Traffic scheduling method and device, electronic equipment and readable storage medium

Abstract

The application discloses a traffic scheduling method and device, electronic equipment and a readable storage medium. The method comprises the following steps: reconstructing a lane of a road based on real-time road conditions to obtain a reconstructed lane topology; grouping traffic tools based on the destination and / or driving path of the traffic tools to obtain at least one fish school team; determining a target fish school team corresponding to the reconstructed lane topology based on the reconstructed lane topology and the at least one fish school team; and scheduling traffic of the target fish school team based on the reconstructed lane topology.

Description

Technical Field

[0001] This application belongs to the field of intelligent transportation technology, specifically relating to a traffic scheduling method, device, electronic equipment, and readable storage medium. Background Technology

[0002] Traditional roads use fixed lane divisions, speed limits, and traffic directions, which cannot adapt to dynamically changing traffic conditions. For example, during peak hours, insufficient lanes in a certain direction can cause congestion, while during off-peak hours, a large number of lanes may be idle. This results in low utilization of traffic resources, low traffic scheduling efficiency, and low vehicle throughput. Summary of the Invention

[0003] The purpose of this application is to provide a traffic scheduling method, device, electronic device, and readable storage medium that can solve the problems of low traffic scheduling efficiency and low vehicle passage efficiency caused by fixed lanes.

[0004] In a first aspect, embodiments of this application provide a traffic scheduling method, which includes: reconstructing lanes on a road based on real-time traffic conditions to obtain a reconstructed lane topology; platooning vehicles based on their destinations and / or travel routes to obtain at least one fish swarm; determining a target fish swarm corresponding to the reconstructed lane topology based on the reconstructed lane topology and the at least one fish swarm; and performing traffic scheduling on the target fish swarm based on the reconstructed lane topology.

[0005] Secondly, embodiments of this application provide a traffic scheduling device, which includes: a lane reconstruction module for reconstructing lanes based on real-time traffic conditions to obtain a reconstructed lane topology; a platooning module for platooning vehicles based on their destinations and / or travel routes to obtain at least one fish swarm; a determination module for determining a target fish swarm corresponding to the reconstructed lane topology based on the reconstructed lane topology and the at least one fish swarm; and a scheduling module for performing traffic scheduling on the target fish swarm based on the reconstructed lane topology.

[0006] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0007] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0008] Fifthly, embodiments of this application provide a chip, the chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the steps of the method described in the first aspect.

[0009] In a sixth aspect, embodiments of this application provide a computer program product, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including a program or instructions, which, when executed, implement the steps of the method described in the first aspect.

[0010] In this embodiment, lane reconstruction is performed on the road based on real-time traffic conditions to obtain a reconstructed lane topology. Vehicles are then grouped according to their destination and / or travel path to obtain at least one "fish school" formation. Based on the reconstructed lane topology and the at least one "fish school" formation, a target "fish school" formation corresponding to the reconstructed lane topology is determined. Traffic scheduling is then performed on the target "fish school" formation based on the reconstructed lane topology. This embodiment uses real-time road traffic conditions to specifically plan lane resources and reconstruct lanes adapted to those conditions. Vehicles are grouped according to their destination and / or travel path, and the vehicles on the road are planned as at least one "fish school" formation. Traffic scheduling is then performed on the target "fish school" formation within the reconstructed lane topology. By combining lane reconstruction with vehicle grouping, and using traffic scheduling of "fish school" formations to achieve traffic scheduling for all vehicles, traffic flow disturbances caused by frequent lane changes can be reduced, improving traffic efficiency. Attached Figure Description

[0011] Figure 1 This is a schematic flowchart of a traffic scheduling method provided in an embodiment of this application; Figure 2 This is a flowchart illustrating a method for reserving an adjustable security area according to an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a traffic dispatching device provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0012] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0013] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0014] The traffic scheduling method, apparatus, electronic device, and readable storage medium provided in this application will be described in detail below with reference to the accompanying drawings and through specific embodiments and application scenarios.

[0015] Figure 1 The diagram illustrates a traffic scheduling method according to an embodiment of this application. This method can be executed by an electronic device, which may include an edge node, such as a vehicle. See also... Figure 1 The method may include the following steps.

[0016] Current scheduling schemes rely excessively on the cloud, placing excessive computational burden on it. Furthermore, limitations in the stability of cloud-to-edge node transmission lead to significant latency in handling complex traffic scenarios, hindering timely service delivery to edge nodes. In this embodiment, traffic scheduling is preferentially implemented through edge nodes, with the cloud intervening in complex conflict scenarios, thereby effectively reducing latency.

[0017] Step 102: Reconstruct lanes based on real-time traffic conditions to obtain the reconstructed lane topology; Real-time traffic conditions can include the number of lanes, lane direction, lane gradient, lane speed limit, traffic density, traffic volume, congestion index, weather, lane slipperiness, and the presence of any abnormal events. Different traffic conditions can determine different load patterns, including pulse, dissipative, and tidal patterns. Each load pattern has its own characteristics, and targeted analysis of traffic resource allocation based on these characteristics can improve the utilization rate of traffic resources.

[0018] Step 104: Form the vehicles into a platoon based on their destination and / or travel route to obtain at least one swarm of fish.

[0019] The vehicles can include cars, drones, electric vehicles, etc. This application embodiment groups vehicles with similar destinations into a swarm, and vehicles can travel towards their destination by joining, switching, or leaving the swarm. Once a stable swarm is formed, vehicles with similar destinations can switch routes at the convoy level, rather than changing lanes individually, thus improving the overall stability of the traffic flow.

[0020] Step 106: Based on the reconstructed lane topology and the at least one fish swarm, determine the target fish swarm corresponding to the reconstructed lane topology.

[0021] Among them, the target fish swarm is a fish swarm formed by vehicles with similar destinations in the reconstructed lane topology.

[0022] Step 108: Perform traffic scheduling on the target fish group based on the reconstructed lane topology.

[0023] In this embodiment, lane reconstruction is performed on the road based on real-time traffic conditions to obtain a reconstructed lane topology. Vehicles are then grouped according to their destination and / or travel path to obtain at least one "fish school" formation. Based on the reconstructed lane topology and the at least one "fish school" formation, a target "fish school" formation corresponding to the reconstructed lane topology is determined. Traffic scheduling is then performed on the target "fish school" formation based on the reconstructed lane topology. This embodiment uses real-time road traffic conditions to specifically plan lane resources and reconstruct lanes adapted to those conditions. Vehicles are grouped according to their destination and / or travel path, and the vehicles on the road are planned as at least one "fish school" formation. Traffic scheduling is then performed on the target "fish school" formation within the reconstructed lane topology. By combining lane reconstruction with vehicle grouping, and using traffic scheduling of "fish school" formations to achieve traffic scheduling for all vehicles, traffic flow disturbances caused by frequent lane changes can be reduced, improving traffic efficiency.

[0024] In one implementation, step 102 above, which reconstructs lanes based on real-time traffic conditions to obtain the reconstructed lane topology, may include the following steps.

[0025] Step 1021: Determine the target load pattern corresponding to the road based on the real-time traffic conditions.

[0026] Different load patterns can be determined based on different road conditions. These load patterns can include pulse patterns, dissipative patterns, and tidal patterns. Each load pattern has its own characteristics. Targeted analysis of traffic resource allocation based on the characteristics of different load patterns is beneficial to improving the utilization rate of traffic resources.

[0027] Step 1022: Adjust the position and function of the lanes based on the target load pattern corresponding to the road to obtain the reconstructed lane topology.

[0028] In this embodiment of the application, the target load pattern corresponding to the road is determined according to the real-time traffic conditions, and the position and function of the lane are adjusted based on the target load pattern corresponding to the road to obtain the reconstructed lane topology. This allows the position and function of the lane in the reconstructed lane topology to adapt to the real-time traffic conditions instead of being fixed, thereby improving the utilization rate of traffic resources.

[0029] In one implementation, step 1021 above, which determines the target load pattern corresponding to the road based on the real-time traffic conditions, may include the following step a1.

[0030] Step a1: If the traffic density of the vehicles during the first preset time period is less than a first density threshold, the average distance between the vehicles is greater than a first distance threshold, and the proportion of traffic flow of the vehicles in the same direction is less than a first flow threshold, then the target load pattern of the road is determined to be a pulse pattern; wherein, the vehicles include a first vehicle and a second vehicle traveling on the road.

[0031] Among them, the pulse pattern is characterized by the fact that when traffic volume is extremely low, lane sharing is less common, and lane resources tend to be monopolized or temporarily shared, so that the lane direction and speed limit are fully used to serve the travel vehicles, reducing the time of road load and improving traffic efficiency.

[0032] Table 1 shows a pulse morphology determination criterion provided in an embodiment of this application.

[0033] Table 1.

[0034] In some embodiments, the first preset time period can be 00:00-05:00 or other extreme off-peak periods. The first density threshold can be 0.2 vehicles / 100 meters². The first distance threshold can be 150 meters (highway section) or 80 meters (urban section). The first traffic flow threshold can be 40%.

[0035] Therefore, step 1022 above adjusts the position and function of the lane based on the target load pattern corresponding to the road to obtain the reconstructed lane topology, which may include the following steps a2-a4.

[0036] Step a2: Determine the speed buffer based on the speed of the first vehicle, and determine the load buffer based on the available area of ​​the road and the traffic density of the pulse pattern; The speed buffer is related to the speed of the first vehicle and represents the area that ensures the safety of the first vehicle at the current speed. The load buffer is related to the available area and load of the road in the pulse pattern and represents the area that ensures the safety of the first vehicle under the current road conditions. Table 2 shows the buffer determination conditions for different scenarios provided by embodiments of this application.

[0037] Table 2.

[0038] The greater the speed of the primary mode of transportation, the larger the speed buffer zone becomes. Similarly, the greater the load, the larger the load buffer zone becomes.

[0039] Step a3: If there are no intersecting obstacles between the first vehicle and the second vehicle, determine the position of the lane based on the speed buffer zone, the load buffer zone, and the real-time position of the first vehicle.

[0040] Specifically, under the constraints of speed buffer and load buffer, when it is determined that there are no intersecting obstacles between the first vehicle and other second vehicles on the road, that is, the first vehicle does not intersect with the second vehicle during its travel, the position of the lane is determined based on the real-time position of the first vehicle.

[0041] Step a4: Adjust the function of the lanes so that the reconstructed lane topology satisfies at least one of the following: (1) The speed limit of the lane shall not exceed the speed limit of the road.

[0042] The speed limit is 80 km / h for urban roads and 120 km / h for highway roads. The lane speed limit is the minimum of 1.5 times the road speed limit and 120 km / h.

[0043] (2) When the distance between the first vehicle and the second vehicle adjacent to the first vehicle is less than the second distance threshold, the speed limit of the lane is reduced to the first speed value.

[0044] The first speed value is related to the distance, the lane's speed limit, and the speed of the first vehicle. When the distance between the first vehicle and the second vehicle adjacent to it is less than a second distance threshold, driving safety is ensured by reducing the lane's speed limit.

[0045] In this embodiment of the application, when the traffic density of vehicles on the road is less than a first density threshold, the average distance between the vehicles is greater than a first distance threshold, and the proportion of traffic flow of the vehicles in the same direction is less than a first flow threshold during a first preset time period, the load pattern of the road is determined to be a pulse pattern. The traffic flow of a road in a pulse pattern is extremely low during the first preset time period, and the available traffic resources are abundant. Thus, the position and function of the lanes are adjusted on the basis of ensuring safety to meet the traffic resources required for vehicles to travel to their destination and avoid the idle and wasted traffic resources.

[0046] In one implementation, step 1021 above, which determines the target load pattern corresponding to the road based on the real-time traffic conditions, may include the following step b1.

[0047] Step b1: If the traffic flow density of the vehicle is greater than or equal to the second density threshold and less than the third density threshold, the traffic flow ratio of the vehicle in both directions is less than the second flow threshold, the lane change frequency of the vehicle is less than the first frequency threshold, and the cosine similarity between the destinations of the vehicle is less than the first similarity threshold, then the target load pattern of the road is determined to be an escaping pattern. The dispersion mode is characterized by a relatively uniform load pressure in both directions. Under this dispersion mode, lanes are reconstructed by configuring one or more virtual lanes in different directions according to the needs of the vehicles. These virtual lanes can interweave to reduce traffic disruption caused by continuous lane changes.

[0048] Table 3 shows the criteria for determining one type of emission mode provided in the embodiments of this application.

[0049] Table 3.

[0050] Therefore, step 1022 above adjusts the position and function of the lane based on the target load pattern corresponding to the road to obtain the reconstructed lane topology, which may include the following steps b2-b5.

[0051] Step b2: Determine the number of lanes in the reconstructed lane topology based on the traffic density of the road with the described eccentricity.

[0052] The number of lanes N in the reconstructed lane topology can be determined using the following formula. lane : N lane = 2+4×(ρ 0.3 / 0.4) .

[0053] Step b3: Cluster the vehicles in the road with the eccentric pattern to determine the main driving direction of the reconstructed lane topology.

[0054] The clustering method can be k-means, where k=3.

[0055] Step b4: Determine the speed limit of the reconstructed lane topology based on the speed of the vehicle, the traffic density, the preset congested traffic density, and the speed limit of the road.

[0056] The speed limit of the reconstructed lane topology can be determined using the following formula: V lane =min(V terminal ×(1+ρ current / ρ jam ),V max Among them, V terminal For the speed of the vehicle, ρ current The traffic flow density of the road is preset to the congested traffic flow density ρ. jam =0.8 vehicles / 100 meters², the speed limit V on the road max =100km / h.

[0057] Step b5: If the average speed of the vehicle on the downstream section of the road in the escape pattern is less than the second speed value, reduce the speed limit of the reconstructed lane topology to the third speed value.

[0058] The second speed value can be 30 km / h, and the third speed value can be (1-10%) * V. lane .

[0059] In this embodiment, when the traffic density of vehicles on the road is greater than or equal to a second density threshold and less than a third density threshold, the proportion of traffic flow of vehicles in both directions is less than a second flow threshold, the lane-changing frequency of vehicles is less than a first frequency threshold, and the cosine similarity between the destinations of vehicles is less than a first similarity threshold, the target load pattern of the road is determined to be an escaping pattern. In an escaping pattern, the load pressure is basically evenly distributed in both directions. The number of lanes in the reconstructed lane topology is determined based on the traffic density of the escaping pattern road. The vehicles on the escaping pattern road are clustered to determine the main driving direction of the reconstructed lane topology. The speed limit of the reconstructed lane topology is determined based on the vehicle speed, the traffic density, the preset congestion traffic density, and the road's speed limit. If the average speed of vehicles on the downstream section of the escaping pattern road is less than a second speed value, the speed limit of the reconstructed lane topology is reduced to a third speed value. The number of lanes, main driving direction, and speed limit of the reconstructed lane topology need to consider real-time road conditions to meet the traffic resources required for vehicles to travel to their destinations while avoiding conflicts between vehicles.

[0060] In one implementation, step 1021 above, which determines the target load pattern corresponding to the road based on the real-time traffic conditions, may include the following step c1.

[0061] Step c1: If the traffic flow density of the vehicles in the second time period is greater than or equal to the third density threshold, the traffic flow ratio of the vehicles in both directions is greater than or equal to the third flow threshold, and the average speed of the vehicles on the downstream section of the road is greater than or equal to the fourth speed value, then the target load pattern of the road is determined to be a tidal pattern.

[0062] Among them, tidal flow refers to the traffic flow on a road with a clear direction. The traffic flow on roads with this tidal flow pattern exhibits a convection pattern similar to boiling water. It is necessary to ensure that the traffic flow on roads with this tidal flow pattern is transported to the maximum extent possible, while also taking into account the capacity of downstream sections of the road and adjusting the opposite lanes in a timely manner to avoid congestion.

[0063] Table 4 shows the criteria for determining a tidal pattern provided in an embodiment of this application.

[0064] Table 4.

[0065] Therefore, step 1022 above adjusts the position and function of the lane based on the target load pattern corresponding to the road to obtain the reconstructed lane topology, which may include the following steps c2-c4.

[0066] Step c2: Determine the number of lanes in the reconstructed lane topology in both directions based on the traffic density of the road in the tidal pattern in both directions.

[0067] The two directions include the primary driving direction and the secondary driving direction. The number of lanes in the primary driving direction after reconstruction can be determined by the following formula: N main = N total ×Q main / (Q main +Q reverse ) +1. Where N total Q represents the total number of lanes on the road. main Traffic density in the main driving direction, Q reverse The traffic flow density is denoted as . The number of lanes in the secondary driving direction of the reconstructed lane topology is obtained by subtracting the number of lanes in the primary driving direction from the total number of lanes on the road, with at least one lane retained.

[0068] Step c3: If the average speed of the vehicle on the downstream section of the road with the tidal pattern is less than the fourth speed value, the number of lanes in the main direction of travel will be gradually reduced to a first preset value according to the speed of the vehicle.

[0069] The fourth speed value can be 20 km / h. The first preset value can be determined by the following formula: N main_new =N main ×V terminal / 30. Wherein, V terminal The speed of the primary mode of transportation.

[0070] Step c4: If the increase in traffic flow in the secondary driving direction of the reconstructed lane topology is greater than the fourth traffic flow threshold, the number of lanes in the secondary driving direction will be increased to the second preset value.

[0071] This involves reducing the number of lanes in the main driving direction and increasing the number of lanes in the secondary driving direction to cope with the increased traffic flow in the secondary driving direction and avoid congestion in the secondary driving direction.

[0072] In this embodiment, when the traffic density of vehicles on the road during the second time period is greater than or equal to a third density threshold, the proportion of traffic flow of vehicles in both directions is greater than or equal to a third flow threshold, and the average speed of vehicles on the downstream section of the road is greater than or equal to a fourth speed value, the target load pattern of the road is determined to be a tidal pattern. A tidal pattern road has a clear directional orientation, meaning that the load pressure is high in one direction and low in the other. Based on the traffic density of the tidal pattern road in both directions, the number of lanes in the reconstructed lane topology in both directions is determined. When the average speed of vehicles on the downstream section of the tidal pattern road is less than the fourth speed value, the number of lanes in the main direction of travel in the reconstructed lane topology is gradually reduced to a first preset value based on the vehicle speed. When the increase in traffic flow in the secondary direction of travel in the reconstructed lane topology is greater than the fourth flow threshold, the number of lanes in the secondary direction of travel is increased to a second preset value. By dynamically adjusting the number of lanes in the main and secondary directions of the reconstructed lane topology through changes in load pressure, the lanes with high load pressure are alleviated, and the lanes with low load pressure are fully utilized.

[0073] In one implementation, traffic scheduling of the target fish group based on the reconstructed lane topology described in step 108 above may include the following steps.

[0074] Step 1081: The motion state of the target fish swarm and the reconstructed lane topology are processed by a deep reinforcement learning model to obtain a set of actions for the vehicles in the target fish swarm to reach their destination from their current position; wherein, the set of actions includes maintaining the target fish swarm, switching the target fish swarm, or leaving the target fish swarm.

[0075] Before processing the motion state of the target fish group and the reconstructed lane topology through a deep reinforcement learning model, the above data can be feature-encoded. This includes abstracting vehicles and roads into a graph structure, with nodes representing vehicles or road segments and edges representing interaction relationships, such as following or changing lanes. The motion state of vehicles is sampled using a sliding window and statistical features are extracted. Roads are mapped into rasterized features to enhance spatial perception of roads.

[0076] In this model, the motion state of the target fish group and the reconstructed lane topology are used as the state space of the deep reinforcement learning model. Based on the analysis of the state space, the corresponding action space is predicted, namely, maintaining the target fish group, switching the target fish group, or leaving the target fish group. Multiple actions in the action space constitute the target driving path of the vehicle, providing the vehicle with the optimal path planning adapted to the real-time road conditions.

[0077] Step 1082: Drive towards the destination based on the target driving path formed by the set of actions.

[0078] In this embodiment, the motion state of the target fish swarm and the reconstructed lane topology are processed using a deep reinforcement learning model to obtain a set of actions for vehicles in the target fish swarm to travel from their current position to their destination. This set of actions includes maintaining the target fish swarm, switching the target fish swarm, or exiting the target fish swarm. Based on the target travel path formed by the set of actions, the vehicle travels to its destination. This embodiment uses a deep reinforcement learning model to analyze and predict strategies for switching fish swarms from the motion state of the target fish swarm and the reconstructed lane topology. Specifically, it uses the set of actions for vehicles to travel from their current position to their destination, switching lanes through the fish swarm, and traveling to the destination along the target travel path formed by the set of actions. This reduces traffic flow disturbance caused by frequent lane changes and improves vehicle traffic efficiency.

[0079] In one implementation, before processing the motion state of the target fish swarm and the reconstructed lane topology using a deep reinforcement learning model to obtain the action set of the vehicles in the target fish swarm from their current position to their destination, training the deep reinforcement learning model may include the following steps d1-d3: Step d1: Obtain the motion state sample of the at least one fish swarm, the reconstructed lane topology sample, and the action set sample of the at least one fish swarm; Step d2: The motion state samples of the at least one fish swarm and the reconstructed lane topology samples are processed by an initial deep reinforcement learning model to obtain the predicted action set of the at least one fish swarm and the reward corresponding to the predicted action set. Step d3 involves training the initial deep reinforcement learning model using a loss function on the predicted action set of the at least one fish group, the action set samples of the at least one fish group, and the reward, until the convergence condition is met, thus obtaining the deep reinforcement learning model.

[0080] In this embodiment, an initial deep reinforcement learning model is trained using motion state samples of at least one fish swarm, the reconstructed lane topology samples, and action set samples of the at least one fish swarm. The motion state samples of the at least one fish swarm and the reconstructed lane topology samples are input into the initial deep reinforcement learning model for processing to obtain a predicted action set for the at least one fish swarm and a corresponding reward. The predicted action set, action set samples, and reward are then iteratively optimized using a loss function to optimize the parameters of the initial deep reinforcement learning model until convergence is achieved, resulting in the deep reinforcement learning model. This deep reinforcement learning model, trained using motion state samples of at least one fish swarm, the reconstructed lane topology samples, and action set samples of the at least one fish swarm, is adaptable to the current real-time road conditions and fish swarms, enabling accurate prediction of the target fish swarm's action set and providing an accurate basis for planning the target fish swarm's target travel path.

[0081] In one implementation, step 104 above, which involves platooning the vehicles based on their destination and / or travel route to obtain at least one swarm of fish, may include the following steps.

[0082] Step 1041: Determine whether the destinations and / or travel paths of the vehicles are the same based on the cosine similarity between the destinations of the vehicles, the road segment overlap rate between the travel paths of the vehicles, the speed difference between the vehicles, and the distance between the current positions of the vehicles.

[0083] Step 1042: Group the vehicles whose cosine similarity is greater than or equal to a second similarity threshold, whose road segment overlap rate is greater than or equal to a third preset value, whose speed difference is less than or equal to a fourth speed value, and whose distance between their current positions is less than or equal to a third distance threshold into a fish swarm; wherein the vehicles in the fish swarm have the same destination and / or travel path.

[0084] The second similarity threshold can be 0.95, the third preset value can be 0.7, the fourth speed value can be 5, and the third distance threshold can be 3.

[0085] In some embodiments, the fish school is determined through the following steps.

[0086] S1 discretizes the travel path of the vehicle into a sequence of key nodes P={p1,p2,...,pn}, where pi=(xi,yi).

[0087] S2 extracts the short-term and long-term directions of the driving path.

[0088] S3. Determine the cosine similarity between each pair of vehicles based on the short-term and long-term directions of their travel paths.

[0089] Wherein, cosine similarity Simcos = (vA) vB) / (∥vA∥∥vB∥).

[0090] S4. Determine the path overlap rate between any two vehicles based on the sequence of key nodes corresponding to the vehicle's travel path.

[0091] Among them, the path overlap rate Overlap = number of shared key nodes / max(total number of nodes A, total number of nodes B).

[0092] S5, vehicles with cosine similarity and path overlap rate greater than or equal to a preset threshold are classified into the same candidate fleet.

[0093] The preset threshold for cosine similarity can be 0.95, and the preset threshold for path overlap rate can be 0.7.

[0094] Furthermore, the scores of the cosine similarity and the path overlap rate can be compared with a preset threshold to determine whether they belong to the same fish group: Score = α Simcos+β Overlap, where α + β = 1, α can be 0.6, and β can be 0.4. The preset threshold corresponding to the score can be 0.85.

[0095] In some embodiments, the health status of the fish school can also be assessed, and the health value of the fish school can be determined by the following formula: Hschool = (∑v∈school(v vehicle condition score + v communication score)) / platoon size, where the vehicle condition score is determined based on real-time road conditions, and the communication score is determined based on the communication quality of vehicles within the swarm. The second preset value can be 0.8.

[0096] In this embodiment of the application, the similarity between the destinations of the vehicles is measured by determining whether the destinations and / or travel paths of the vehicles are the same, the road segment overlap rate between the travel paths of the vehicles, the speed difference between the vehicles, and the distance between the current positions of the vehicles. Vehicles whose cosine similarity is greater than or equal to a second similarity threshold, whose road segment overlap rate is greater than or equal to a third preset value, whose speed difference is less than or equal to a fourth speed value, and whose distance between the current positions is less than or equal to a third distance threshold are grouped into a fish swarm. The vehicles in the fish swarm have the same destination and / or travel path.

[0097] In one implementation, before step 108 above, which performs traffic scheduling on the target fish swarm based on the reconstructed lane topology, the method further includes: determining that the first vehicle has the right to use the lane if the distance between the first vehicle and the second vehicle adjacent to it in front is greater than a fourth distance threshold and the priority of the first vehicle is higher than the priority of the third vehicle currently using the lane. After obtaining the right to use the reconstructed lane topology, the first vehicle can use the lane to travel along the target driving path constituted by the action set, thereby avoiding lane abuse and traffic anomalies.

[0098] In one implementation, when the vehicle switches to the target fish group, the traffic scheduling method further includes the following steps.

[0099] Step e1: Determine the longitudinal safety distance of the vehicle based on its speed, preset time limit, preset acceleration limit, and preset distance limit.

[0100] Among them, the longitudinal safety distance d of the vehicle long It can be determined using the following formula: d long =v t react + v 2 / 2a max +δ res .

[0101] Where v is the speed of the vehicle, and t is the preset time limit. react =1.5s, preset limit acceleration a max =3m / s 2 Preset limit distance δ res =5m.

[0102] Step e2: Determine the lateral safety width of the vehicle based on the vehicle's steering angle when switching the convoy, the width of the vehicle, and the longitudinal safety distance.

[0103] Among them, the lateral safety distance w of vehicles lat It can be determined using the following formula: w lat =w vehicle +2 tan(θ steer ) d long .

[0104] Where, θ steer For the steering angle, w vehicle Width of vehicles.

[0105] Step e3: Determine the adjustable safety zone of the vehicle based on the longitudinal safety distance and the lateral safety width.

[0106] The adjustable safety zone may include a rectangular area corresponding to when a vehicle changes lanes in a straight line, and the length of this rectangular area is d. long Width is w lat The adjustable safety zone can also include a sector area corresponding to when a vehicle changes lanes on a curve, the radius of which is d. long Arc length is based on steering angle θ steer Sure.

[0107] In some alternative embodiments, the longitudinal safety distance d long It is also affected by weather or light intensity, and the corrected longitudinal safety distance d final =d long (1+k rain I rain +k night I night ), where the weight k corresponds to the weather. rain =0.2, the weight k corresponding to the light intensity night =0.1, I rain I is the indicator variable corresponding to the weather. night This is an indicator variable corresponding to light intensity.

[0108] Step e4: Reserve the adjustable safety area for the vehicle.

[0109] In this embodiment, when the vehicle switches to the target fish school, the longitudinal and lateral safety distances of the vehicle are determined based on its motion state and real-time road conditions. Then, an adjustable safety zone is determined based on these distances to reserve for the vehicle when switching to the target fish school. This adjustable safety zone can be reserved for the vehicle based on its priority or time information. For example, if a vehicle has a high priority, the adjustable safety zone is reserved for it when switching to the target fish school. Alternatively, if vehicles with the same priority switch to the target fish school earlier, the adjustable safety zone is reserved for that vehicle, ensuring safe switching of the target fish school.

[0110] Figure 2 This document illustrates a flowchart of a method for reserving an adjustable security area according to an embodiment of this application. (See attached diagram.) Figure 2 The method may include the following steps.

[0111] Step 210: The first vehicle broadcasts a request to reserve an adjustable safety zone.

[0112] Step 220: The second vehicle responds to the request, indicating whether it agrees to reserve the vehicle. If yes, proceed to step 230; otherwise, return to step 210.

[0113] Step 230: The first vehicle switches to the convoy to obtain the reserved adjustable safety zone, and the second vehicle adjusts its position based on the adjustable safety zone.

[0114] Step 240: The first vehicle broadcasts a request to reserve an adjustable security area to expand its communication range, and the edge node broadcasts the request to reserve an adjustable security area via a secondary broadcast.

[0115] Step 250: The third mode of transportation responds to the request, indicating whether it agrees to the reservation. If yes, proceed to step 260; otherwise, return to step 240.

[0116] Step 260: The first vehicle switches convoys to obtain the reserved, adjustable safety zone.

[0117] Step 270: In the event that coordination between the first, second, and third vehicles fails, the edge node sends a global coordination request to the cloud.

[0118] Step 280: The cloud dynamically adjusts the lanes based on a global coordination request.

[0119] In one implementation, the above method may further include step e5, where if the number of failures to reserve the adjustable safety area for the vehicle is greater than or equal to a fourth preset value or the range of the adjustable safety area is greater than or equal to half the width of the road, a global coordination request is sent to the cloud so that the cloud adjusts the reconstructed lane topology.

[0120] In this embodiment of the application, if the number of failures in reserving the adjustable safety area for the vehicle is greater than or equal to a fourth preset value, or if the range of the adjustable safety area is greater than or equal to half the width of the road, a global coordination request is sent to the cloud. After receiving the global coordination request, the cloud knows that the current situation requires cloud intervention, and thus the cloud adjusts the reconstructed lane topology, including adjusting the lane width or reducing the lane speed, to avoid being unable to cope with the failure of edge node negotiation.

[0121] In one implementation, the traffic scheduling method described above may further include step f1: when the target fish group has the highest priority, a communication method corresponding to the distance between the target fish group and other fish groups is set for the target fish group, so as to remind other fish groups to give way to the target fish group through the communication method; wherein, the priority is determined based on the identification information of the fish group and a preset priority matrix, and the preset priority matrix is ​​determined based on the task type of the vehicle, the traffic demand intensity of the vehicle, and the segment value of the road.

[0122] The mission type of a vehicle can include emergency, special, and general. The traffic demand intensity of a vehicle can be classified as high, medium, and low. The value of a road segment can be assessed based on factors such as cost and ability to handle traffic accidents.

[0123] The communication methods corresponding to the distance may include: ① 5G-C-V2X global broadcast (1km range, 1Hz frequency); ② LTE-V2N emergency channel (5km range, 100ms latency); ③ DSRC point-to-point interaction (200m range, 3ms ultra-low latency).

[0124] The avoidance action that reminds other fish groups to avoid the target fish group can adopt the 8 types of avoidance action codes defined by the ASN.1 standard, such as lane clearance, speed synchronization, path freezing, etc.

[0125] The avoidance behavior of other fish groups towards the target fish group can include: ① When the distance between other fish groups and the target fish group is close (<100m): emergency avoidance is triggered, and other fish groups around the target fish group perform hard lane clearing; ② When the distance between other fish groups and the target fish group is medium (100-500m): flexible speed coordination is initiated; ③ When the distance between other fish groups and the target fish group is long (>500m): path optimization and diversion are induced through edge nodes. Furthermore, an adjustable safety zone of 1.2 times the size is reserved for other fish groups before avoidance to ensure their safety. After avoidance, credit points are awarded to other fish groups that cooperate in avoiding the target fish group, optimizing the acceptance of other fish groups sharing road usage rights with the target fish group and encouraging other fish groups to avoid the target fish group with higher priority.

[0126] In one implementation, the traffic scheduling of the target fish group based on the reconstructed lane topology in step 108 above may include the following steps.

[0127] Step f2: Construct a target travel path from the current location to the destination for the target fish swarm based on the Bézier curve.

[0128] Specifically, based on Bézier curves, the safe zone of the target fish convoy can be predicted for the next 10-30 seconds, indicating whether the convoy should switch to reach the safe zone relative to its current location. During the convoy's movement, the safe zone for the next 10-30 seconds is continuously predicted. Multiple consecutive safe zones constitute the target travel path from the current location to the destination, thus providing the target fish convoy with a path from its current location to its destination.

[0129] Step f3: Drive towards the destination based on the target driving route.

[0130] In this embodiment, when the target fish group has the highest priority, a communication method corresponding to the distance between the target fish group and other fish groups is set up to remind other fish groups to avoid the target fish group. Then, a target travel path from the current location to the destination is constructed for the target fish group based on a Bézier curve, and traffic resources are allocated to the target fish group so that it travels to the destination according to the target travel path, thus satisfying the traffic resource needs of the highest priority target fish group.

[0131] In one implementation, the traffic scheduling method described above may further include: sending a global coordination request to the cloud when at least two of the fish groups have the highest priority, so that the cloud generates a target driving path for each of the fish groups based on the Pareto optimal path algorithm, wherein the target driving paths of each fish group are staggered on the road.

[0132] In this embodiment of the application, when at least two of the fish groups have the highest priority, there is a decision conflict among the vehicles in multiple fish groups. By sending a global coordination request to the cloud, the cloud receives the global coordination request and learns that the current situation requires cloud intervention. The cloud then generates the target driving path for each fish group based on the Pareto optimal path algorithm, so that the target driving paths of each fish group are interspersed on the road, thus satisfying the traffic resource needs of multiple fish groups.

[0133] In addition, in the event of cloud-based failure, the manual control channel is automatically activated to ensure that traffic scheduling remains uninterrupted.

[0134] In one implementation, the traffic scheduling method described above may further include: gradually reducing the priority of the target fish swarm based on the travel progress of the target travel path.

[0135] In this embodiment of the application, as the vehicles in the target fish swarm travel towards their destination along the target travel path, the priority of the target fish swarm is gradually reduced according to the travel progress in order to avoid excessive road occupation.

[0136] In one implementation, the target fish swarm with the highest priority can be split and reorganized. For example, the number of vehicles in the target fish swarm can be reduced to three, making the target fish swarm more flexible and thus enabling it to travel from its current location to its destination efficiently.

[0137] In one implementation, the road includes a three-dimensional road space; the three-dimensional road space includes multiple three-dimensional lanes; step 102 above reconstructs the lanes based on real-time traffic conditions to obtain the reconstructed lane topology, which may include: when the traffic density of the three-dimensional lanes in the three-dimensional road space is greater than a fourth density threshold, adjusting the lane position, lane height and lane function of the three-dimensional lanes to obtain the reconstructed three-dimensional lane topology.

[0138] The number of lanes in the reconstructed three-dimensional lane topology is determined based on the traffic density of the three-dimensional lanes and the maximum traffic density of the three-dimensional road space.

[0139] Among them, the number of lanes N in the reconstructed 3D lane topology newIt can be determined using the following formula: N new = ρ max / ρ z Wherein, ρ z Let ρ be the traffic density of the three-dimensional lane. max This represents the maximum traffic density in the three-dimensional road space.

[0140] The lane height of the reconstructed three-dimensional lane topology is determined based on the height of the three-dimensional lanes and the number of lanes in the reconstructed three-dimensional lane topology.

[0141] The lane height ΔZ′ of the reconstructed 3D lane topology can be determined by the following formula: ΔZ′=ΔZ / N new Where ΔZ is the height of the three-dimensional lane. For example, ΔZ = {10m (small drone); 30m (cargo aircraft); 50m (manned flying car)}, where the vehicles in the three-dimensional road space can be extended to small drones, cargo aircraft, and manned flying cars.

[0142] In one implementation, based on the adjustable safety zone constructed in two-dimensional space, a height variable is added to the three-dimensional road space. The adjustable safety zone can be expanded into a three-dimensional ellipsoidal safety cavity, as shown in the following formula: (x x0) 2 / a 2 + (y y0) 2 / b 2 +(z z0) 2 / c 2 ≤1. Where a=1.2d long b=1.5w lat c = ΔZ / 2.

[0143] In one implementation, the traffic scheduling method described above may further include: determining the height safety distance between three-dimensional lanes in the reconstructed three-dimensional lane topology based on the speed of the vehicle and a preset height limit of the three-dimensional road space.

[0144] In this embodiment, the vehicle can be extended to a small drone, a cargo aircraft, or a manned flying car. In the three-dimensional road space, a certain distance must be maintained between the three-dimensional lanes in the reconstructed three-dimensional lane topology. The height safety distance between the three-dimensional lanes in the reconstructed three-dimensional lane topology can be determined based on the speed of the vehicle and the preset height limit of the three-dimensional road space, ensuring that adjacent three-dimensional lanes also have a safety space in height, thereby guaranteeing the safety of the vehicle when switching target fish groups across layers.

[0145] In one implementation, based on the aforementioned construction of a target travel path from the current location to the destination for the highest priority vehicle in two-dimensional space, a similar three-dimensional travel path can be constructed for the vehicle in three-dimensional road space, allowing the vehicle to travel to the destination based on this three-dimensional travel path. In the three-dimensional road space, the avoidance of the target fish school by other fish schools can include: ① When the height between other fish schools and the target fish school is close (<50m): other fish schools below the target fish school immediately descend, while other fish schools above the target fish school ascend; ② When the height between other fish schools and the target fish school is medium (50-200m): other fish schools in the three-dimensional lanes adjacent to the target fish school's lane laterally disperse to form a circular channel; ③ When the height between other fish schools and the target fish school is far (>200m): instructing other fish schools on the periphery to enter the three-dimensional lane at a spare height.

[0146] In this embodiment, edge nodes perceive and calculate the motion state of vehicles and real-time road conditions in real time, transforming roads into dynamically reconfigurable traffic resources. This breaks through the fixed limitations of lanes, allowing vehicles to autonomously plan travel routes within a reasonable space and maximizing the flexible utilization of traffic resources. A vehicle platooning mechanism is employed, grouping vehicles into platoons with the same destination. Lane changes are achieved by switching platoons, reducing traffic flow disturbances caused by frequent lane changes and improving overall traffic flow stability. A dual-elastic buffer mechanism, consisting of speed and load buffers, is also introduced to adapt to vehicle speed and road load, providing safety redundancy for vehicle platooning.

[0147] It should be noted that the traffic dispatching method provided in this application embodiment can be executed by a traffic dispatching device or a control module within that traffic dispatching device for executing the traffic dispatching method. This application embodiment uses the execution of the method by a traffic dispatching device as an example to illustrate the traffic dispatching device provided in this application embodiment.

[0148] Figure 3A schematic diagram of a traffic dispatching device according to an embodiment of this application is shown. See also: Figure 3 The device 300 may include: a lane reconstruction module 31, a platooning module 32, a determination module 33, and a scheduling module 34.

[0149] The system includes: a lane reconstruction module 31 for reconstructing lanes based on real-time traffic conditions to obtain a reconstructed lane topology; a platooning module 32 for platooning vehicles based on their destination and / or travel path to obtain at least one fish swarm; a determination module 33 for determining a target fish swarm corresponding to the reconstructed lane topology based on the reconstructed lane topology and the at least one fish swarm; and a scheduling module 34 for performing traffic scheduling on the target fish swarm based on the reconstructed lane topology.

[0150] In one implementation, the lane reconstruction module 31 described above can be used to: determine the target load pattern corresponding to the road based on the real-time traffic conditions; adjust the position and function of the lane based on the target load pattern corresponding to the road, and obtain the reconstructed lane topology.

[0151] In one implementation, the lane reconstruction module 31 described above can be specifically used to: determine the target load pattern of the road as a pulse pattern when the traffic density of the vehicles during a first preset time period is less than a first density threshold, the average distance between the vehicles is greater than a first distance threshold, and the proportion of traffic flow of the vehicles in the same direction is less than a first flow threshold; wherein the vehicles include a first vehicle and a second vehicle traveling on the road; determine a speed buffer based on the speed of the first vehicle, and determine a load buffer based on the available area and traffic density of the road in the pulse pattern; determine the position of the lane based on the speed buffer, the load buffer, and the real-time position of the first vehicle when there are no intersecting obstacles between the first vehicle and the second vehicle; adjust the function of the lane so that the reconstructed lane topology satisfies at least one of the following: the speed limit of the lane does not exceed the speed limit of the road; and reduce the speed limit of the lane to a first speed value when the distance between the first vehicle and the second vehicle adjacent to the first vehicle is less than a second distance threshold.

[0152] In one implementation, the lane reconstruction module 31 described above can be specifically used to: determine the target load pattern of the road as an escaping pattern when the traffic flow density of the vehicle is greater than or equal to a second density threshold and less than a third density threshold, the traffic flow ratio of the vehicle in both directions is less than a second flow threshold, the lane change frequency of the vehicle is less than a first frequency threshold, and the cosine similarity between the destinations of the vehicle is less than a first similarity threshold; determine the number of lanes in the reconstructed lane topology based on the traffic flow density of the road with the escaping pattern; cluster the vehicles on the road with the escaping pattern to determine the main driving direction of the reconstructed lane topology; determine the speed limit of the reconstructed lane topology based on the speed of the vehicle, the traffic flow density, a preset congestion traffic flow density, and the speed limit of the road; and reduce the speed limit of the reconstructed lane topology to a third speed value when the average speed of the vehicle on the downstream segment of the road with the escaping pattern is less than a second speed value.

[0153] In one implementation, the lane reconstruction module 31 described above can be specifically used to: determine that the target load pattern of the road is a tidal pattern when the traffic flow density of the vehicle in the second time period is greater than or equal to a third density threshold, the traffic flow ratio of the vehicle in both directions is greater than or equal to a third flow threshold, and the average speed of the vehicle on the downstream section of the road is greater than or equal to a fourth speed value; determine the number of lanes in the reconstructed lane topology in both directions based on the traffic flow density of the road in the tidal pattern in both directions; when the average speed of the vehicle on the downstream section of the road in the tidal pattern is less than the fourth speed value, gradually reduce the number of lanes in the main driving direction to a first preset value based on the speed of the vehicle; and when the increase in traffic flow in the secondary driving direction of the reconstructed lane topology is greater than the fourth flow threshold, increase the number of lanes in the secondary driving direction to a second preset value.

[0154] In one implementation, the scheduling module 34 described above can be used to: process the motion state of the target fish swarm and the reconstructed lane topology through a deep reinforcement learning model to obtain a set of actions for the vehicles in the target fish swarm to travel from their current position to their destination; wherein, the set of actions includes maintaining the target fish swarm, switching the target fish swarm, or exiting the target fish swarm; and travel to the destination based on the target driving path formed by the set of actions.

[0155] In one implementation, the aforementioned apparatus 300 may further include a training module, which can be used to: acquire motion state samples of the at least one school of fish, the reconstructed lane topology samples, and action set samples of the at least one school of fish; process the motion state samples of the at least one school of fish and the reconstructed lane topology samples through an initial deep reinforcement learning model to obtain a predicted action set of the at least one school of fish and a reward corresponding to the predicted action set; train the initial deep reinforcement learning model with the predicted action set of the at least one school of fish, the action set samples of the at least one school of fish, and the reward through a loss function until convergence is achieved, thereby obtaining the deep reinforcement learning model.

[0156] In one implementation, the formation module 32 described above can be used to: determine whether the destinations and / or travel paths of the vehicles are the same based on the cosine similarity between the destinations of the vehicles, the road segment overlap rate between the travel paths of the vehicles, the speed difference between the vehicles, and the distance between the current positions of the vehicles; and form a fish swarm team by grouping the vehicles whose cosine similarity is greater than or equal to a second similarity threshold, whose road segment overlap rate is greater than or equal to a third preset value, whose speed difference is less than or equal to a fourth speed value, and whose distance between the current positions is less than or equal to a third distance threshold; wherein the destinations and / or travel paths of the vehicles in the fish swarm team are the same.

[0157] In one implementation, the determining module 33 described above can also be used to: determine the longitudinal safety distance of the vehicle based on the vehicle's speed and preset time limit, preset acceleration limit, and preset distance limit; determine the lateral safety width of the vehicle based on the vehicle's turning angle when switching to the target fish school, the vehicle's width, and the longitudinal safety distance; determine the adjustable safety area of ​​the vehicle based on the longitudinal safety distance and the lateral safety width; and reserve the adjustable safety area for the vehicle.

[0158] In one implementation, the aforementioned device 300 may further include a sending module, which can be used to: send a global coordination request to the cloud when the number of failures to reserve the adjustable safety area for the vehicle is greater than or equal to a fourth preset value or the range of the adjustable safety area is greater than or equal to half the width of the road, so that the cloud can adjust the reconstructed lane topology.

[0159] In one implementation, the aforementioned device 300 may further include a configuration module, which can be used to: when the target fish group has the highest priority, set a communication method corresponding to the distance between the target fish group and other fish groups, so as to remind other fish groups to avoid the target fish group through the communication method; wherein, the priority is determined based on the identification information of the fish group and a preset priority matrix, and the preset priority matrix is ​​determined based on the task type of the vehicle, the traffic demand intensity of the vehicle, and the segment value of the road.

[0160] In one implementation, the scheduling module 34 described above can also be used to: construct a target travel path from the current position to the destination for the target fish group based on a Bézier curve; and travel to the destination based on the target travel path.

[0161] In one implementation, the scheduling module 34 described above can also be used to: send a global coordination request to the cloud when at least two of the fish groups have the highest priority, so that the cloud generates a target driving path for each of the fish groups based on the Pareto optimal path algorithm; wherein the target driving paths of each fish group are staggered on the road.

[0162] In one implementation, the scheduling module 34 described above can also be used to: gradually reduce the priority of the target fish swarm based on the travel progress of the target travel path.

[0163] In one implementation, the road includes a three-dimensional road space; the three-dimensional road space includes multiple three-dimensional lanes; the lane reconstruction module 31 described above can be used to: adjust the lane position, lane height, and lane function of the three-dimensional lanes when the traffic density of the three-dimensional lanes in the three-dimensional road space is greater than a fourth density threshold, to obtain a reconstructed three-dimensional lane topology; wherein, the number of lanes in the reconstructed three-dimensional lane topology is determined based on the traffic density of the three-dimensional lanes and the maximum traffic density of the three-dimensional road space; the lane height of the reconstructed three-dimensional lane topology is determined based on the height of the three-dimensional lanes and the number of lanes in the reconstructed three-dimensional lane topology.

[0164] In one implementation, the determining module 33 described above can also be used to: determine the height safety distance between the three-dimensional lanes in the reconstructed three-dimensional lane topology based on the speed of the vehicle and the preset height limit of the three-dimensional road space.

[0165] The traffic dispatching device in this application embodiment can be an electronic device or a component of an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a mobile phone, tablet computer, laptop computer, handheld computer, in-vehicle electronic device, mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc. It can also be a server, network attached storage (NAS), personal computer (PC), television (TV), ATM, or self-service machine, etc. The embodiments of this application do not specifically limit the device.

[0166] The traffic dispatching device in this application embodiment can be a device with an operating system. This operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit the specific operating system used.

[0167] The traffic dispatching device provided in this application embodiment can achieve... Figures 1 to 2 The various processes implemented in the method implementation examples will not be described again here to avoid repetition.

[0168] Based on the same technical concept, embodiments of this application also provide an electronic device for performing the above-described traffic scheduling method. Figure 4 This is a schematic diagram of the structure of an electronic device to implement the various embodiments of this application. The electronic device can vary significantly due to differences in configuration or performance, and may include a processor 401, a communications interface 402, a memory 403, and a communication bus 404. The processor 401, communications interface 402, and memory 403 communicate with each other via the communication bus 404. The processor 401 can call a computer program stored in the memory 403 and executable on the processor 401 to perform the various steps of the traffic scheduling method embodiments described above, achieving the same technical effects. To avoid repetition, further details are omitted here.

[0169] It should be noted that the electronic devices in the embodiments of this application include servers, terminals, or other devices besides terminals. For example, automobiles, robots, and handheld devices.

[0170] The above electronic device structure does not constitute a limitation on the electronic device. An electronic device may include more or fewer components than illustrated, or combine certain components, or arrange them differently. For example, an input unit may include a Graphics Processing Unit (GPU) and a microphone, and a display unit may use a liquid crystal display (LCD), organic light-emitting diode (OLED), or other similar display panels. User input units include at least one of a touch panel and other input devices. A touch panel is also called a touchscreen. Other input devices may include, but are not limited to, physical keyboards, function keys (such as volume control buttons, power buttons, etc.), trackballs, mice, and joysticks, which will not be elaborated further here.

[0171] Memory can be used to store software programs and various data. Memory can primarily include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area can store the operating system, application programs or instructions required for at least one function (such as sound playback, image playback, etc.). Furthermore, memory can include volatile memory or non-volatile memory, or both. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (Synchlink DRAM, SLDRAM), and direct memory bus RAM (DRRAM).

[0172] The processor may include one or more processing units; optionally, the processor integrates an application processor and a modem processor, wherein the application processor mainly handles operations related to the operating system, user interface, and applications, while the modem processor mainly handles wireless communication signals, such as a baseband processor. It is understood that the aforementioned modem processor may also not be integrated into the processor.

[0173] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described traffic scheduling method embodiments and achieve the same technical effects. To avoid repetition, they will not be described again here.

[0174] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0175] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above-described traffic scheduling method embodiments and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0176] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0177] This application also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes a program or instructions. When the program or instructions are executed, they implement the various processes of the above-described traffic scheduling method embodiments and can achieve the same technical effects. To avoid repetition, they will not be described again here.

[0178] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0179] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0180] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A traffic scheduling method, characterized in that, include: Based on real-time traffic conditions, lane reconstruction is performed on the road to obtain the reconstructed lane topology; The vehicles are grouped together based on their destination and / or travel route to obtain at least one fish swarm. Based on the reconstructed lane topology and the at least one fish swarm, determine the target fish swarm corresponding to the reconstructed lane topology; Traffic scheduling is performed on the target fish group based on the reconstructed lane topology.

2. The method according to claim 1, characterized in that, The process of reconstructing lanes based on real-time traffic conditions to obtain the reconstructed lane topology includes: The target load pattern corresponding to the road is determined based on the real-time traffic conditions. The position and function of the lanes are adjusted based on the target load pattern corresponding to the road to obtain the reconstructed lane topology.

3. The method according to claim 2, characterized in that, Determining the target load pattern corresponding to the road based on the real-time traffic conditions includes: If the traffic density of the vehicles is less than a first density threshold, the average distance between the vehicles is greater than a first distance threshold, and the proportion of traffic flow of the vehicles in the same direction is less than a first flow threshold during a first preset time period, the target load pattern of the road is determined to be a pulse pattern; wherein the vehicles include a first vehicle and a second vehicle traveling on the road. The process of adjusting the position and function of the lanes based on the target load pattern corresponding to the road to obtain the reconstructed lane topology includes: A speed buffer is determined based on the speed of the first vehicle, and a load buffer is determined based on the available area of ​​the road and the traffic density of the pulse pattern. When there are no intersecting obstacles between the first vehicle and the second vehicle, the position of the lane is determined based on the speed buffer, the load buffer, and the real-time position of the first vehicle; Adjust the functionality of the lanes so that the reconstructed lane topology satisfies at least one of the following: The speed limit of the lane shall not exceed the speed limit of the road. If the distance between the first vehicle and the second vehicle adjacent to the first vehicle is less than a second distance threshold, the speed limit of the lane is reduced to a first speed value.

4. The method according to claim 2, characterized in that, Determining the target load pattern corresponding to the road based on the real-time traffic conditions includes: If the traffic flow density of the vehicle is greater than or equal to the second density threshold and less than the third density threshold, the traffic flow ratio of the vehicle in both directions is less than the second flow threshold, the lane change frequency of the vehicle is less than the first frequency threshold, and the cosine similarity between the destinations of the vehicle is less than the first similarity threshold, then the target load pattern of the road is determined to be an escaping pattern. The process of adjusting the position and function of the lanes based on the target load pattern corresponding to the road to obtain the reconstructed lane topology includes: The number of lanes in the reconstructed lane topology is determined based on the traffic density of the road with the described eccentricity pattern. Cluster the vehicles on the road with the aforementioned dispersion pattern to determine the main driving direction of the reconstructed lane topology; The speed limit of the reconstructed lane topology is determined based on the speed of the vehicle, the traffic density, the preset congested traffic density, and the speed limit of the road. If the average speed of the vehicle on the downstream section of the road in the escape pattern is less than the second speed value, the speed limit of the reconstructed lane topology is reduced to the third speed value.

5. The method according to claim 2, characterized in that, Determining the target load pattern corresponding to the road based on the real-time traffic conditions includes: If the traffic flow density of the vehicles in the second time period is greater than or equal to the third density threshold, the traffic flow ratio of the vehicles in both directions is greater than or equal to the third flow threshold, and the average speed of the vehicles on the downstream section of the road is greater than or equal to the fourth speed value, the target load pattern of the road is determined to be a tidal pattern. The process of adjusting the position and function of the lanes based on the target load pattern corresponding to the road to obtain the reconstructed lane topology includes: The number of lanes in the reconstructed lane topology in both directions is determined based on the traffic density in both directions of the road with the tidal pattern. If the average speed of the vehicle on the downstream section of the road with the tidal pattern is less than the fourth speed value, the number of lanes in the main direction of travel will be gradually reduced to a first preset value according to the speed of the vehicle. If the increase in traffic flow in the secondary driving direction after the reconstructed lane topology exceeds the fourth traffic flow threshold, the number of lanes in the secondary driving direction will be increased to the second preset value.

6. The method according to claim 1, characterized in that, The traffic scheduling of the target fish group based on the reconstructed lane topology includes: The motion state of the target fish swarm and the reconstructed lane topology are processed by a deep reinforcement learning model to obtain a set of actions for the vehicles in the target fish swarm to reach their destination from their current position; wherein, the set of actions includes maintaining the target fish swarm, switching the target fish swarm, or leaving the target fish swarm. Travel to the destination based on the target driving path formed by the set of actions.

7. The method according to claim 6, characterized in that, The method further includes: Obtain motion state samples of the at least one fish swarm, the reconstructed lane topology samples, and action set samples of the at least one fish swarm; The motion state samples of the at least one fish swarm and the reconstructed lane topology samples are processed through an initial deep reinforcement learning model to obtain the predicted action set of the at least one fish swarm and the reward corresponding to the predicted action set. The predicted action set of the at least one fish group, the action set sample of the at least one fish group, and the reward are used to train the initial deep reinforcement learning model through a loss function until the convergence condition is met, thus obtaining the deep reinforcement learning model.

8. The method according to claim 1, characterized in that, The process of grouping the vehicles based on their destination and / or travel route to obtain at least one fish swarm includes: Based on the cosine similarity between the destinations of the vehicles, the road segment overlap rate between the travel paths of the vehicles, the speed difference between the vehicles, and the distance between the current positions of the vehicles, it is determined whether the destinations and / or travel paths of the vehicles are the same. Vehicles whose cosine similarity is greater than or equal to a second similarity threshold, whose road segment overlap rate is greater than or equal to a third preset value, whose speed difference is less than or equal to a fourth speed value, and whose distance between their current positions is less than or equal to a third distance threshold are grouped into a fish swarm; wherein the vehicles in the fish swarm have the same destination and / or travel path.

9. The method according to claim 6, characterized in that, When the vehicle switches to the target fish school, the method further includes: The longitudinal safety distance of the vehicle is determined based on its speed, preset time limit, preset acceleration limit, and preset distance limit. The lateral safety width of the vehicle is determined based on the turning angle of the vehicle switching the target fish school, the width of the vehicle, and the longitudinal safety distance. The adjustable safety zone of the vehicle is determined based on the longitudinal safety distance and the lateral safety width. The adjustable safety area is reserved for the vehicle.

10. The method according to claim 9, characterized in that, The method further includes: If the number of failures to reserve the adjustable safety zone for the vehicle is greater than or equal to a fourth preset value, or if the range of the adjustable safety zone is greater than or equal to half the width of the road, a global coordination request is sent to the cloud so that the cloud can adjust the reconstructed lane topology.

11. The method according to claim 1, characterized in that, The method further includes: When the target fish group has the highest priority, a communication method corresponding to the distance between the target fish group and other fish groups is set for the target fish group. This communication method is used to remind other fish groups to avoid the target fish group. The priority is determined based on the fish group's identification information and a preset priority matrix. The preset priority matrix is ​​determined based on the vehicle's task type, the vehicle's traffic demand intensity, and the road segment value.

12. The method according to claim 11, characterized in that, The traffic scheduling of the target fish group based on the reconstructed lane topology includes: Based on Bézier curves, construct a target travel path for the target fish group from its current location to its destination; Drive towards the destination based on the target driving route.

13. The method according to claim 1, characterized in that, The method further includes: If at least two of the fish swarms have the highest priority, a global coordination request is sent to the cloud so that the cloud generates a target driving path for each fish swarm based on the Pareto optimal path algorithm; wherein the target driving paths of each fish swarm are interspersed on the road.

14. The method according to claim 12, characterized in that, The method further includes: Based on the progress of the target travel path, the priority of the target fish group is gradually reduced.

15. The method according to claim 1, characterized in that, The road includes a three-dimensional road space; the three-dimensional road space includes multiple three-dimensional lanes; the lane reconstruction based on real-time traffic conditions to obtain the reconstructed lane topology includes: When the traffic density of the three-dimensional lanes in the three-dimensional road space is greater than the fourth density threshold, the lane position, lane height and lane function of the three-dimensional lanes are adjusted to obtain the reconstructed three-dimensional lane topology. The number of lanes in the reconstructed three-dimensional lane topology is determined based on the traffic density of the three-dimensional lanes and the maximum traffic density of the three-dimensional road space; the lane height of the reconstructed three-dimensional lane topology is determined based on the height of the three-dimensional lanes and the number of lanes in the reconstructed three-dimensional lane topology.

16. The method according to claim 15, characterized in that, The method further includes: The height safety distance between the three-dimensional lanes in the reconstructed three-dimensional lane topology is determined based on the speed of the vehicle and the preset height limit of the three-dimensional road space.

17. A traffic dispatching device, characterized in that, include: The lane reconstruction module is used to reconstruct lanes based on real-time traffic conditions to obtain the reconstructed lane topology. The formation module is used to form the vehicles based on their destination and / or travel route to obtain at least one fish swarm. The determination module is used to determine the target fish swarm corresponding to the reconstructed lane topology based on the reconstructed lane topology and the at least one fish swarm; The scheduling module is used to perform traffic scheduling for the target fish swarm based on the reconstructed lane topology.

18. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the traffic scheduling method as described in any one of claims 1 to 16.

19. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the traffic scheduling method as described in any one of claims 1 to 16.

20. A computer program product, characterized in that, The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program including programs or instructions that, when executed, implement the steps of the traffic scheduling method as described in any one of claims 1 to 16.

Images