Low-altitude unmanned aerial vehicle assisted peak period urban intersection coordination control system and method

Abstract

The invention relates to a low-altitude unmanned aerial vehicle assisted peak period urban intersection coordination control system and method, and belongs to the field of traffic control. The system comprises an equipment deployment layer, a sensing layer, a decision-making layer and a control layer. Based on infrastructures deployed by the equipment deployment layer, the unmanned aerial vehicle cruises, shoots and preliminarily analyzes an image, and transmits structured data to the edge computing equipment; the sensing layer determines intersection state information and transmits the intersection state information to the decision-making layer, and the decision-making layer generates a signal timing strategy and a regional scheduling scheme; and the control layer controls the signal lamp according to the information and feeds back an execution result, and the system can perform self-adaptive optimization and early warning. The method comprises the steps that the unmanned aerial vehicle acquires and processes an image, transmits data to a decision-making layer to optimize timing, issues a strategy to the signal lamp controller and feeds back an execution effect. According to the scheme, the sensing range and precision can be enhanced, the communication load is reduced, the signal timing is optimized, multi-intersection collaboration is supported, a closed-loop control system is constructed, deployment is flexible, and adaptability is high.

Description

Technical Field

[0001] The present invention belongs to the field of traffic control and relates to a low-altitude unmanned aerial vehicle (UAV)-assisted urban intersection coordination control system and method during peak hours. Background Art

[0002] Traffic management at urban intersections is one of the most complex and critical links in the urban traffic system. With the continuous increase in the number of motor vehicles in cities, especially during the morning and evening rush hours on weekdays, intersections have become traffic bottlenecks, and traffic congestion is particularly prominent. In order to alleviate congestion and improve traffic efficiency, various urban intersection control technologies have been developed and applied one after another. At present, the mainstream traffic signal control methods mainly include the following categories: (1) Adaptive signal light control technology: Real-time collection of traffic flow data through ground sensors or cameras, dynamic adjustment of traffic light duration. This method can alleviate congestion to a certain extent, but the response time is limited by the coverage range and processing speed of ground sensing equipment, and it is difficult to adapt to emergencies or large-scale traffic changes. (2) Green wave band control technology: Coordinate multiple signal lights on continuous sections of road, set a "green wave" traffic strategy, and improve the traffic efficiency of main roads. However, this technology has high requirements for vehicle speed consistency, is difficult to implement, and has limited effect in areas with dense intersections. (3) Vehicle speed guidance and induction system: Guide vehicles to choose reasonable routes and speeds through information release platforms (such as variable information boards, APPs, etc.). This system relies on the cooperation of drivers, making it difficult to achieve forced intervention and rapid response. (4) On-site traffic control by traffic police: During peak hours or emergencies, traffic is manually directed by traffic police. Although it has a certain degree of flexibility, it has high labor costs, low efficiency, and poor timeliness, making it difficult to achieve large-scale coordination.

[0003] In summary, while existing urban intersection traffic management technologies have achieved some success, they still face significant shortcomings in addressing sudden congestion during peak hours and improving overall coordination efficiency. In particular, they lack comprehensive situational awareness and dynamic decision-making capabilities from an aerial perspective, making it impossible to achieve rapid response and coordinated control at the regional level.

[0004] Existing technology: A method for optimizing intersection signal timing based on drone-generated aerial video capture uses drones to capture intersection video information. Deep learning and image processing algorithms extract vehicle trajectory data and road structure data to assist in optimizing signal timing. The method focuses on analyzing historical and real-time traffic flow information through image recognition to support traffic management.

[0005] Another existing technology, a drone-based video-based method for detecting dynamic intersection signal timing schemes, uses drone-generated traffic imagery to extract and process information by building a dataset and neural network model. This method is used to detect and assist in optimizing signal timing. This method emphasizes automated recognition and timing analysis driven by video data.

[0006] Both solutions are centered around drone aerial photography and image analysis, enabling automated processing of data collection and timing detection, and improving data acquisition efficiency and accuracy.

[0007] Although the above-mentioned existing technologies have introduced drones into urban intersection management, the following deficiencies still exist: They are limited to perception and auxiliary analysis, and lack a closed-loop control system: Existing patents focus on drone video acquisition and timing scheme detection, lack the ability to directly intervene in the signal control terminal, and are unable to achieve an integrated linkage of "analysis-decision-control". Perception capabilities tend to be processed offline and lack real-time performance: Image data usually needs to be transmitted back to the center for in-depth processing, and the response to sudden congestion incidents during peak hours is not flexible and timely enough. A systematic and layered architecture has not been established: The existing solutions do not clearly divide the functions of each level of perception, decision-making, and control, and lack a comprehensive coordination mechanism based on the system architecture. Summary of the Invention

[0008] In view of this, the purpose of the present invention is to provide a low-altitude drone-assisted urban intersection coordination control system and method during peak hours, aiming to achieve real-time dynamic optimization and regional collaborative control of traffic lights, build a "perception-decision-control" closed-loop architecture, and realize real-time full-process linkage response; adopt a low-altitude drone + edge computing + traffic light intelligent controller fusion architecture to ensure response speed and coverage; support coordinated scheduling of regional intersections rather than isolated optimization to achieve optimal global traffic efficiency; act as an "electronic traffic police" to support immediate intervention and emergency timing adjustment in emergencies (such as accidents, temporary construction), and improve the robustness and adaptability of intersections.

[0009] In order to achieve the above object, the present invention provides the following technical solutions:

[0010] On the one hand, a low-altitude drone-assisted urban intersection coordination control system during peak hours is proposed. The system includes: equipment deployment layer, perception layer, decision layer and control layer. The perception layer, decision layer and control layer implement corresponding data collection and data processing based on the basic settings deployed in the equipment deployment layer;

[0011] The device deployment layer includes the deployment of low-altitude drones, intelligent traffic light controllers, communication modules, and edge computing devices. The low-altitude drones and edge computing devices communicate through the communication modules. The drones obtain video information and perform front-end processing tasks. The edge computing devices then extract structured data and upload it to the perception layer.

[0012] The perception layer determines intersection status information based on structured data and passes it to the decision layer, which generates real-time signal timing strategies and regional scheduling plans based on the intersection status information.

[0013] The control layer controls traffic lights based on the real-time signal timing plan and regional scheduling plan output by the decision layer, and the execution results of the traffic lights are fed back to the perception layer through ground equipment or drones; in the event of continued congestion or strategy execution failure, the system triggers adaptive optimization and early warning mechanisms to achieve dynamic correction.

[0014] Furthermore, the low-altitude drones at the equipment deployment layer are equipped with high-definition cameras, embedded computing platforms, communication modules, and flight control systems. They use high-definition cameras to conduct aerial patrols and photography of intersections and their surrounding areas. The embedded computing platform then performs preliminary image analysis, including vehicle detection and recognition, and generates localized structured data.

[0015] The intelligent traffic light controller includes an external communication interface, remotely receives a control strategy through the external communication interface, and dynamically adjusts the duration and phase sequence of the traffic lights according to the received control strategy;

[0016] The edge computing device obtains the structured information transmitted by the drone through the communication module, and completes time analysis and control strategy generation based on the structured information;

[0017] The communication module uses high-speed communication technology to support data synchronization between drones and MEC nodes, and information transmission in the case of command distribution between MEC and traffic lights.

[0018] Furthermore, the perception layer performs corresponding operations based on the various infrastructures deployed by the basic equipment layer. The perception layer includes: drone front-end video perception and preliminary analysis module, multi-source data fusion module, and event detection module;

[0019] The drone's front-end video perception and preliminary analysis module is deployed in the drone's embedded computing platform. It completes vehicle target detection, traffic density estimation, and initial screening of abnormal behavior based on the acquired video information, and transmits the structured data to the edge computing node.

[0020] The multi-source data fusion module integrates the structured perception data from different UAVs and the perception data from various ground perception sources;

[0021] The event detection module determines whether an emergency occurs based on the structured data stream and rule model and generates a description of the current intersection status.

[0022] Furthermore, in the UAV front-end video perception and preliminary analysis module, the structured data D output by the UAV l l (t) is expressed as:

[0023] D l (t) = {n veh ,ρ lane ,E abn}

[0024]

[0025] Where: n veh is the number of vehicles, indicating the number of vehicles on the road per unit time; ρ lane is the lane occupancy rate, reflecting the degree of traffic congestion; L is the total length of the observed road section, L i is the length of vehicle i; E abn It is a sign of abnormal events, such as driving against traffic and traffic accidents;

[0026] The structured perception data of different perception sources after integration by the multi-source data fusion module is:

[0027] D k (t)={m veh ,D l (t)}

[0028] Among them, m veh The number of vehicles released upstream of the intersection;

[0029] The event detection module calculates the lane occupancy rate ρ lane The threshold value is used to determine whether there is congestion at the current intersection and record it; the congestion determination threshold is set to ρ thresh , if ρ lane >ρ thresh , it is judged to be congested, otherwise it is judged to be non-congested; if the above detection result belongs to E abn , then record abnormal events; structured perception data D from different perception sources k (t) is used to guide the decision-making layer to specify the signal timing plan.

[0030] Furthermore, the decision-making layer is responsible for generating traffic control strategies, which include a timing optimization module and a strategy evaluation and adaptive feedback module. When the traffic lights are rotated according to a preset plan, the timing optimization module analyzes the current traffic state based on perception data and uses a dynamic timing algorithm to generate the optimal timing strategy for the intersection. The duration of each phase is adjusted according to the optimal timing strategy.

[0031] The strategy evaluation and adaptive feedback module continuously evaluates the effectiveness of the current strategy and makes adaptive corrections based on signal control results and real-time traffic feedback.

[0032] Furthermore, in the configuration optimization module, the green light duration s of the traffic light corresponding to lane j under the optimal timing strategy is j The calculation method is:

[0033]

[0034] Among them, n vehIndicates the number of vehicles in the queue; s m is the distance traveled by the front vehicle when the rear vehicle starts; s1 is the distance traveled by the vehicle after it starts and accelerates to the required speed; v d and t loss represent the expected speed and loss time respectively; is the starting acceleration; t is the system time; L i is the vehicle length; m veh Indicates the number of vehicles input from upstream, t add Indicates the delay time of upstream vehicles; Δτ k represents the strategy adjustment amount; Δs * is the safety distance, and the intelligent driver model is used to describe the car following behavior, where the predicted acceleration Defined as:

[0035]

[0036]

[0037] Among them, a max 、v max , s0, σ and b represent the maximum acceleration, maximum speed, congestion distance, safety headway and expected acceleration respectively; Indicates the instantaneous speed of the vehicle; Δd n (t), Δv n (t), Δs * They represent the distance, speed difference, and safety distance between vehicle n and vehicle n-1 respectively.

[0038] Furthermore, the strategy evaluation and adaptive feedback module evaluates the effectiveness of the strategy according to the following formula:

[0039] E=n veh +α1m veh -n real

[0040] Among them, E represents the difference between the target number of vehicles passing through and the actual number of vehicles passing after a green light is released, α1 represents the weight of upstream vehicles, and n real Indicates the actual number of vehicles passing through;

[0041] Then, the signal timing strategy is modified based on the strategy error feedback:

[0042] Δτ k (t) = γ·E

[0043] Where Δτ k (t) is the strategy adjustment amount, and γ is the feedback adjustment coefficient.

[0044] Furthermore, the control layer includes an instruction issuing module, a traffic light execution module and a feedback confirmation module. Among them, the instruction issuing module deployed in the edge computing device sends the timing strategy generated by the decision-making layer to the traffic light controller of the corresponding intersection in real time through a wireless link. The traffic light execution module in the traffic light controller automatically completes the traffic light cycle, phase sequence and phase adjustment according to the timing strategy received by the signal controller; the feedback confirmation module in the traffic light controller feeds back the execution results to the edge computing node, and performs closed-loop verification through drone perception to achieve closed-loop optimization of the strategy.

[0045] On the other hand, a low-altitude drone-assisted coordinated control method for urban intersections during peak hours is also proposed. The method is based on the aforementioned low-altitude drone-assisted coordinated control system for urban intersections during peak hours. The method includes:

[0046] Obtain intersection video images through drones, perform front-end perception and data processing at the perception layer, and pass the processed structured data to the decision layer;

[0047] The edge computing device receives traffic status data from different drones, builds an intersection status matrix, and determines whether the lane occupancy exceeds a threshold, triggering a strategy optimization process if so;

[0048] Signal timing optimization and coordinated scheduling are performed at the decision-making level. Based on the perception data, the current traffic status is analyzed and the optimal timing strategy for the intersection is generated using a dynamic timing algorithm.

[0049] The generated signal control strategy is sent to the traffic light controllers at each intersection; the controllers adjust the light control parameters in real time, and the controller execution status is fed back to the edge computing device through the status confirmation module, and the execution effect is perceived and verified by the drone.

[0050] The beneficial effects of the present invention are:

[0051] (1) Enhanced sensing range and accuracy: By deploying low-altitude drone systems to conduct dynamic aerial photography at intersections and surrounding airspace, the system overcomes the blind spots of traditional ground-based cameras and induction coils, enabling large-scale, high-resolution, and multi-angle real-time monitoring of traffic flows. This system, combined with ground-based sensing sources, builds an air-ground fusion sensing system, significantly improving the integrity and accuracy of traffic information at urban intersections.

[0052] (2) The communication load is greatly reduced. The present invention adopts a "cooperative processing architecture" and integrates an embedded computing module on the drone side to complete the local extraction of structured data such as traffic flow, queue length, and abnormal behavior, and only uploads necessary information to the edge computing device, thereby effectively reducing the communication load, avoiding network congestion caused by large-scale high-definition video transmission, and improving the overall real-time performance and stability of the system.

[0053] (3) Optimize signal timing: By using edge computing devices to aggregate various structured data in real time, combining current traffic conditions with historical data, and using dynamic timing optimization algorithms to quickly generate optimal signal control strategies. The system has rapid response capabilities and can adjust traffic light timing based on emergencies (such as accidents and abnormal traffic flow), alleviating queue pressure and improving intersection efficiency.

[0054] (4) Supporting coordinated control of multiple intersections; the decision-making layer of the present invention supports signal timing coordination for multiple adjacent intersections to avoid phenomena such as "green light conflicts" or "traffic conflicts", thereby improving the continuity and stability of regional traffic flow, and is particularly suitable for the management needs of dense road networks and areas with dense main road intersections.

[0055] (5) Build a closed-loop control system; the system builds a full-process closed-loop mechanism from the perception layer, decision layer to the control layer. The execution results of the traffic lights can be captured and fed back by the perception layer again, supporting strategy evaluation and automatic correction, ensuring the continuous optimization of the control effect, and having the ability of adaptive evolution.

[0056] (6) Flexible deployment and strong adaptability: This system supports zoning deployment according to urban road grades and traffic flow changes. Low-altitude drones can be flexibly dispatched, do not rely on fixed infrastructure, and adapt to various road types, making it easier for urban traffic management departments to promote and apply them.

[0057] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0059] Figure 1 Schematic diagram of the architecture of a low-altitude drone-assisted urban intersection coordination control system during peak hours according to an embodiment of the present invention;

[0060] Figure 2 Schematic diagram of the hierarchical structure of a low-altitude drone-assisted urban intersection coordination control system during peak hours according to an embodiment of the present invention;

[0061] Figure 3 The figure is a flow chart of a method for coordinated control of urban intersections during peak hours assisted by a low-altitude drone according to an embodiment of the present invention. DETAILED DESCRIPTION

[0062] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0063] Among them, the accompanying drawings are only for illustrative purposes and represent only schematic diagrams rather than actual pictures, and should not be understood as limiting the present invention. In order to better illustrate the embodiments of the present invention, some parts of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the dimensions of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions may be omitted in the accompanying drawings.

[0064] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "back", etc. indicating directions or positional relationships, they are based on the directions or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operate in a specific direction. Therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0065] See also Figures 1 to 3 , which is a low-altitude drone-assisted coordinated control system and method for urban intersections during peak hours.

[0066] Example 1

[0067] This embodiment first describes in detail the specific structure of the low-altitude drone-assisted urban intersection coordination control system during peak hours of the present invention. Figure 1 The overall system architecture diagram shown in the figure includes a device deployment layer, a perception layer, a decision layer, and a control layer. The perception layer, the decision layer, and the control layer implement corresponding data collection and data processing based on the basic settings deployed in the device deployment layer.

[0068] The device deployment layer includes the deployment of low-altitude drones, intelligent traffic light controllers, communication modules, and edge computing devices. The low-altitude drones and edge computing devices communicate through the communication modules. The drones obtain video information and perform front-end processing tasks. The edge computing devices then extract structured data and upload it to the perception layer.

[0069] The perception layer determines intersection status information based on structured data and passes it to the decision layer, which generates real-time signal timing strategies and regional scheduling plans based on the intersection status information.

[0070] The control layer controls traffic lights based on the real-time signal timing plan and regional scheduling plan output by the decision layer, and the execution results of the traffic lights are fed back to the perception layer through ground equipment or drones; in the event of continued congestion or strategy execution failure, the system triggers adaptive optimization and early warning mechanisms to achieve dynamic correction.

[0071] In this embodiment, if Figure 2 The following diagram shows the hierarchical structure of a low-altitude drone-assisted coordinated control system for urban intersections during peak hours. The equipment deployment layer, the system's infrastructure support, primarily includes: a low-altitude drone system equipped with a high-definition camera, an embedded computing platform (such as the Jetson Nano), and a flight control system. It can cruise and capture images at the intersection and its surrounding area, performing local preliminary image analysis, such as vehicle detection and lane recognition, to generate some structured data locally. Drone deployment is primarily during peak traffic flow periods (such as morning and evening rush hours). An intelligent traffic light controller has an external communication interface, supports remote control policy reception, and dynamically adjusts parameters such as traffic light duration and phase sequence. Edge computing devices (MEC nodes) are deployed near the signal control center and are responsible for structured data aggregation, event analysis, and control policy generation, possessing strong real-time processing capabilities. The communication module utilizes high-speed communication technologies such as 5G and V2X to support data synchronization between drones and MEC nodes (uploading only structured information) and command distribution between the MEC and traffic lights, ensuring low-latency and highly reliable data exchange.

[0072] In this embodiment, the perception layer is responsible for real-time monitoring of traffic conditions and data collection, mainly including: UAV front-end video perception and preliminary analysis module: on the UAV, embedded computing is used to implement vehicle target detection, traffic density estimation, abnormal behavior screening and other functions, and only structured information (such as the number of vehicles, congestion level, and suspected abnormal event signs) is sent to the MEC node, significantly reducing the transmission load. Assume that the structured data D output by UAV 1 is l (t) is:

[0073] D l (t) = {n veh ,ρ lane ,E abn}

[0074]

[0075] Where: n vehis the number of vehicles, indicating the number of vehicles on the road per unit time; ρ lane is the lane occupancy rate, reflecting the degree of traffic congestion; L is the total length of the observed road section, L i is the length of vehicle i; E abn It is a sign of abnormal events, such as wrong-way driving and traffic accidents.

[0076] Multi-source data fusion module: Integrates structured perception data from different drones and can also introduce ground perception sources such as ground induction coils, surveillance cameras, and RSUs to improve overall perception coverage and reliability. Suppose the structured perception data from different perception sources is:

[0077] D k (t)={m veh ,D l (t)}

[0078] where m veh The number of vehicles released upstream of the intersection.

[0079] Event detection module: The edge computing end determines whether there is an emergency (such as serious queues, wrong-way traffic, accidents, etc.) based on the structured data stream and rule model, and generates a description of the current intersection state. Specifically, the edge computing end calculates the lane occupancy rate ρ lane The threshold value is used to determine whether there is congestion at the current intersection and record it. lane >ρ thresh , it is judged as congestion. thresh is the set congestion determination threshold. In addition, if the above detection result belongs to E abn , then record abnormal events; structured perception data D from different perception sources k (t) is used to guide the decision-making layer to specify the signal timing plan.

[0080] In this embodiment, the decision-making layer is responsible for generating traffic control strategies, which mainly include:

[0081] Timing Optimization Module: First, the signal light sequence is rotated according to the preset plan. Then, the duration of each phase is based on the following plan: Based on the perception data, the current traffic state is analyzed and the dynamic timing algorithm is used to generate the optimal timing strategy for the intersection. Assume that the green light duration s of the signal light corresponding to lane j is j The calculation formula is:

[0082]

[0083] Among them, n veh Indicates the number of vehicles in the queue; s m =5m is the distance the front vehicle has traveled when the rear vehicle starts; s1 is the distance the vehicle travels after starting and accelerating to the required speed; vd and t loss =3s represents the expected speed and lost time respectively; is the starting acceleration; t is the system time; L i is the vehicle length; m veh Indicates the number of vehicles input from upstream, t add =2s means taking into account the delay time of upstream vehicles; Δτ k represents the strategy adjustment amount, and its value is determined by referring to formula (8); Δs * is a safe distance. This embodiment uses the intelligent driver model (IDM) to describe the car following behavior, where the predicted acceleration Defined as:

[0084]

[0085]

[0086] Among them, a max , v max , s0, σ and b represent the maximum acceleration, maximum speed, congestion distance, safe headway and expected acceleration respectively; Indicates the instantaneous speed of the vehicle; Δd n (t), Δv n (t), Δs * They represent the distance, speed difference, and safety distance between vehicle n and vehicle n-1 respectively.

[0087] Strategy Evaluation and Adaptive Feedback Module: Based on signal control results and real-time traffic feedback, it continuously evaluates the effectiveness of the current strategy and makes adaptive corrections. The effectiveness of the strategy is evaluated according to the following formula, and the strategy parameters are adjusted according to the actual situation:

[0088] E=n veh +α1m veh -n real

[0089] E represents the difference between the target number of vehicles passing through and the actual number of vehicles passing after a green light is released. α1 represents the weight of upstream vehicles, n real Indicates the actual number of vehicles passing. According to the strategy error, feedback is used to correct the signal timing strategy:

[0090] Δτ k (t) = γ·E

[0091] Where Δτ k (t) is the strategy adjustment amount, and γ is the feedback adjustment coefficient.

[0092] The adaptive feedback regulation mechanism can ensure that the system has strong adaptability to the dynamic changes in traffic conditions and guarantee the real-time and effectiveness of traffic control strategies.

[0093] In this embodiment, the control layer is responsible for executing control commands issued by the upper layer. Specifically, it includes the following: The command issuance module: The edge computing device transmits the timing strategy generated by the decision-making layer to the traffic light controller at the corresponding intersection in real time via a wireless link. The traffic light execution module: Automatically adjusts the traffic light cycle, phase sequence, and phase according to the timing strategy received by the traffic light controller. The feedback confirmation mechanism: The traffic light controller feeds back the execution results to the edge computing node and performs closed-loop verification through drone perception to achieve closed-loop strategy tuning.

[0094] Example 2

[0095] This embodiment is directed to the system in Example 1, taking two consecutive cross-shaped urban intersections (respectively referred to as intersection A and intersection B) as an example, and combining it with a typical traffic scene during the morning rush hour to introduce the workflow and application effects of the system of the present invention.

[0096] 1) Implementation environment and parameter settings

[0097] Traffic scenario: Intersections A and B are 300 meters apart, both with six lanes in both directions, and are located on a major urban road. During the morning rush hour (7:30-9:00), traffic is dense, with significant queues and bottlenecks along the main road. Table 1 shows the traffic flow configuration at the intersection, and Table 2 shows the fixed signal timing plan for intersections A and B:

[0098] Table 1

[0099]

[0100] Table 2

[0101]

[0102] Drone deployment: One low-altitude drone is deployed at each intersection, flying at an altitude of 40 meters. The cruising path covers the intersection and the two lanes in front and behind.

[0103] Drone hardware: Equipped with a 1080p HD camera, a Jetson Nano embedded computing module, and a 5G communication terminal.

[0104] Edge computing device (MEC): deployed in the traffic management control center near intersection B, communicates with the outside world through RSU, has a GPU acceleration module, and supports real-time strategy generation.

[0105] Traffic light controller: supports V2X communication interface and can receive remote signal timing control instructions.

[0106] 2) Workflow

[0107] Figure 3 This is an overall control flow chart under the system collaborative processing architecture, that is, a low-altitude drone-assisted urban intersection coordinated control method during peak hours proposed in this embodiment, which includes the following steps:

[0108] Step 1: Front-end perception and data processing (perception layer)

[0109] The drone cruises in an S-shaped path over intersections A and B, capturing traffic images. The embedded computing module analyzes the images in real time to identify lane traffic conditions (number of vehicles, average speed, queue length, whether there are any delays, etc.). The extracted structured data is packaged into a "traffic status package" and sent to the MEC node via the 5G network.

[0110] Step 2: Data fusion and event judgment (edge ​​perception fusion)

[0111] The MEC node receives traffic status data from UAVs A and B and establishes the intersection status matrix D k (t), determine whether the following events exist: the lane occupancy rate exceeds the threshold (such as ρ lane ≥0.6), which triggers the strategy optimization process.

[0112] The perception layer is responsible for real-time monitoring of traffic conditions and data collection, mainly including: UAV front-end video perception and preliminary analysis module: vehicle target detection, traffic density estimation, abnormal behavior screening and other functions are realized through embedded computing on the UAV, and only structured information (such as the number of vehicles, congestion level, and suspected abnormal event signs) is sent to the MEC node, significantly reducing the transmission load.

[0113] The multi-source data fusion module integrates structured perception data from different drones, and can also introduce ground perception sources such as ground induction coils, surveillance cameras, and RSUs to improve the overall perception coverage and reliability.

[0114] The edge computing end of the event detection module determines whether there is an emergency (such as serious queues, wrong-way traffic, accidents, etc.) based on the structured data stream and rule model, and generates a description of the current intersection state. Specifically, the edge computing end calculates the lane occupancy rate ρ lane The threshold value is used to determine whether there is congestion at the current intersection and record it. lane >ρ thresh , it is judged as congestion. thresh is the set congestion determination threshold. In addition, if the above detection result belongs to E abn , then record the abnormal event.

[0115] Step 3: Signal Timing Optimization and Coordinated Scheduling (Decision-making Layer)

[0116] First, the traffic light sequence rotates according to a preset plan. Then, the duration of each phase is determined based on the following plan: based on the current traffic state analyzed by sensor data, a dynamic timing algorithm is used to generate the optimal timing strategy for the intersection.

[0117] Strategy Evaluation and Adaptive Feedback Module: Based on signal control results and real-time traffic feedback, it continuously evaluates the effectiveness of the current strategy and makes adaptive corrections. This adaptive feedback adjustment mechanism ensures the system's strong adaptability to dynamic changes in traffic conditions, ensuring the real-time and effectiveness of traffic control strategies.

[0118] Step 4: Policy issuance and control execution (control layer)

[0119] The generated signal control strategy is sent by MEC to the signal light controllers at each intersection; the controller adjusts the light control parameters in real time to implement operations such as dynamic release priority adjustment and green light time extension; the controller execution status is fed back to MEC through the status confirmation module, and the execution effect is verified by drone perception.

[0120] 3) Experimental data and effect verification

[0121] In order to verify the control effect of the system of the present invention during peak hours, the following two indicators were selected for comparative experiments. The comparison results are shown in Table 3:

[0122] Table 3

[0123] Test items Traditional fixed time control solution Solution of the present invention Improvement rate Average queue length (meters) 110m 65m ↓40.9％ Number of vehicles passing through (30 minutes) 630 vehicles 880 vehicles ↑39.7％ Average speed (in the intersection) 12km / h 18km / h ↑50％

[0124] 4) Results analysis and summary

[0125] As can be seen from the experimental results in Table 3, the present invention effectively alleviates the traffic bottleneck problem during peak hours through the collaborative perception and control mechanism assisted by low-altitude drones. Its core advantages are reflected in:

[0126] Wider and more flexible perception: avoiding blind spots of ground sensors; more efficient communication: only structured data is transmitted to reduce bandwidth pressure; strategies are more dynamic and intelligent: with the ability to respond to emergencies; and more globally coordinated management and control: supporting unified scheduling of multiple intersections to avoid policy conflicts.

[0127] The present invention can effectively alleviate traffic congestion during peak hours in the morning and evening: by using aerial drones to obtain the traffic flow status of multiple intersections in real time, dynamically analyze congestion trends, adjust signal timing in a timely manner, effectively guide vehicle traffic, and improve intersection traffic efficiency. Improve the intelligence level of signal control at urban intersections: build a closed-loop system architecture from perception, analysis to control, integrate drones, edge computing and intelligent signal control equipment, and realize the real-time generation and execution of signal timing plans. Enhance the responsiveness and adaptability of the traffic management system: It has a rapid response mechanism for sudden traffic events (such as accidents, abnormal traffic flow), can temporarily adjust signal strategies, and issue guidance instructions, thereby improving system robustness and traffic safety assurance levels. Realize linkage scheduling control between intersections: through the mobile perception capability of drones and the regional data sharing mechanism, support unified coordinated control of multiple adjacent intersections, and avoid the situation where local optimization leads to a decrease in global efficiency.

[0128] The present invention overcomes the limitations of traditional ground-based sensing methods in terms of field of view, information update frequency, and real-time performance; it makes up for the shortcomings of existing drone sensing solutions in that they fail to form an automatic control closed loop and lack a real-time control execution mechanism; and it establishes a traffic collaborative management and control system with rapid perception, intelligent decision-making, and precise control for peak hours, multiple intersections, and complex traffic conditions.

[0129] This embodiment fully verifies the practicality, innovation and significant effect of the present invention in the field of intelligent control of urban traffic signals.

[0130] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A low-altitude drone-assisted coordinated control system for urban intersections during peak hours, characterized by: The system includes: a device deployment layer, a perception layer, a decision layer and a control layer. The perception layer, the decision layer and the control layer implement corresponding data collection and data processing based on the basic settings deployed by the device deployment layer; wherein, The device deployment layer includes the deployment of low-altitude drones, intelligent traffic light controllers, communication modules, and edge computing devices. The low-altitude drones and edge computing devices communicate through the communication modules. The drones obtain video information and perform front-end processing tasks. The edge computing devices then extract structured data and upload it to the perception layer. The perception layer determines intersection status information based on structured data and passes it to the decision layer, which generates real-time signal timing strategies and regional scheduling plans based on the intersection status information. The control layer controls traffic lights based on the real-time signal timing plan and regional scheduling plan output by the decision layer, and the execution results of the traffic lights are fed back to the perception layer through ground equipment or drones; in the event of continued congestion or strategy execution failure, the system triggers adaptive optimization and early warning mechanisms to achieve dynamic correction.

2. The low-altitude drone-assisted urban intersection coordination control system during peak hours according to claim 1 is characterized by: The low-altitude drones at the equipment deployment layer are equipped with high-definition cameras, embedded computing platforms, and flight control systems. They use the high-definition cameras to conduct aerial patrols and photography of intersections and their surrounding areas. The embedded computing platform then performs preliminary image analysis, including vehicle detection and recognition, and generates localized structured data. The intelligent traffic light controller includes an external communication interface, remotely receives a control strategy through the external communication interface, and dynamically adjusts the duration and phase sequence of the traffic lights according to the received control strategy; The edge computing device obtains the structured information transmitted by the drone through the communication module, and completes time analysis and control strategy generation based on the structured information; The communication module uses high-speed communication technology to support data synchronization between drones and MEC nodes, and information transmission in the case of command distribution between MEC and traffic lights.

3. The low-altitude drone-assisted urban intersection coordination control system during peak hours according to claim 1 is characterized by: The perception layer performs corresponding operations based on the various infrastructure deployed by the basic equipment layer. The perception layer includes: drone front-end video perception and preliminary analysis module, multi-source data fusion module, and event detection module; The drone's front-end video perception and preliminary analysis module is deployed in the drone's embedded computing platform. It completes vehicle target detection, traffic density estimation, and initial screening of abnormal behavior based on the acquired video information, and transmits the structured data to the edge computing node. The multi-source data fusion module integrates structured perception data from different UAVs; The event detection module determines whether an emergency occurs based on the structured data stream and rule model and generates a description of the current intersection status.

4. The low-altitude drone-assisted urban intersection coordination control system during peak hours according to claim 3 is characterized by: In the UAV front-end video perception and preliminary analysis module, the structured data D output by the UAV l l (t) is expressed as: D l (t)={n veh ,ρ lane ,E abn } Where: n veh is the number of vehicles, indicating the number of vehicles on the road per unit time; ρ lane is the lane occupancy rate, reflecting the degree of traffic congestion; L is the total length of the observed road section, L i is the length of vehicle i; E abn It is a sign of abnormal events, including wrong-way events and traffic accidents; The structured perception data of different perception sources after integration by the multi-source data fusion module is: D k (t)={m veh ,D l (t)} Among them, m veh The number of vehicles released upstream of the intersection; The event detection module calculates the lane occupancy rate ρ lane The threshold value is used to determine whether there is congestion at the current intersection and record it; the congestion determination threshold is set to ρ thresh , if ρ lane >ρ thresh , it is judged to be congested, otherwise it is judged to be non-congested; if the above detection result belongs to E abn , then record abnormal events; structured perception data D from different perception sources k (t) is used to guide the decision-making layer to specify the signal timing plan.

5. The low-altitude drone-assisted urban intersection coordination control system during peak hours according to claim 1 is characterized by: The decision-making layer is responsible for generating traffic control strategies. It includes a timing optimization module and a strategy evaluation and adaptive feedback module. When traffic lights are rotated according to a preset plan, the timing optimization module analyzes the current traffic state based on sensor data and uses a dynamic timing algorithm to generate the optimal timing strategy for the intersection. The duration of each phase is adjusted based on the optimal timing strategy. The strategy evaluation and adaptive feedback module continuously evaluates the effectiveness of the current strategy and makes adaptive corrections based on signal control results and real-time traffic feedback.

6. The low-altitude drone-assisted urban intersection coordination control system during peak hours according to claim 5, characterized in that: In the configuration optimization module, the green light duration s of the traffic light corresponding to lane j under the optimal timing strategy is j The calculation method is: Among them, n veh Indicates the number of vehicles in the queue; s m is the distance traveled by the front vehicle when the rear vehicle starts; s1 is the distance traveled by the vehicle after it starts and accelerates to the required speed; v d and t loss represent the expected speed and loss time respectively; is the starting acceleration; t is the system time; L i is the vehicle length; m veh Indicates the number of vehicles input from upstream, t add Indicates the delay time of upstream vehicles; Δτ k represents the strategy adjustment amount; Δs * is the safety distance, and the intelligent driver model is used to describe the car following behavior, where the predicted acceleration Defined as: Among them, a max 、v max , s0, σ and b represent the maximum acceleration, maximum speed, congestion distance, safety headway and expected acceleration respectively; Indicates the instantaneous speed of the vehicle; Δd n (t), Δv n (t), Δs * They represent the distance, speed difference, and safety distance between vehicle n and vehicle n-1 respectively.

7. The low-altitude drone-assisted urban intersection coordination control system during peak hours according to claim 6, characterized in that: The strategy evaluation and adaptive feedback module evaluates the effectiveness of the strategy according to the following formula: E=n veh +α1m veh -n real Among them, E represents the difference between the target number of vehicles passing through and the actual number of vehicles passing after a green light is released, α1 represents the weight of upstream vehicles, and n real Indicates the actual number of vehicles passing through; Then, the signal timing strategy is modified based on the strategy error feedback: Dt k (t)=γ·E Where Δτ k (t) is the strategy adjustment amount, and γ is the feedback adjustment coefficient.

8. The low-altitude drone-assisted urban intersection coordination control system during peak hours according to claim 1, characterized in that: The control layer includes an instruction issuing module, a traffic light execution module, and a feedback confirmation module. The instruction issuing module deployed in the edge computing device sends the timing strategy generated by the decision-making layer to the traffic light controller of the corresponding intersection in real time through a wireless link. The traffic light execution module in the traffic light controller automatically completes the traffic light cycle, phase sequence, and phase adjustment according to the timing strategy received by the signal controller; the feedback confirmation module in the traffic light controller feeds back the execution results to the edge computing node, and performs closed-loop verification through drone perception to achieve closed-loop optimization of the strategy.

9. A low-altitude drone-assisted coordinated control method for urban intersections during peak hours, characterized by: A low-altitude drone-assisted urban intersection coordination control system during peak hours according to any one of claims 1 to 8, the method comprising: Obtain intersection video images through drones, perform front-end perception and data processing at the perception layer, and pass the processed structured data to the decision layer; The edge computing device receives traffic status data from different drones, builds an intersection status matrix, and determines whether the lane occupancy exceeds a threshold, triggering a strategy optimization process if so; Signal timing optimization and coordinated scheduling are performed at the decision-making level. Based on the perception data, the current traffic status is analyzed and the optimal timing strategy for the intersection is generated using a dynamic timing algorithm. The generated signal control strategy is sent to the traffic light controllers at each intersection; the controllers adjust the light control parameters in real time, and the controller execution status is fed back to the edge computing device through the status confirmation module, and the execution effect is perceived and verified by the drone.

Images