planning optimization: Based on path optimization algorithms such as Dijkstra, A*, or reinforcement learning, combined with real-time geographical data, dynamically plan the optimal flight path. Considering external factors such as weather, traffic flow, and obstacles, adjust the route priority to avoid resource waste.

[0117] Traffic flow scheduling: Use traffic flow prediction algorithms (such as LSTM or Transformer models) to predict traffic flow peaks in advance and dynamically allocate resources. For specific periods, adjust the hierarchical diversion strategy of the airspace to balance the traffic flow pressure in each airspace.

[0118] Anomaly detection and response: Conduct anomaly event detection, such as weather changes, equipment failures, or congestion. Generate corresponding emergency response strategies according to the detection results to optimize the flight path and resource allocation of aircraft.

[0119] (2.1) Use Geographic Information System (GIS) to obtain geographical data such as terrain, buildings, and roads, combined with real-time weather, traffic flow, obstacles, etc. obtained through sensors, satellites, drones, etc., and use them after preprocessing such as data cleaning, fusion, and standardization.

[0120] Weather factors: Real-time weather data (such as wind speed, rainfall, etc.). Evaluate the impact of weather on the flight path and adjust the path priority. Dynamically adjust the flight path according to weather changes.

[0121] Traffic flow factor: Real-time traffic flow data. Evaluate the impact of traffic flow on route selection to avoid congestion. Dynamically adjust route priorities according to traffic flow changes.

[0122] Obstacle factor: Real-time obstacle data (such as buildings, trees, other unmanned devices). Evaluate the impact of obstacles on the route and plan detour routes. Dynamically adjust the operating route according to obstacle changes.

[0123] 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 the route planning results.

[0124] Performance optimization: Use technologies such as multi-threading and GPU acceleration to improve the algorithm calculation efficiency. Cache common route planning results to reduce repeated calculations.

[0125] Continuous improvement: Establish user feedback and system logs to continuously improve the algorithm and model. Regularly update the model to adapt to new geographical data and external factors.

[0126] (2.2) The core of anomaly detection and response lies in quickly identifying abnormal events and generating effective emergency response strategies. The specific methods for generating emergency response strategies include:

[0127] Anomaly detection:

[0128] (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.

[0129] (2.2.2) Anomaly classification includes: Weather anomalies: such as storms, fog, etc. Equipment failures: such as insufficient battery power, equipment failures, abnormal load data, etc.

[0130] The specific content of adjusting routes and resources through emergency response strategies includes:

[0131] (2.3.1) Route adjustment: Detour route, generate a detour route to avoid the abnormal area. Spare route, pre-plan multiple spare routes and select the optimal route according to the abnormal situation. Dynamic adjustment, dynamically adjust the route according to real-time data to ensure that the aircraft is always on the optimal route.

[0132] (2.3.2)Resource Allocation: Spare devices are configured at critical nodes to ensure quick replacement in case of device failures. Resource scheduling is carried out according to abnormal situations, dynamically scheduling resources (such as increasing the number of aircraft or readjusting aircraft tasks). Resource optimization is performed using optimization algorithms (such as linear programming and integer programming) to optimize resource allocation and ensure maximum resource utilization.

[0133] (2.3.3)Priority Adjustment: Task priority is adjusted according to the impact of anomalies (such as giving priority to urgent tasks and delaying non-urgent tasks). Path priority is adjusted according to the path conditions (such as preferentially selecting safe paths and avoiding high-risk paths). Resource priority is adjusted according to the resource conditions (such as preferentially allocating resources to high-priority tasks).

[0134] (3)Event Response Mechanism

[0135] Real-time Response to Emergencies: Through the digital twin model, the impact of emergencies on the traffic system is simulated in real time (such as node interruptions or traffic surges).

[0136] According to the events and corresponding emergency specifications, emergency specifications are triggered dynamically (such as flight restricted airspace and reallocation of flight paths).

[0137] (4)Intelligent Operation Management

[0138] Real-time Visualization of Traffic Status: The low-altitude traffic conditions are dynamically presented on the digital twin model to intuitively identify key issues.

[0139] Multi-modal Fusion Regulation: Using a rule engine, the operation specifications are dynamically adjusted, such as airspace priority and aircraft speed.

[0140] Resource Integration: Intelligent allocation of low-altitude resources (such as infrastructure, equipment, and aircraft) is carried out to avoid idleness or overload.

[0141] 5. Adapt to complex task requirements.

[0142] Through structured management and multi-level planning, it flexibly adapts to different types of task scenarios such as commercial tasks, public safety, and emergency response, meeting the diverse needs of low-altitude industry equipment traffic.

[0143] 6. Promote the development of the low-altitude economy.

[0144] Provide efficient and standardized technical support for various enterprises in the low-altitude economy field, improve the overall industrialization level of the low-altitude field, and accelerate the construction and popularization of the low-altitude three-dimensional traffic network. Description of the Drawings

[0145] Figure 1 It is a flowchart of the method for dividing the low-altitude three-dimensional traffic road network according to the embodiment of the present invention.

[0146] Figure 2 It is the topology diagram of the low-altitude three-dimensional traffic road network division method of the embodiment of the present invention.

[0147] Figure 3 It is the logical architecture diagram of the dynamic division of the embodiment of the present invention.

[0148] Figure 4 It is the logical architecture diagram of the stratification of the three-dimensional traffic road network of the embodiment of the present invention.

[0149] Figure 5 It is the logical architecture diagram of the traffic flow diversion of the three-dimensional traffic road network of the embodiment of the present invention.

[0150] Figure 6 It is the logical architecture diagram of the nodes of the three-dimensional traffic road network of the embodiment of the present invention.

[0151] Figure 7 It is the logical architecture diagram of the classification of the three-dimensional traffic road network of the embodiment of the present invention.

[0152] Figure 8 It is the logical architecture diagram of the classification of the three-dimensional traffic road network of the embodiment of the present invention.

[0153] Figure 9 It is the logical architecture diagram of the traffic lights of the three-dimensional traffic road network of the embodiment of the present invention.

[0154] Figure 10 It is the logical architecture diagram of the operation specifications of the embodiment of the present invention. Detailed implementation manners

[0155] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with embodiments and the accompanying drawings.

[0156] Refer to Figure 1 、 Figure 2 , this embodiment provides a low-altitude three-dimensional traffic road network division method and system. By implementing a series of low-altitude three-dimensional traffic road network division contents and combining low-altitude multi-modal resources, a low-altitude three-dimensional traffic road network and its intelligent operation specifications for unmanned devices are constructed.

[0157] Figure 1 It is the flowchart of the low-altitude three-dimensional traffic road network division method of the embodiment of the present invention, showing the specific process of the method in this embodiment. Figure 2 It is the topology diagram of the low-altitude three-dimensional traffic road network division method of the embodiment of the present invention, which is a star structure. To complete the division of the three-dimensional traffic road network, five basic elements of data collection, dynamic division, multi-modal data integration, structured management, and standardized management are required; the method in this embodiment includes:

[0158] S1, regional data collection and data processing, and the specific process includes:

[0159] S11. Confirm the regional scope of the low-altitude three-dimensional traffic road network division, summarize the unmanned equipment and task information within this region, and design the data content to be collected based on the task information. The data content includes road surface information for road inspections, terrain information for forest inspections, water body information for water conservancy inspections, building information for urban management inspections, environmental information for environmental protection inspections, etc.

[0160] S12. Use computing equipment to perform data cleaning, format standardization, data conversion, and classified storage on the collected data.

[0161] S13. Evaluate whether the processed data meets the requirements for the division of the three-dimensional traffic road network. The data that meets the requirements will undergo subsequent dynamic division and multimodal resource integration, while the data that does not meet the requirements will be reprocessed through the data collection and processing procedures of steps S11 - S12 until the processed data meets the requirements.

[0162] S2. Perform dynamic division on the processed data to obtain the divided logical data; simultaneously, perform multimodal resource integration by integrating low-altitude air traffic data, low-altitude road surface data, and low-altitude water body data to form three-dimensional traffic data.

[0163] Dynamic division:

[0164] 1. Dynamic data collection and update

[0165] Adopt real-time data collection technology to collect and update the data generated by unmanned equipment, including sensors, drones, and ground equipment, etc., and monitor the traffic situation in real time. Continuously update the data according to factors such as road traffic flow, weather changes, and emergencies to ensure the timeliness and accuracy of the data. For example, real-time data such as road traffic flow, vehicle speed, and traffic congestion are used to dynamically update and adjust the current data of the traffic road network.

[0166] 2. Data cleaning

[0167] Clean the data from different sources, eliminate error and redundant information, and perform standardization processing to ensure the integrity and consistency of the data. Use big data technology to fuse various types of data such as road traffic flow, accident information, and equipment status in multiple dimensions to form a comprehensive three-dimensional traffic data, providing an accurate basis for subsequent analysis.

[0168] 3. Real-time road network state modeling

[0169] Regularize the data into six real-time state models (stratification, diversion, node, grading, classification, traffic lights) to reflect the current traffic state of each node or region.

[0170] The update frequency of each model can be dynamically adjusted according to traffic conditions. For example, the update frequency can be increased during peak hours.

[0171] 4. Dynamic data support

[0172] Provide six types of data support and feedback for unmanned devices.

[0173] 5. Dynamic optimization

[0174] Analyze the laws of dynamic data and optimize the road network division plan. Introduce a self-learning mechanism to continuously improve response and accuracy.

[0175] 6. Dynamic data display and interaction

[0176] Build a real-time interactive interface to dynamically present key indicators such as traffic status, diversion results, and node pressure, assisting decision-makers to respond quickly.

[0177] See Figure 3 , and the content of dynamic division includes the following types:

[0178] Stratification of the three-dimensional traffic road network, such as Figure 4 shown. Preferably, the 1000-meter low-altitude area is divided from high to low into a long-distance layer (100 meters - 1000 meters), a medium-distance layer (50 meters - 100 meters), a short-distance layer (10 meters - 50 meters), a ground layer (below 10 meters to the surface of the terrain), and an underwater layer (regardless of altitude, with the water surface of the terrain as the boundary).

[0179] Diversion of the three-dimensional traffic road network, such as Figure 5 shown. According to the use of the channels, it is divided into operation channels, dispatching channels, task channels, and emergency channels. Among them, the operation channel is the channel where unmanned devices are normally operating between nodes, the dispatching channel is the channel where unmanned devices are located during non-task transfer, the task channel is the channel where unmanned devices are located when performing tasks, and the emergency channel is the fast channel for unmanned devices in special emergency situations. (The channels are divided into four, namely up, down, left, and right. Among them, left and right are the forward and reverse operation flows, and up and down are the geographical directions of east-west and north-south).

[0180] Nodes of the three-dimensional traffic road network, such as Figure 6 shown. Nodes can be divided into normal nodes and temporary nodes. Normal nodes are further divided into task nodes and exchange nodes, and temporary nodes are also divided into activity nodes and exchange nodes. Among them, the task node is a small area temporarily occupied by unmanned devices when performing normal task items, and the activity node is a small area temporarily occupied by unmanned devices when performing temporary activities. The areas of the task nodes and activity nodes are exchanged through the exchange nodes.

[0181] Grading of the three-dimensional traffic road network, such as Figure 7As shown in the figure, the classification by priority order can be divided into P level, C level, S level, and G level. Among them, the P level has the lowest priority, and the G level has the highest priority. Devices operating on the high-priority road network have the right of way. The P level can only operate non-commercial unmanned devices, the C level can only operate enterprise-level unmanned devices, the S level can only operate government-level unmanned devices, and the G level is a controlled-level unmanned device.

[0182] Classification of the three-dimensional transportation road network, such as Figure 8 As shown in the figure, the classification includes vector type, plane type, three-dimensional type, 4DT type, and special type. Among them, the vector type is the one-way straight running direction from point to point; the plane type is the free running area within a small range (within 10 meters) up and down; the three-dimensional type is the free running area within a large range (more than 10 meters) up and down; the 4DT type adds time restrictions to the plane type and three-dimensional type and can only operate within the specified time; the special type can be any of the previous four types, but can only pass through the bound unmanned device.

[0183] Traffic lights of the three-dimensional transportation road network, such as Figure 9 As shown in the figure, the traffic lights are divided into red lights and green lights. Red lights mean stop, and green lights mean go.

[0184] S3, Structuring: Structurally process the three-dimensional transportation data for dynamically divided logical data and multimodal resource integration to obtain the structured result of the low-altitude three-dimensional transportation road network.

[0185] 1. Structured processing of dynamically divided data

[0186] Objective: Convert the collected data related to the low-altitude three-dimensional transportation road network (such as devices, locations, roads, etc.) into a key-value based processable structured data form for subsequent analysis, query, and application.

[0187] 2. Data standardization

[0188] Unify data from different sources (such as traffic cameras, drone images, sensors) into a standard format, such as unifying timestamps, spatial coordinate systems, and data units.

[0189] 3. Hierarchical classification

[0190] According to the dynamic division logic of the low-altitude road network (layering, flow diversion, nodes, grading, classification, traffic lights), label the data to distinguish different types of data streams.

[0191] 4. Data modeling

[0192] Represent the low-altitude three-dimensional transportation road network in the form of graph data. For example, the layering is represented from height a to height b, and 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, passable time, etc.).

[0193] Build a multi-dimensional data model, dynamically incorporate time and space data into the structured results, and facilitate querying the three-dimensional traffic road network data at a certain time and in a certain area.

[0194] 5. Multi-modal resource integration

[0195] Multi-modal resources include videos, images, unmanned devices, and sensor data, etc. Extract objects from videos and images (such as vehicles, roads, bridges, traffic lights). And align data from different modalities, for example, align video frames and sensor streams based on timestamps, and integrate aerial images and unmanned device data based on geographical coordinates.

[0196] 6. Generation and management of structured results

[0197] Use standardized data formats (such as GeoJSON, Shapefile) to represent spatial information, and combine tabular data to represent dynamic indicators.

[0198] Database storage:

[0199] The database architecture adopts a hybrid database system, which supports both graph data (storing the traffic road network structure) and relational data (storing dynamic attributes) at the same time.

[0200] S4, Standardization: Calculate the corresponding action specifications based on the structured results. As Figure 10 shown, Figure 10 This is the logical architecture diagram of the action specifications in the embodiments of the present invention, showing that the key components of the action specifications include:

[0201] Operation specifications: Formulate operation specifications, including but not limited to pre-start inspection of unmanned devices, operating conditions, operation monitoring, and fault tolerance mechanisms. Among them, the pre-start inspection includes that the device hardware is intact, the sensor works normally, the battery power is sufficient, the communication module is normal, the self-check program passes, and the system is free of faults, etc. The operating conditions confirm whether the environment, road network and other conditions meet the operating requirements. The operation monitoring monitors the device attitude, speed, remaining battery power, communication signal strength, etc. in real time. The fault tolerance mechanism includes standby stop, fault detection, safety mode, etc.

[0202] Stop specifications: Formulate stop specifications, including but not limited to manual stop, automatic stop, and safe parking. Among them, the manual stop is that the remote control terminal sends a stop instruction, and the device immediately responds and safely shuts down. The automatic stop is to complete the automatic stop after the task is executed. Safe parking includes return flight after stopping, parking in a specific area, and battery charge and discharge management, etc.

[0203] Waiting specifications: The unmanned device is in a standby state and waits in a low-power state in the temporary parking area of the switching node.

[0204] Switching specifications: That is, the access specifications of the switching node, and the principle of first-in first-out and sequence switching is executed.

[0205] Scheduling Specification: Scheduling is carried out based on principles such as task allocation, conflict avoidance, and coordinated operation, and the device location information during the scheduling process is fed back in real time.

[0206] Emergency Specification: Emergency specifications are formulated, including device self-emergency and emergency tasks. The device is equipped with fault detection and emergency recovery (such as emergency stop and automatic return), and the emergency tasks are strictly operated according to the task requirements.

[0207] The calculation ideas and formulas for the operation specifications are as follows:

[0208] 1. Operation Specification

[0209] Calculation Idea: The conditions for allowing operation on a road section include that the traffic flow has not reached 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 depending on the actual environment. The preferential passage path for the road section and the dynamic node entry and exit plan are obtained.

[0210] 2. Stop Specification

[0211] Calculation Idea: Based on the dynamically divided data status (abnormal, faulty) and path interruption data, the stop conditions include: road section interruption (such as abnormal weather, obstacle detection); node failure or stop signal (such as device failure); traffic flow dynamically exceeding the threshold (such as exceeding the safety carrying capacity); setting a safe stop area and an alternative detour plan based on dynamic time prediction; dynamically prompting the stop area and the detour plan.

[0212] 3. Waiting Specification

[0213] Calculation Idea: The waiting rule is based on the signal light time and the node capacity, and the waiting time is dynamically adjusted.

[0214] 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 flow.

[0215] 4. Exchange Node Entry and Exit Specification

[0216] Calculation Idea: Road network node connectivity and node capacity data.

[0217] Exchange Priority Reference Formula: P 交换 = F 流量 * F 节点距离 − F 节点负载, Adjust the in - out direction of nodes according to the priority to avoid congestion in the switching nodes.

[0218] 5. Scheduling Specification

[0219] Calculation idea: Global traffic distribution, node load rate.

[0220] Dynamic adjustment scheduling frequency reference formula: F 调度 =Q 总流量 / T 预测时间窗口 , to obtain the optimal scheduling path and frequency.

[0221] 6. Emergency Specification

[0222] Calculation idea: Abnormal detection data (weather, obstacles, equipment failures).

[0223] Set the emergency level (such as low, medium, high), and define emergency actions in different situations.

[0224] Generate a detour path based on dynamic information.

[0225] Emergency response time reference formula: T 应急 =D 目标距离 / V 应急速度 , emergency response plan and path planning.

[0226] S5. Operation: Based on the above - mentioned structured and standardized results, guide the overall operation of low - altitude unmanned devices.

[0227] This embodiment also discloses a low - altitude three - dimensional traffic road network division system, including:

[0228] Unmanned devices and multi - source sensing devices, used to collect data for the area of the low - altitude three - dimensional traffic road network division;

[0229] Computing devices, used to perform real - time processing on the collected data;

[0230] Dynamic division module, used to perform dynamic division on the processed data to obtain the divided logical data;

[0231] Multi - modal resource integration module, used to integrate low - altitude air traffic data, low - altitude road surface data, and low - altitude waterway data to obtain three - dimensional traffic data;

[0232] Structured management module; used to perform structured design and management on the logically divided data and the three - dimensional traffic data integrated by multi - modal resources; obtain the structured results of the low - altitude three - dimensional traffic road network;

[0233] Standardized management module; used to calculate the corresponding action specifications based on the structured results.

[0234] The above embodiments are only for illustrating the technical idea of the present invention, and the protection scope of the present invention cannot be limited thereby. Any modification made on the basis of the technical solution according to the technical idea proposed by the present invention falls 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 low-altitude air traffic data, low-altitude road data, and low-altitude water data to form three-dimensional traffic data; Dynamic partition types include: layering, diversion, node, grading, classification, and traffic light; 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 stop, green lights go; 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: 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.

5. 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 specifications: The conditions under which a road section is allowed to operate include that the traffic volume does not reach the maximum threshold and the node connectivity is normal; W operation = a1*f traffic capacity + a2*f traffic share − a3*f congestion index; 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; Twait = (C current node capacity / Q arrival flow) × F signal light time; Switching node entry and exit specifications: P exchange = F traffic * F node distance − F node load, adjust the node entry and exit direction according to the priority to avoid congestion of the exchange node; Scheduling specifications: dynamically adjust the scheduling frequency; F scheduling = Q total flow / T prediction time window; Emergency specifications: Set emergency levels based on abnormal detection data and define emergency actions in different situations; Generate detour paths based on dynamic information; Emergency response time: Temergency = Dtarget distance / Vemergency speed, emergency response plan and path planning.

6. The method for dividing a low-altitude three-dimensional traffic network according to any one of claims 1, 4 and 5, 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; The exchange node entry and exit specifications implement the first-in-first-out and sequential exchange principles; 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.

7. 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.

8. 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 7 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