A real-time traffic signal control method and system
By deploying 3D camera units at intersections to identify vehicles and pedestrians in real time, and dynamically adjusting traffic light status in conjunction with a conflict rule base, the problems of low traffic efficiency and poor safety in existing traffic signal control methods are solved, achieving real-time optimization and safe passage at the lane level.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- TAIZHOU CONVENIENT ENTERPRISE MANAGEMENT CO LTD
- Filing Date
- 2026-04-29
- Publication Date
- 2026-07-31
AI Technical Summary
Existing traffic signal control methods cannot adapt to the dynamic and random nature of traffic flow, resulting in low traffic efficiency and difficulty in handling complex conflicts when multiple lanes in different directions request passage simultaneously. In particular, the response to priority passage of special vehicles is delayed, wasting road resources.
By deploying 3D camera units to collect video stream data in real time, identifying and tracking vehicle and pedestrian targets, and combining them with a predefined conflict rule base, the traffic light status is dynamically adjusted to achieve lane-level real-time control, ensuring first-come-first-served passage and avoiding conflicts.
It significantly improves the efficiency and safety of intersection traffic, ensures priority passage for special vehicles, optimizes the use of road resources, reduces frequent vehicle starts and stops, and improves the response speed and safety of the traffic system.
Smart Images
Figure CN122493673A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent traffic control technology, and in particular to a real-time traffic signal control method and system. Background Technology
[0002] Urban intersections are bottlenecks in traffic flow. Currently, the most widely used traffic signal control methods are fixed timing and simple inductive control. Fixed timing methods preset the signal cycle and green light ratio based on historical statistical data, operating in a fixed, cyclical pattern. However, traffic flow is highly dynamic and random, and fixed timing cannot adapt to sudden congestion or empty lanes, resulting in a low probability of vehicles encountering a green light when arriving at the intersection, frequent starts and stops, and low traffic efficiency.
[0003] Existing technologies include some sensor-based control schemes that detect vehicle presence by burying inductive loops or installing radar detectors. However, these methods typically only detect vehicles at close range, lacking foresight regarding the overall situation at intersections and struggling to handle complex conflicts when multiple lanes simultaneously request passage. Furthermore, for priority passage of emergency vehicles such as fire trucks and ambulances, existing systems usually rely on manual intervention or single GPS positioning, resulting in delayed responses and an inability to dynamically and safely insert priority phases based on real-time traffic conditions, easily leading to secondary conflicts. More importantly, traditional control systems bind traffic lights in pairs or groups, switching them according to fixed phase combinations. When a lane in one direction is empty, the lane bound to it in the other direction, even if it has a need, cannot obtain the right-of-way, significantly wasting road space and time resources. Summary of the Invention
[0004] This application proposes a real-time traffic signal control method and system, which addresses the problem that existing technologies cannot optimize real-time dynamic signals based on the fine trajectory of a single vehicle, resulting in low traffic efficiency and poor safety. It achieves "green light on when a vehicle arrives, yellow light flashing when a vehicle leaves," significantly improving the traffic efficiency and safety of intersections.
[0005] To achieve the above objectives, this application provides the following technical solution: Firstly, this application proposes a real-time traffic signal control method, comprising: Real-time acquisition of target request information within multiple preset lanes at the intersection; the target request information includes the time when the target enters the lane and the identifier of the lane entry; A conflict rule base is predefined and stored; the conflict rule base defines the conflict correspondence between different lane signs; Based on the target request information, add the lanes requesting passage to the processing queue in the order of the request time; In response to the current request lane in the pending queue, query the conflict rule base to determine at least one conflicting lane that has a conflict relationship with the current request lane; If the conflict lane is not currently permitted to pass, then a first control command and a second control command are generated; In response to the first control command, the traffic light corresponding to the currently requested lane is switched from a first state representing waiting to pass to a second state representing permission to pass; and In response to the second control command, the traffic light corresponding to the conflict lane is switched from the first state to the third state, which represents prohibition of passage.
[0006] In conjunction with the first aspect, the real-time acquisition of target request information within multiple preset lanes at the intersection includes: Video stream data is collected in real time by 3D camera units deployed at intersections; The video stream data is processed based on the target detection model to identify and track the requesting targets entering the intersection; the requesting targets include vehicle targets and pedestrian targets. Based on the tracking trajectory of the requested target and in conjunction with the preset lane numbering rules, determine the lane identifier where the requested target is located; In response to the preset conflict zone of the requested target entering the intersection, target request information including lane markings and the time of entry is generated.
[0007] In conjunction with the first aspect, the step of responding to the currently requested lane in the pending queue, querying the conflict rule base, and determining at least one conflicting lane that has a conflict relationship with the currently requested lane includes: Based on the tracking trajectory, determine whether the target has completely left the preset conflict area; If so, then generate a third control command; In response to the third control command, the traffic lights corresponding to the currently requesting lane and the conflicting lane are restored from the second or third state back to the first state.
[0008] In conjunction with the first aspect, the step of responding to the currently requested lane in the pending queue, querying the conflict rule base, and determining at least one conflicting lane that has a conflict relationship with the currently requested lane includes: If the conflicting lane is already permitted to pass, the currently requested lane and the time of the request are stored in the pending queue, and the traffic light of the currently requested lane is switched to the third state.
[0009] In conjunction with the first aspect, the step of responding to the currently requested lane in the pending queue, querying the conflict rule base, and determining at least one conflicting lane that has a conflict relationship with the currently requested lane includes: In response to the detection that the traffic demand at the intersection has reached a preset peak condition, a time-limited traffic mode is triggered. In time-limited passage mode, a preset green light passage time is assigned to all lanes or multiple selected lanes; After the preset time period ends, the dynamic control logic based on the pending queue and conflict rule base is restored.
[0010] In conjunction with the first aspect, the conflict rule base is a conflict matrix or adjacency table pre-constructed based on the physical path intersection relationships of lanes in each direction at the intersection.
[0011] In conjunction with the first aspect, the real-time acquisition of target request information within multiple preset lanes at the intersection also includes: The turn signal status of the target vehicle is identified using a 3D camera unit; For a combined lane containing multiple driving direction signs, the target driving direction corresponding to the target request information is determined based on the turn signal status, and the conflict rule base is queried based on the target driving direction.
[0012] Secondly, this application also proposes a real-time traffic signal control system, comprising: The perception module is configured to acquire target request information in multiple preset lanes at the intersection in real time; the target request information includes the time when the target enters the lane and the identifier of the lane entry. The storage module is configured to store predefined rules and a conflict rule base; the conflict rule base defines the conflict correspondence between different lane signs. The decision module, which is communicatively connected to both the perception module and the storage module, is configured as follows: Based on the target request information, add the lanes requesting passage to the processing queue in the order of the request time; In response to the current request lane in the pending queue, query the conflict rule base to determine at least one conflicting lane that has a conflict relationship with the current request lane; If the conflict lane is not currently permitted to pass, then a first control command and a second control command are generated; The execution module, communicatively connected to the decision module, is configured to, in response to a first control command, switch the traffic light corresponding to the currently requested lane from a first state representing a wait-to-pass state to a second state representing permission to pass; and In response to the second control command, the traffic light corresponding to the conflict lane is switched from the first state to the third state, which represents prohibition of passage.
[0013] In conjunction with the second aspect, the sensing module includes: Roadside video acquisition unit: Used to acquire video stream data in real time through 3D camera units deployed at intersections; Request target recognition unit: used to process video stream data based on target detection model, identify and track request targets entering the intersection; where request targets include vehicle targets and pedestrian targets; Lane marking recognition unit: used to determine the lane marking where the requested target is located based on the tracking trajectory of the requested target and in combination with preset lane numbering rules; Conflict determination unit: In response to a preset conflict area where a target is requested to enter the intersection, it generates target request information including lane markings and the time of entry.
[0014] Thirdly, this application also proposes a computer-readable storage medium containing computer-executable instructions that, when executed by one or more processors, cause the processors to perform the aforementioned real-time traffic signal control method.
[0015] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.
[0016] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description
[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.
[0018] In the attached diagram: Figure 1 This is a flowchart of a real-time traffic signal control method according to an embodiment of the present invention; Figure 2 This is a system composition diagram of a real-time traffic signal control system according to an embodiment of the present invention; Figure 3 This is a schematic diagram of intersecting lanes in an embodiment of the present invention. Detailed Implementation
[0019] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.
[0020] To facilitate understanding of the technical solutions in the embodiments of this application, the rules for the status lights at intersections in this application are first given.
[0021] Traffic lights at intersections have three states. The first state is the default "waiting to proceed" state, meaning there are no vehicles or pedestrians in that lane or any conflicting lanes. The second state is the "permitted to proceed" state, indicating that there are vehicles or pedestrians in the lane. The third state is the "prohibited to proceed" state in response to the conflicting lanes in the second state. Therefore, the traffic light for the lane where the target is stopped and waiting is in the third state, which is the red light state.
[0022] Each lane's traffic light transitions between a first state (default waiting state where neither the lane nor the conflicting lane has any vehicles or pedestrians) and a second state (green light allowing passage) or a third state (red light prohibiting passage in response to the second state).
[0023] Example 1, as Figure 1 As shown, this embodiment provides a real-time traffic signal control method for achieving lane-level, real-time dynamic control of traffic at intersections. This application deconstructs the traditional phase-bundled control of traffic lights into lane-independent control, and achieves traffic light control at intersections through a first-come, first-served approach, thereby improving traffic efficiency.
[0024] The specific steps in this embodiment are as follows: Step 1: Acquire target request information in multiple preset lanes at the intersection in real time; the target request information includes the time when the target enters the lane and the identifier of the lane entry. In one specific embodiment, the target request information represents the digital request information of the passing subject, namely a vehicle or pedestrian, requesting to obtain the right-of-way in the current lane. It includes: a lane identifier, uniquely identifying which specific lane the request occurred in; and a request time, representing the timestamp precisely recording the target request information being captured by the system. Specifically, at traffic intersections, this application, unlike simply sensing the presence of vehicles in a certain direction, can determine which lane and at what time there is a need for passage based on the target request information.
[0025] Step 2: Predefine and store the conflict rule base; whereby the conflict rule base defines the conflict correspondence between different lane signs; In one specific embodiment, the conflict rule base is a static data model based on the physical topology of the intersection; it defines whether there is a path intersection between two (or more) lane signs, i.e. whether right-of-way can be granted, to prevent the risk of traffic accidents caused by control logic errors.
[0026] like Figure 3 As shown, Figure 3This is a diagram of the intersecting lanes. When a vehicle is detected in lane 7 (west) of Fuying East Road, the traffic light for lane 7 turns green, the conflicting lane (eastbound lane) turns red, and the traffic light for lane 8 (eastbound lane) of Fuying East Road turns yellow. When a vehicle is detected entering the intersection of Fuying East Road via the pedestrian crossing at lane 0 (eastbound lane), the traffic light for the corresponding lane returns to yellow. Step 3: Based on the target request information, add the lanes requesting passage to the processing queue according to the order of the request time. In one specific embodiment, the processing queue is a dynamic data queue structure used to receive and store target request information for different lanes. The queue is arranged sequentially based on the order of request timestamps, implementing first-come, first-served service. By introducing a timestamp-based queue mechanism, resource contention issues during concurrent requests from multiple lanes can be resolved, preventing lanes from being permanently inefficient due to priority limitations.
[0027] Step 4: Identify conflicting relationships; In response to the currently requested lane in the queue, the system queries the conflict rule base to identify at least one conflicting lane that has a conflict relationship with the currently requested lane. In one specific embodiment, the currently requested lane is the first request in the queue. The system uses the lane identifier of the request as the search keyword and matches it with the conflict rule base to identify all conflicting lane identifiers that have a trajectory intersection with the requested lane on the physical path, thereby clarifying the set of lanes that cannot be used simultaneously.
[0028] Step 5: Determine the right of way and issue instructions; If the conflicting lane is not currently permitted to pass, a first control command and a second control command are generated. In one specific embodiment, the system needs to verify that the current state of all identified conflicting lanes is not "permitted to pass," which is a core prerequisite for ensuring that the currently requesting lane obtains safe right-of-way. Once it is confirmed that there is no potential conflict risk, the system determines that the currently requesting lane has the right-of-way and generates a control command. This process aims to ensure that the currently requesting lane and any conflicting lanes with intersecting paths do not simultaneously obtain the right-of-way, thereby completely eliminating the risk of intersection collisions at the physical level and ensuring traffic safety.
[0029] In a preferred embodiment, the initial state of all traffic lights at the current intersection is the first state of waiting to pass by default. When a target lane is detected to have a target passage request, the conflict detection checks all conflicting lanes of the first target lane. If it is determined that none of the conflicting lanes are allowed to pass, the traffic light of the first target lane changes from the first state to the second state of allowing passage, while the traffic lights of the lanes that conflict with the first target lane change from the first state to the third state.
[0030] When a target request is detected in another lane, namely the second target lane, the conflict detection mechanism will check all conflicting lanes of the second target lane. If all conflicting lanes of the second target lane are found to be in conflict with the first target lane, the second target lane and the target request will be stored in the priority queue for processing. At this time, the traffic light of the second target lane is already in the third state.
[0031] Step 6: In response to the first control command, switch the traffic light corresponding to the currently requested lane from the first state representing waiting to pass to the second state representing permission to pass; In one specific embodiment, the first state refers to a yellow light or a non-lit state (when set to non-lit, a text sign indicating "Wait" can be displayed at the intersection); when there is a red and green light transition, the current traffic signal rules still apply, treating the "yellow light" as an intermediate transition light. The first state also represents a waiting state, which is the system's default safe initial state. The green light is not part of the cycle but rather an immediate request-authorization response, preventing idle driving and waiting.
[0032] Step 7: In response to the second control command, switch the traffic light corresponding to the conflict lane from the first state to the third state, which represents prohibition of passage.
[0033] In one specific embodiment, the third state represents a red light, indicating that passage is prohibited. This state is executed simultaneously with step 6, locking all lanes that conflict with the currently passing lane to ensure the absolute right-of-way for the currently passing lane. This application achieves safe and stable passage through a one-to-one authorization and locking operation, ensuring that at any given time, only non-conflicting lane combinations can pass through the intersection, thereby improving the temporal utilization rate of the intersection space.
[0034] Example 2, in the real-time process of step 1, specifically by acquiring target request information in multiple preset lanes at the intersection in real time, includes: Video stream data is collected in real time by 3D camera units deployed at intersections; In one embodiment of this application, the 3D camera unit includes a 3D camera, which is a sensing front end used to capture real-time RGB images of the intersection and depth information of each pixel, thereby obtaining the three-dimensional spatial position of the target vehicle or pedestrian. Compared to a 2D camera, the depth information from the 3D camera can determine the position, size, and motion state of the target vehicle and pedestrian in the real world, thus avoiding the scale blur problem of monocular vision, especially the scale blur under changes in lighting and inclement weather.
[0035] The video stream data is processed based on the target detection model to identify and track the requesting targets entering the intersection; the requesting targets include vehicle targets and pedestrian targets. In one embodiment of this application, a deep learning model is used to detect targets in each frame of a video stream, thereby identifying participants such as vehicles, pedestrians, and non-motorized vehicles. At the same time, a target tracking algorithm is used, preferably the Deepsort algorithm, to assign a unique ID to each detected target and continuously update its motion trajectory, thereby determining where the target that made the request came from and where it went.
[0036] Based on the tracking trajectory of the requested target and in conjunction with the preset lane numbering rules, determine the lane identifier where the requested target is located; In one embodiment of this application, the preset lane numbering rule maps physical lanes to a digitally represented lane code. Then, by using target tracking, combined with lane line detection results and predefined lane geometry regions, it is possible to determine which lane number the requesting target is located in or heading towards. This allows conflict lookup and command generation to be accurate down to each lane.
[0037] In response to the preset conflict zone of the requested target entering the intersection, target request information including lane markings and the time of entry is generated.
[0038] In one embodiment of this application, the preset conflict area is one or more trigger areas pre-defined in space based on the geometric features of the intersection. These are typically located near the stop line or inside the intersection. A request event is only generated when the target's trajectory first touches or enters this area. The spatiotemporal triggering mechanism introduced in this application can avoid false triggering and degree-based triggering, reducing computational pressure on the system.
[0039] Example 3: The step of responding to the current request lane in the waiting queue, querying the conflict rule base, and determining at least one conflicting lane that has a conflict relationship with the current request lane includes: Based on the tracking trajectory, determine whether the target has completely left the preset conflict area; In one embodiment of this application, by continuously tracking a target vehicle or pedestrian that has obtained the right of way and determining its position in real time, the passage ends when the target vehicle or pedestrian has completely left the pre-defined conflict area while being tracked. Compared to the end of a fixed time, determining the actual spatial position maximizes the use of green light time, ensuring the target safely and completely crosses the intersection.
[0040] If so, then generate a third control command; In response to the third control command, the traffic lights corresponding to the currently requesting lane and the conflicting lane are restored from the second or third state back to the first state.
[0041] In one embodiment of this application, when the target vehicle or pedestrian completely leaves the preset conflict area, a third control command is automatically generated, indicating that the authorized passage has ended and resource reclamation is required. The third control command restores all lanes previously set to the second state (green light) and all conflict lanes set to the third state to the default safe state (yellow light). This achieves real-time release of resource control or rapid reset of the intersection state.
[0042] Example 4: The step of responding to the current request lane in the waiting queue, querying the conflict rule base, and determining at least one conflicting lane that has a conflict relationship with the current request lane includes: If the conflicting lane is already permitted to pass, the currently requested lane and the time of the request are stored in the pending queue, and the traffic light of the currently requested lane is switched to the third state.
[0043] In one embodiment of this application, when processing the current lane request, if a query of the conflict rule base reveals that at least one conflicting lane is currently in a green light state, it indicates that the current request does not meet the conditions for immediate execution. Therefore, no authorization instruction is generated; instead, the current request is placed back into the processing queue, awaiting the next scheduling opportunity. Simultaneously, the traffic light corresponding to that lane remains red. This ensures that when the intersection is within a safe passage space, subsequent conflicting requests will not forcibly intervene but will wait in an orderly manner.
[0044] Example 5: The step of responding to the current request lane in the waiting queue, querying the conflict rule base, and determining at least one conflicting lane that has a conflict relationship with the current request lane includes: In response to the detection that the traffic demand at the intersection has reached a preset peak condition, a time-limited traffic mode is triggered. In time-limited passage mode, a preset green light passage time is assigned to all lanes or multiple selected lanes; After the preset time period ends, the dynamic control logic based on the pending queue and conflict rule base is restored.
[0045] In one embodiment of this application, preset peak conditions are set such that the length of the queue to be processed exceeds a threshold M, or the number of requests per unit time exceeds a threshold N, or lanes in a specific direction are continuously requested but cannot obtain right-of-way for an extended period. When these conditions are met, the system determines that it has entered a peak state. This leads to a time-limited passage system to prevent situations where, under extreme peak traffic conditions, a first-come, first-served micro-scheduling approach would result in some directions of traffic being unable to obtain right-of-way for an extended period.
[0046] This application temporarily suspends the queue-based conflict scheduling logic, forcibly allocating a fixed green light duration to all lanes or a specific set of lanes. During this period, each lane takes turns releasing traffic according to a preset phase scheme. In other words, through a macro-level forced release strategy, it quickly resolves the queuing vehicles backlogged in all directions, bringing the intersection back from the brink of congestion to a controllable state.
[0047] After the peak-hour time limit mode ends, it automatically switches off, clears or retains the queue of pending tasks, and reactivates the dynamic scheduling logic based on lane-level requests and conflict rule base, thus achieving a seamless connection between macro-control and micro-scheduling.
[0048] Example 6: The conflict rule base is a conflict matrix or adjacency table pre-constructed based on the physical path intersection relationship of lanes in each direction at the intersection.
[0049] In one embodiment of this application, For an intersection with N lanes to be controlled, the conflict matrix is constructed as an N×N two-dimensional Boolean matrix. The rows and columns of the matrix represent lane identifiers. If the i-th lane conflicts with the j-th lane, the matrix element Miijj = 1 (or True); otherwise, it is 0 (or False). The matrix is typically diagonally symmetric. For each lane identifier, an adjacency list is maintained, storing all other lane identifiers that conflict with that lane. The conflict matrix and adjacency list are used to store conflict rules, reducing the computational complexity when querying conflicting lanes and ensuring real-time performance and low latency under high concurrency requests.
[0050] Example 7: The real-time acquisition of target request information within multiple preset lanes at an intersection further includes: The turn signal status of the target vehicle is identified using a 3D camera unit; In one embodiment of this application, the 3D camera unit, based on target detection and tracking, uses image recognition technology to determine whether the tracked vehicle has turned on its left or right turn signal, i.e., the turn signal status, so that the perception is converted from data such as position and speed into a specific turning intention.
[0051] For a combined lane containing multiple driving direction signs, the target driving direction corresponding to the target request information is determined based on the turn signal status, and the conflict rule base is queried based on the target driving direction.
[0052] In one embodiment of this application, a combined lane is a lane marked with both straight and left turns, corresponding to multiple possible driving trajectories and conflict relationships. By recognizing the intent of the turn signal status, the ambiguity of intent in the combined lane scenario is solved, achieving on-demand allocation. That is, different driving directions on the same physical lane are matched with the most accurate conflict rules, thereby achieving more refined right-of-way allocation without increasing the number of physical lanes.
[0053] Example 8: As Figure 2 As shown, this application proposes a real-time traffic signal control system, comprising: The perception module is configured to acquire target request information in multiple preset lanes at the intersection in real time; the target request information includes the time when the target enters the lane and the identifier of the lane entry. The storage module is configured to store predefined rules and a conflict rule base; the conflict rule base defines the conflict correspondence between different lane signs. The decision module, which is communicatively connected to both the perception module and the storage module, is configured as follows: Based on the target request information, add the lanes requesting passage to the processing queue in the order of the request time; In response to the current request lane in the pending queue, query the conflict rule base to determine at least one conflicting lane that has a conflict relationship with the current request lane; If the conflict lane is not currently permitted to pass, then a first control command and a second control command are generated; The execution module, communicatively connected to the decision module, is configured to, in response to a first control command, switch the traffic light corresponding to the currently requested lane from a first state representing a wait-to-pass state to a second state representing permission to pass; and In response to the second control command, the traffic light corresponding to the conflict lane is switched from the first state to the third state, which represents prohibition of passage.
[0054] The sensing module includes: Roadside video acquisition unit: Used to acquire video stream data in real time through 3D camera units deployed at intersections; Request target recognition unit: used to process video stream data based on target detection model, identify and track request targets entering the intersection; where request targets include vehicle targets and pedestrian targets; Lane marking recognition unit: used to determine the lane marking where the requested target is located based on the tracking trajectory of the requested target and in combination with preset lane numbering rules; Conflict determination unit: In response to a preset conflict area where a target is requested to enter the intersection, it generates target request information including lane markings and the time of entry.
[0055] The above system can run on a processor, which can perform the methods described in the embodiments shown above.
[0056] This application also provides a computer-readable storage medium storing a computer program or instructions (also referred to as code). When the computer program or instructions are executed, the methods described in the embodiments shown above can be implemented.
[0057] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A real-time traffic signal control method, characterized by, include: Real-time acquisition of target request information within multiple preset lanes at the intersection; the target request information includes the time when the target enters the lane and the identifier of the lane entry; A conflict rule base is predefined and stored; the conflict rule base defines the conflict correspondence between different lane signs; Based on the target request information, add the lanes requesting passage to the processing queue in the order of the request time; In response to the current request lane in the pending queue, query the conflict rule base to determine at least one conflicting lane that has a conflict relationship with the current request lane; If the conflict lane is not currently permitted to pass, then a first control command and a second control command are generated; In response to the first control command, the traffic light corresponding to the currently requested lane is switched from a first state representing waiting to pass to a second state representing permission to pass; In response to the second control command, the traffic light corresponding to the conflict lane is switched from the first state to the third state, which represents prohibition of passage.
2. The real-time traffic signal control method as described in claim 1, characterized in that, The real-time acquisition of target request information within multiple preset lanes at the intersection includes: Video stream data is collected in real time by 3D camera units deployed at intersections; The video stream data is processed based on the target detection model to identify and track the requesting targets entering the intersection; the requesting targets include vehicle targets and pedestrian targets. Based on the tracking trajectory of the requested target and in conjunction with the preset lane numbering rules, determine the lane identifier where the requested target is located; In response to the preset conflict zone of the requested target entering the intersection, target request information including lane markings and the time of entry is generated.
3. The real-time traffic signal control method as described in claim 2, characterized in that, The step of responding to the currently requested lane in the queue by querying the conflict rule base and determining at least one conflicting lane that has a conflict relationship with the currently requested lane includes: Based on the tracking trajectory, determine whether the target has completely left the preset conflict area; If so, then generate a third control command; In response to the third control command, the traffic lights corresponding to the currently requesting lane and the conflicting lane are restored from the second or third state back to the first state.
4. The real-time traffic signal control method as described in claim 1, characterized in that, The step of responding to the currently requested lane in the queue by querying the conflict rule base and determining at least one conflicting lane that has a conflict relationship with the currently requested lane includes: If the conflicting lane is already permitted to pass, the currently requested lane and the time of the request are stored in the pending queue, and the traffic light of the currently requested lane is switched to the third state.
5. The real-time traffic signal control method as described in claim 1, characterized in that, The step of responding to the currently requested lane in the queue by querying the conflict rule base and determining at least one conflicting lane that has a conflict relationship with the currently requested lane includes: In response to the detection that the traffic demand at the intersection has reached a preset peak condition, a time-limited traffic mode is triggered. In time-limited passage mode, a preset green light passage time is assigned to all lanes or multiple selected lanes; After the preset time period ends, the dynamic control logic based on the pending queue and conflict rule base is restored.
6. The real-time traffic signal control method as described in claim 1, characterized in that, The conflict rule base is a conflict matrix or adjacency table pre-built based on the physical path intersection relationships of lanes in each direction at the intersection.
7. The real-time traffic signal control method as described in claim 1, characterized in that, The real-time acquisition of target request information within multiple preset lanes at the intersection also includes: The turn signal status of the target vehicle is identified using a 3D camera unit; For a combined lane containing multiple driving direction signs, the target driving direction corresponding to the target request information is determined based on the turn signal status, and the conflict rule base is queried based on the target driving direction.
8. A real-time traffic signal control system, characterized in that, include: The perception module is configured to acquire target request information in multiple preset lanes at the intersection in real time; the target request information includes the time when the target enters the lane and the identifier of the lane entry. The storage module is configured to store predefined rules and a conflict rule base; the conflict rule base defines the conflict correspondence between different lane signs. The decision module, which is communicatively connected to both the perception module and the storage module, is configured as follows: Based on the target request information, add the lanes requesting passage to the processing queue in the order of the request time; In response to the current request lane in the pending queue, query the conflict rule base to determine at least one conflicting lane that has a conflict relationship with the current request lane; If the conflict lane is not currently permitted to pass, then a first control command and a second control command are generated; The execution module, communicatively connected to the decision module, is configured to, in response to a first control command, switch the traffic light corresponding to the currently requested lane from a first state representing a wait-to-pass state to a second state representing permission to pass; and In response to the second control command, the traffic light corresponding to the conflict lane is switched from the first state to the third state, which represents prohibition of passage.
9. The system according to claim 8, characterized in that, The sensing module includes: Roadside video acquisition unit: Used to acquire video stream data in real time through 3D camera units deployed at intersections; Request target recognition unit: used to process video stream data based on target detection model, identify and track request targets entering the intersection; where request targets include vehicle targets and pedestrian targets; Lane marking recognition unit: used to determine the lane marking where the requested target is located based on the tracking trajectory of the requested target and in combination with preset lane numbering rules; Conflict determination unit: In response to a preset conflict area where a target is requested to enter the intersection, it generates target request information including lane markings and the time of entry.
10. A computer-readable storage medium comprising computer-executable instructions that, when executed by one or more processors, cause the processors to perform the real-time traffic signal control method according to any one of claims 1 to 7.