A lane matching method in a split-flow scene based on automatic driving real-time mapping

Abstract

The application provides a lane matching method in a split-flow scene based on automatic driving real-time mapping, which comprises the following steps: step 1, calculating the lateral comfort value comfort(lane) of the lane of the ego vehicle and the subsequent lane; step 2, calculating the lateral navigation efficiency value efficiency(lane) of the lane of the ego vehicle and the subsequent lane; step 3, calculating the traffic congestion degree value congestion(lane) of the lane of the ego vehicle and the subsequent lane; step 4, according to the cost equation cost (lane) =δ com *comfort(lane)+δ e *efficiency(lane)+δ con *congestion(lane), calculating the cost value of each subsequent lane after being connected with its previous lane after splitting; and step 5, selecting the subsequent lane with the minimum cost value as the matched subsequent lane. The technical problem of selecting the optimal subsequent lane in the real-time mapping road splitting situation in the prior art is solved.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving. Background Technology

[0002] Currently, autonomous driving solutions relying on high-precision maps dominate the industry, with the accuracy and efficiency of these maps directly impacting the performance of autonomous vehicles. However, high-precision maps suffer from limitations such as high acquisition costs, difficulty in maintaining their freshness, and inability to be expanded to smaller cities. To overcome the bottlenecks imposed by high-precision maps on autonomous vehicles, a large amount of research has emerged on "map-free" solutions, such as real-time mapping. A challenging task in real-time mapping is constructing road topology in real time, especially in scenarios involving road diversion. Road diversion scenarios include intersection diversion, elevated and highway ramps, and roundabout diversion. In these scenarios, improper subsequent lane selection can lead to unnecessary steering radii, driving into lanes with long queues, entering the wrong turning lane at an intersection, or taking the wrong ramp, thus affecting the autonomous driving experience. Summary of the Invention

[0003] In view of this, the present invention provides a lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving, so as to solve the technical problem of selecting the optimal successor lane in the case of real-time mapping and traffic diversion in the prior art.

[0004] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving. The method includes: Step 1, extracting lane path-related information in real time through vehicle sensors, and calculating the root mean square error (Angle) of the angles of path points on the current lane centerline and the subsequent lane centerline based on the lane path-related information. MSE The steering angle steer, which is the current vehicle's target distance from the next lane, is substituted into the preset function. In the calculation, the lateral comfort value comfort(lane) between the driving lane and the following lane is calculated, where steer standard These are standard values ​​for the comfort turning angle of different vehicle models. The road-related information includes the lane waypoint heading angle, the number of waypoints, and the lane end and start points. Step 2: Obtain lane duration and distance information through the vehicle navigation system in the smart cockpit, and then substitute the lane duration and distance information into a preset function. In this process, the lateral navigation efficiency value (efficiency(lane)) between the vehicle lane and subsequent lanes is calculated, where ETA... lane It is the expected arrival time for selecting this lane, ETA. min It is the shortest expected arrival time at present, δeta It is the expected arrival time weight, Distance lane This selects the distance to reach the lane. min It is the shortest distance to reach, δ dis This is the nearest arrival distance weight. Step 3: Obtain road-related information through the vehicle's navigation system in the smart cockpit, and then substitute this road-related information into the function. In the calculation, the congestion value (congestion(lane)) is used to determine the degree of traffic congestion between the vehicle lane and the subsequent lanes, where μ lane This represents the traffic flow speed on that road, μ. f Indicates the road speed limit, N lane L represents the number of vehicles currently on the road. lane K represents the length of the road. j Representing the blockage density; Step 4, based on the cost equation. (lane) =δ com *comfort(lane)+δ e *efficiency(lane)+δ con *congestion(lane) calculates the cost of connecting each subsequent lane to its preceding lane after the traffic split. (lane) , where δ com It is the weight of the lateral comfort cost, δ e The weight representing navigation efficiency, δ con The weights are based on the degree of traffic congestion; Step 5, select the cost. (lane) The lane with the smallest value is selected as the matching lane.

[0005] Furthermore, the formula for calculating the root mean square error of the angles of the path points on the current lane centerline and the subsequent lane centerline is as follows: Among them, Angle mean It is the average heading angle of the path points on the current lane centerline and the subsequent lane centerline. ego_end It is the end of the bicycle lane, Lane ego_local It is the shortest path point in the vehicle lane, Angle point The heading angle of the waypoint, Lane ego_pointnum Lane represents the number of waypoints in a vehicle lane. childrenpointnum This indicates the number of waypoints for the subsequent lane.

[0006] Furthermore, the feature is that the calculation function for the average heading angle of the path points on the current lane centerline and the subsequent lane centerline is: Among them, Lane children_endIt is the end of the following lane, Lane children_start It is the starting point of the subsequent lane.

[0007] Furthermore, the calculation function for the steering angle of the current vehicle's pre-aimed steering at a preset distance in the subsequent lane is as follows:

[0008] Where wheelbase_len is the vehicle's wheelbase, Angle children_lookaheadpoint This indicates a path point that is aimed forward at a preset distance from the vehicle's current position.

[0009] Furthermore, step 4 also includes normalizing the lateral comfort value, lateral navigation efficiency value, and traffic congestion value based on the current lane data.

[0010] Furthermore, the indices affecting the lateral comfort cost include curvature magnitude and curvature continuity.

[0011] Furthermore, step 3 also includes the maximum number of vehicles per unit length of road when traffic flow comes to a complete standstill.

[0012] Furthermore, the lane path-related information includes abrupt changes or distorted values.

[0013] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving. In diversion scenarios, this invention comprehensively considers multiple factors such as lateral comfort, navigation efficiency, and traffic congestion to match the optimal successor lane. It can provide autonomous vehicles with a smooth, efficient, and comfortable guidance route in real time without relying on high-precision maps. This solves the road topology matching problem in real-time mapping road diversion scenarios in existing technologies, as well as the problem of comprehensively considering multiple factors such as vehicle lateral comfort, navigation efficiency, and traffic congestion in diversion scenarios. Attached Figure Description

[0014] Figure 1 This is a flowchart of a lane matching method for a traffic diversion scenario based on real-time mapping of autonomous driving, provided by the present invention.

[0015] Figure 2 This is a flowchart of another lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving provided by the present invention;

[0016] Figure 3 This is a schematic diagram of lane diversion provided by the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1

[0019] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving, such as... Figure 1 As shown, the method includes the following steps.

[0020] Step 1: Extract lane path information in real time using vehicle sensors. Based on this information, calculate the root mean square error (Angle) of the angles of the path points on the current lane centerline and the subsequent lane centerlines. MSE The steering angle steer, which is the current vehicle's target distance from the next lane, is substituted into the preset function. In the calculation, the lateral comfort value comfort(lane) between the driving lane and the following lane is calculated, where steer standard These are the standard values ​​for the comfort turning angle of different vehicle models. The road-related information includes the lane waypoint heading angle, the number of waypoints, and the lane end point and start point.

[0021] Step 2: Obtain lane duration and distance information through the vehicle navigation system in the smart cockpit, and then substitute the lane duration and distance information into a preset function. In this process, the lateral navigation efficiency value (efficiency(lane)) between the vehicle lane and subsequent lanes is calculated, where ETA... lane It is the expected arrival time for selecting this lane, ETA. min It is the shortest expected arrival time at present, δ eta It is the expected arrival time weight, Distance lane This selects the distance to reach the lane. min It is the shortest distance to reach, δ dis It is the nearest arrival distance weight.

[0022] Step 3: Obtain road-related information through the vehicle navigation system in the smart cockpit, and then substitute the road-related information into the function. In the calculation, the congestion value (congestion(lane)) is used to determine the degree of traffic congestion between the vehicle lane and the subsequent lanes, where μ lane This represents the traffic flow speed on that road, μ. f Indicates the road speed limit, N laneL represents the number of vehicles currently on the road. lane K represents the length of the road. j Indicates blockage density;

[0023] After receiving the perceived lane information, the vehicle begins real-time mapping for the diversion scenario. Confirming lane topology matching within the diversion scenario is itself part of the mapping process, but this mapping is partial, not full-scene mapping. Real-time mapping for autonomous driving is a dynamic map building technology that fuses data from onboard sensors such as LiDAR, cameras, and radar to capture and update detailed information about the vehicle's surroundings in real time, including road topology and traffic signs, thus constructing lane-level maps and achieving near-high-precision map-based lane-level navigation capabilities. Figure 3 As shown, a traffic splitting scenario refers to a situation where a single lane on a road splits into two or more lanes. Lane topology matching mainly realizes the connection of topological relationships (preceding, succeeding, parallel, etc.) between different lanes. In a traffic splitting scenario, the lane before the split has multiple succeeding lanes. At the same time, due to the lack of guidance from high-precision maps, how to select the optimal succeeding lane during the real-time mapping process of autonomous vehicles becomes a challenge.

[0024] Step 4, according to the cost equation (lane) =δ com *comfort(lane)+δ e *efficiency(lane)+δ con *congestion(lane) calculates the cost of connecting each subsequent lane to its preceding lane after the traffic split. (lane) , where δ com It is the weight of the lateral comfort cost, δ e The weight representing navigation efficiency, δ con It is the weight of the degree of traffic congestion;

[0025] Step 5, select cost (lane) The lane with the smallest value is selected as the matching lane.

[0026] like Figure 2 As shown, step 4 further includes normalizing the lateral comfort value, lateral navigation efficiency value, and traffic congestion value based on the current lane data.

[0027] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving. In diversion scenarios, this invention comprehensively considers multiple factors such as lateral comfort, navigation efficiency, and traffic congestion to match the optimal successor lane. It can provide autonomous vehicles with a smooth, efficient, and comfortable guidance route in real time without relying on high-precision maps. This solves the road topology matching problem in real-time mapping road diversion scenarios in existing technologies, as well as the problem of comprehensively considering multiple factors such as vehicle lateral comfort, navigation efficiency, and traffic congestion in diversion scenarios.

[0028] Example 2

[0029] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving, such as... Figure 1 As shown, the method includes the following steps.

[0030] Step 1: Extract lane path information in real time using vehicle sensors. Based on this information, calculate the root mean square error (Angle) of the angles of the path points on the current lane centerline and the subsequent lane centerlines. MSE The steering angle steer, which is the current vehicle's target distance from the next lane, is substituted into the preset function. In the calculation, the lateral comfort value comfort(lane) between the driving lane and the following lane is calculated, where steer standard These are the standard values ​​for the comfort turning angle of different vehicle models. The road-related information includes the lane waypoint heading angle, the number of waypoints, and the lane end point and start point.

[0031] The lateral comfort cost, comfort(lane), of the autonomous vehicle's lane and subsequent lanes is calculated. The influence of the reference path on the lateral comfort of autonomous vehicles is mainly reflected in two indicators: curvature magnitude and curvature continuity. However, the lane line data acquired in real time due to occlusion, errors, etc., may have abrupt changes or distorted values. Therefore, this invention proposes a new method for calculating the lateral comfort cost of the autonomous vehicle's lane and subsequent lanes, as shown in the following formula.

[0032]

[0033]

[0034] Among them, Angle mean It calculates the average heading angle of the path points on the current lane centerline and the subsequent lane centerline. MSE This calculates the mean square error of the angles of path points on the current lane centerline and the centerlines of the next lane. The next lane... children All waypoints will be included in the calculation, but the current lane... egoOnly calculate the path points forward from the vehicle's current position. ego_pointnum Lane represents the number of waypoints in a vehicle lane. childrenpointnum This indicates the number of waypoints in subsequent lanes. The mean squared error of the waypoint angles (Angle) MSE It can reflect the curvature changes of connecting lanes and reduce the impact of curvature abrupt changes or distortions. However, using only the angle mean square error to reflect the lateral comfort cost between the vehicle's lane and the following lane is insufficient; the vehicle's current position and posture need to be considered. The final formula for calculating the comfort cost is shown below.

[0035]

[0036] Where steer is the calculated steering angle of the current vehicle looking ahead at a certain distance from look_ahead_dist in the following lane. standard These are the standard values ​​for the comfort steering angle of different vehicle models. `wheelbase_len` is the vehicle's wheelbase. children_lookaheadpoint This indicates a path point that is projected forward a distance of look_ahead_dist from the vehicle's current position.

[0037] Step 2: Obtain lane duration and distance information through the vehicle navigation system in the smart cockpit, and then substitute the lane duration and distance information into a preset function. In this process, the lateral navigation efficiency value (efficiency(lane)) between the vehicle lane and subsequent lanes is calculated, where ETA... lane It is the expected arrival time for selecting this lane, ETA. min It is the shortest expected arrival time at present, δ eta It is the expected arrival time weight, Distance lane This selects the distance to reach the lane. min It is the shortest distance to reach, δ dis It is the nearest arrival distance weight.

[0038] Step 3: Obtain road-related information through the vehicle navigation system in the smart cockpit, and then substitute the road-related information into the function. In the calculation, the congestion value (congestion(lane)) is used to determine the degree of traffic congestion between the vehicle lane and the subsequent lanes, where μ lane This represents the traffic flow speed on that road, μ. f Indicates the road speed limit, N lane L represents the number of vehicles currently on the road. lane K represents the length of the road. j Indicates blockage density;

[0039] Step 4, according to the cost equation (lane) =δ com *comfort(lane)+δ e *

[0040] efficiency(lane) + δ con *congestion(lane) calculates the cost of connecting each subsequent lane to its preceding lane after the traffic split. (lane) , where δ com It is the weight of the lateral comfort cost, δ e The weight representing navigation efficiency, δ con It is the weight of the degree of traffic congestion;

[0041] Step 5, select cost (lane) The lane with the smallest value is selected as the matching lane.

[0042] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving. In diversion scenarios, this invention comprehensively considers multiple factors such as lateral comfort, navigation efficiency, and traffic congestion to match the optimal successor lane. It can provide autonomous vehicles with a smooth, efficient, and comfortable guidance route in real time without relying on high-precision maps. This solves the road topology matching problem in real-time mapping road diversion scenarios in existing technologies, as well as the problem of comprehensively considering multiple factors such as vehicle lateral comfort, navigation efficiency, and traffic congestion in diversion scenarios.

[0043] Example 3

[0044] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving, such as... Figure 1 As shown, the method includes the following steps.

[0045] Step 1: Extract lane path information in real time using vehicle sensors. Based on this information, calculate the root mean square error (Angle) of the angles of the path points on the current lane centerline and the subsequent lane centerlines. MSE The steering angle steer, which is the current vehicle's target distance from the next lane, is substituted into the preset function. In the calculation, the lateral comfort value comfort(lane) between the driving lane and the following lane is calculated, where steer standard These are the standard values ​​for the comfort turning angle of different vehicle models. The road-related information includes the lane waypoint heading angle, the number of waypoints, and the lane end point and start point.

[0046] Step 2: Obtain lane duration and distance information through the vehicle navigation system in the smart cockpit, and then substitute the lane duration and distance information into a preset function. In this process, the lateral navigation efficiency value (efficiency(lane)) between the vehicle lane and subsequent lanes is calculated, where ETA... lane It is the expected arrival time for selecting this lane, ETA. min It is the shortest expected arrival time at present, δ eta It is the expected arrival time weight, Distance lane This selects the distance to reach the lane. min It is the shortest distance to reach, δ dis It is the nearest arrival distance weight;

[0047] The cost of navigation efficiency mainly includes the expected arrival time (ETA) to the destination after selecting the subsequent lane and the increase in travel distance (the difference between the travel distance of the current optimal route). Among them, ETA lane It is the expected arrival time for selecting this lane, ETA. min It is the shortest expected arrival time at present, δ eta It is the weight of the expected arrival time. Distance lane This selects the distance to reach the lane. min It is the shortest distance to reach, δ dis It is the shortest distance to reach.

[0048] Step 3: Obtain road-related information through the vehicle navigation system in the smart cockpit, and then substitute the road-related information into the function. In the calculation, the congestion value (congestion(lane)) is used to determine the degree of traffic congestion between the vehicle lane and the subsequent lanes, where μ lane This represents the traffic flow speed on that road, μ. f Indicates the road speed limit, N lane L represents the number of vehicles currently on the road. lane K represents the length of the road. j Indicates blockage density;

[0049] Step 4, according to the cost equation (lane) =δ com *comfort(lane)+δ e *

[0050] efficiency(lane) + δ con *congestion(lane) calculates the cost of connecting each subsequent lane to its preceding lane after the traffic split. (lane) , where δ comIt is the weight of the lateral comfort cost, δ e The weight representing navigation efficiency, δ con It is the weight of the degree of traffic congestion;

[0051] Step 5, select cost (lane) The lane with the smallest value is selected as the matching lane.

[0052] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving. In diversion scenarios, this invention comprehensively considers multiple factors such as lateral comfort, navigation efficiency, and traffic congestion to match the optimal successor lane. It can provide autonomous vehicles with a smooth, efficient, and comfortable guidance route in real time without relying on high-precision maps. This solves the road topology matching problem in real-time mapping road diversion scenarios in existing technologies, as well as the problem of comprehensively considering multiple factors such as vehicle lateral comfort, navigation efficiency, and traffic congestion in diversion scenarios.

[0053] Example 4

[0054] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving, such as... Figure 1 As shown, the method includes the following steps.

[0055] Step 1: Extract lane path information in real time using vehicle sensors. Based on this information, calculate the root mean square error (Angle) of the angles of the path points on the current lane centerline and the subsequent lane centerlines. MSE The steering angle steer, which is the current vehicle's target distance from the next lane, is substituted into the preset function. In the calculation, the lateral comfort value comfort(lane) between the driving lane and the following lane is calculated, where steer standard These are the standard values ​​for the comfort turning angle of different vehicle models. The road-related information includes the lane waypoint heading angle, the number of waypoints, and the lane end point and start point.

[0056] Step 2: Obtain lane duration and distance information through the vehicle navigation system in the smart cockpit, and then substitute the lane duration and distance information into a preset function. In this process, the lateral navigation efficiency value (efficiency(lane)) between the vehicle lane and subsequent lanes is calculated, where ETA... lane It is the expected arrival time for selecting this lane, ETA. min It is the shortest expected arrival time at present, δ eta It is the expected arrival time weight, Distance lane This selects the distance to reach the lane. minIt is the shortest distance to reach, δ dis It is the nearest arrival distance weight;

[0057] Step 3: Obtain road-related information through the vehicle navigation system in the smart cockpit, and then substitute the road-related information into the function. In the calculation, the congestion value (congestion(lane)) is used to determine the degree of traffic congestion between the vehicle lane and the subsequent lanes, where μ lane This represents the traffic flow speed on that road, μ. f Indicates the road speed limit, N lane L represents the number of vehicles currently on the road. lane K represents the length of the road. j Indicates blockage density;

[0058] Traffic congestion not only increases travel time but also raises the difficulty and collision risk for autonomous vehicles. Therefore, the degree of traffic congestion is considered as a separate cost. This invention primarily expresses the degree of congestion through traffic flow speed, calculated using the following formula: Where, μ lane This represents the traffic flow speed on that road, μ. f Indicates the road speed limit, N lane L represents the number of vehicles currently on the road. lane K represents the length of the road. j This represents congestion density, which is the maximum number of vehicles per unit length of road when traffic flow is completely stopped; that is, the density when vehicles cannot move.

[0059] Step 4, according to the cost equation (lane) =δ com *comfort(lane)+δ e *

[0060] efficiency(lane) + δ con *congestion(lane) calculates the cost of connecting each subsequent lane to its preceding lane after the traffic split. (lane) , where δ com It is the weight of the lateral comfort cost, δ e The weight representing navigation efficiency, δ con It is the weight of the degree of traffic congestion;

[0061] Step 5, select cost (lane) The lane with the smallest value is selected as the matching lane.

[0062] This invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving. In diversion scenarios, this invention comprehensively considers multiple factors such as lateral comfort, navigation efficiency, and traffic congestion to match the optimal successor lane. It can provide autonomous vehicles with a smooth, efficient, and comfortable guidance route in real time without relying on high-precision maps. This solves the road topology matching problem in real-time mapping road diversion scenarios in existing technologies, as well as the problem of comprehensively considering multiple factors such as vehicle lateral comfort, navigation efficiency, and traffic congestion in diversion scenarios.

[0063] In summary, this invention provides a lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving. This method comprehensively considers multiple factors such as vehicle lateral comfort, navigation efficiency, and traffic congestion levels when selecting subsequent lanes, thus optimizing vehicle comfort and traffic efficiency in a real-time mapping environment for autonomous driving. Furthermore, this method evaluates vehicle lateral comfort in the context of reference route stitching, which can reflect changes in curvature between connecting lanes while reducing the impact of sudden changes or distortions in road curvature caused by occlusion or errors.

[0064] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A lane matching method for traffic diversion scenarios based on real-time mapping for autonomous driving, characterized in that, The method includes: Step 1: Extract lane path information in real time using vehicle sensors. Based on this information, calculate the root mean square error (Angle) of the angles of the path points on the current lane centerline and the subsequent lane centerlines. MSE The steering angle steer, which is the current vehicle's target distance from the next lane, is substituted into the preset function. In the calculation, the lateral comfort value comfort(lane) between the driving lane and the following lane is calculated, where steer standard These are the standard values ​​for the comfort turning angle of different vehicle models. The lane path-related information includes the heading angle of the lane path point, the number of path points, and the lane end point and start point. Step 2: Obtain lane duration and distance information through the vehicle navigation system in the smart cockpit, and then substitute the lane duration and distance information into a preset function. In this process, the lateral navigation efficiency value (efficiency(lane)) between the vehicle lane and subsequent lanes is calculated, where ETA... lane It is the expected arrival time for selecting this lane, ETA. min It is the shortest expected arrival time at present, δ eta It is the expected arrival time weight, Distance lane This selects the distance to reach the lane. min It is the shortest distance to reach, δ dis It is the nearest arrival distance weight; Step 3: Obtain road-related information through the vehicle navigation system in the smart cockpit, and then substitute the road-related information into the function. In the calculation, the congestion value (congestion(lane)) is used to determine the degree of traffic congestion between the vehicle lane and the subsequent lanes, where μ lane This represents the traffic flow speed on that road, μ. f Indicates the road speed limit, N lane L represents the number of vehicles currently on the road. lane K represents the length of the road. j Indicates blockage density; Step 4, according to the cost equation (lane) =δ com *comfort(lane)+δ e *efficiency(lane)+δ con *congestion(lane) calculates the cost of connecting each subsequent lane to its preceding lane after the traffic split. (lane) , where δ com It is the weight of the lateral comfort cost, δ e The weight representing navigation efficiency, δ con It is the weight of the degree of traffic congestion; Step 5, select cost (lane) The lane with the smallest value is selected as the matching lane.

2. The lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving according to claim 1, characterized in that, The formula for calculating the root mean square error of the angles of the path points on the current lane centerline and the subsequent lane centerline is: ， Among them, Angle mean It is the average heading angle of the path points on the current lane centerline and the subsequent lane centerline. ego_end It is the end of the bicycle lane, Lane ego_local It is the shortest path point in the vehicle lane, Angle point The heading angle of the waypoint, Lane ego_pointnum Lane represents the number of waypoints in a vehicle lane. childrenpointnum Lane indicates the number of waypoints in the subsequent lane. children_end It is the end of the following lane, Lane children_start It is the starting point of the subsequent lane.

3. The lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving according to claim 2, characterized in that, The formula for calculating the average heading angle of the path points on the current lane centerline and the subsequent lane centerline is: 。 4. The lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving according to claim 1, characterized in that, The formula for calculating the steering angle of the current vehicle at a preset distance from the next lane is: Where wheelbase_len is the vehicle's wheelbase, Angle children_lookaheadpoint This indicates a path point that is aimed forward at a preset distance from the vehicle's current position.

5. The lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving according to claim 1, characterized in that, Step 4 further includes normalizing the lateral comfort value, lateral navigation efficiency value, and traffic congestion value based on the current lane data.

6. The lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving according to claim 1, characterized in that, The indices affecting the cost of lateral comfort include curvature magnitude and curvature continuity.

7. The lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving according to claim 1, characterized in that, Step 3 also includes the maximum number of vehicles per unit length of road when traffic flow comes to a complete standstill.

8. The lane matching method for traffic diversion scenarios based on real-time mapping of autonomous driving according to claim 1, characterized in that, The lane path-related information includes abrupt changes or abnormal values.

Images