Method, system, device, storage medium and vehicle for performing lane change control in autonomous driving

Abstract

The application discloses a method for performing lane changing control in automatic driving, which comprises the following steps: obtaining current self-vehicle position information if a lane changing cancellation command is received during lane changing execution; judging a current lane changing cancellation level according to the self-vehicle position information; obtaining a target state and a target pose corresponding to the self-vehicle and the current lane changing cancellation level according to the current lane changing cancellation level, jumping a vehicle decision state machine to the target state, and controlling the self-vehicle to reach the target pose. The application further discloses a corresponding system, device, storage medium and vehicle. The application can improve the lane changing success rate, comfort and safety of the automatic driving system.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation technology, specifically to a method, system, device, storage medium, and vehicle for lane change control in autonomous driving. Background Technology

[0002] In current mainstream autonomous driving systems, decision-making, planning, and control generally adopt a line-following-based solution. This type of solution uses the navigation route provided by the user-defined navigation system, downstream decision-making, planning, and control to perceive or obtain high-precision maps in real time, acquire lane edge lines for the current and a certain distance in the future, and generate a virtual reference line for each lane on the navigation route based on these lane edge lines. Based on these reference lines, behavior decisions, path and trajectory planning, and control are realized.

[0003] In autonomous driving systems, everyday human driving behavior is abstracted and summarized into a series of decisions and actions such as lane changing, obstacle avoidance, and lane keeping. The decision-making and action execution process of autonomous driving, based on anthropomorphic thinking, can be subdivided into five states: propose, prepare, execute, finish, and cancel, with lane keeping as the default state. Transitions between these states can be controlled using a hierarchical finite state machine by setting thresholds and rules, thus preventing frequent state switching. The entry and exit processes of individual states can be controlled by binding corresponding logical judgment modules. This method allows for a better simulation of the decision-making and behavioral processes of human drivers.

[0004] Safety is paramount in lane-changing decisions and actions. To ensure safety, if a lane-changing condition is not met during the execution of a lane change, the system will directly cancel the lane change. In existing autonomous driving systems, the general approach for handling the vehicle's behavior after a lane-change cancellation is to directly return the vehicle to the center of the original lane, allowing it to continue driving based on the virtual reference line of the original lane. After a cooling-off period, the system will attempt to initiate another lane-change attempt. In terms of state machine transitions, after entering the cancel state, the system first needs to return to the default lane-keeping state, and then enter the propose state to find a suitable opportunity.

[0005] However, the existing lane-changing decision-making scheme has many shortcomings:

[0006] First, in areas with high traffic density and at the end of navigation lane changes, suitable lane-changing opportunities are fleeting. Existing solutions cannot capture lane-changing opportunities in time, and significant fluctuations in the vehicle's heading angle can cause considerable panic for drivers and passengers. The aforementioned transition process is time-consuming, and during this process, suitable lane-changing opportunities may arise multiple times. A mechanical transition process can cause the vehicle to miss numerous lane-changing opportunities, or even get stuck at the end of the navigation lane change, ultimately requiring manual intervention or system exit.

[0007] Secondly, existing technical solutions do not consider the extent of the lane change action when the vehicle enters the lane change cancellation state. If the vehicle has already crossed the lane edge line, or even if a significant portion of the vehicle has entered the target lane, the current technical solution requires the vehicle to return to the center of the original lane, and the required yaw angle correction is substantial. Externally, this manifests as irregular swerving in high-traffic areas, making it difficult to adjust to a suitable position for the next lane change opportunity. Internally, it results in significant steering wheel back turn, leading to sudden front-end swaying and potentially causing panic-induced takeover. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to provide a method, system, device, storage medium and vehicle for lane change control in autonomous driving, which can improve the success rate, comfort and safety of lane change in autonomous driving systems.

[0009] To address the aforementioned technical problems, as one aspect of the present invention, a method for lane change control in autonomous driving is provided, which includes at least the following steps:

[0010] During the lane change process, if a lane change cancellation command is received, the current vehicle position information is obtained. The vehicle position information includes at least: the lateral offset δl of the vehicle relative to the original lane virtual reference line and the angular value δθ between the vehicle and the original lane virtual reference line.

[0011] Based on the vehicle's location information, determine the current lane change cancellation level;

[0012] Based on the current lane change cancellation level, the target state and target pose of the vehicle corresponding to the current lane change cancellation level are obtained, the vehicle decision state machine is switched to the target state, and the vehicle is controlled to reach the target pose; wherein each lane change cancellation level corresponds to a target state, and the target state includes: lane keeping, lane change proposal, lane change preparation, or target lane keeping.

[0013] The step of determining the current lane change cancellation level based on the lateral offset δl and the included angle in radians δθ further includes:

[0014] When δl is less than or equal to the first offset threshold and δθ is less than or equal to the first radian value threshold, the current lane change cancellation level is determined to be the first level.

[0015] When δl is less than or equal to the second offset threshold and δθ is greater than the first radian value threshold, the current lane change cancellation level is determined to be the second level.

[0016] When δθ is greater than the second offset threshold and δθ is less than or equal to the first radian value threshold, the current lane change cancellation level is determined to be the third level.

[0017] When δl is greater than the second offset threshold and δθ is greater than the first radian value threshold, the current lane change cancellation level is determined to be the fourth level.

[0018] Wherein, the second offset threshold is greater than the first offset threshold.

[0019] The method further includes: obtaining the target state and target pose of the vehicle corresponding to the current lane change cancellation level based on the current lane change cancellation level; transitioning the vehicle decision state machine to the target state; and controlling the vehicle to reach the target pose.

[0020] When the current lane change cancellation level is Level 1, the target state is lane keeping. The state is switched to lane keeping state, and the target pose is: δl=0 and δθ=0.

[0021] When the current lane change cancellation level is the second level, the target state is lane change initiated. The state will be switched to the lane change initiated state, and the vehicle will no longer enter the target lane. The vehicle will straighten its front end in the current lane. The target position is: δθ=0.

[0022] When the current lane change cancellation level is level 3, the target state is lane change preparation. The state is switched to lane change preparation state, and the vehicle continues to enter the target lane. The vehicle is straightened in the target lane. The target position is: δl is greater than half the width of the lane and δθ = 0.

[0023] When the current lane change cancellation level is level four, the target state is target lane keeping. The state is switched to target lane keeping, and the current pose is maintained. When a suitable space appears, the lane moves towards the target lane virtual reference line. The target pose is: δθ=0.

[0024] This further includes:

[0025] After the vehicle jumps into the target state, the judgment logic corresponding to the target state is triggered for subsequent control.

[0026] The process of the vehicle transitioning to the target state and triggering the corresponding judgment logic for subsequent control further includes:

[0027] When the vehicle enters the lane keeping or target lane keeping state, the longitudinal speed decision module of the vehicle is triggered, and then the vehicle determines whether to accelerate or decelerate by judging the movement of the vehicles in front and behind in the lane.

[0028] When a vehicle enters the lane change proposal state, it determines whether the driving environment continuously meets the necessity and legality of changing lanes. If it continues to meet the requirements, it enters the next lane change preparation state; otherwise, it returns to the lane holding state.

[0029] When the vehicle enters the lane change preparation state, the decision-making system determines in each frame whether there is a suitable opportunity and duration for lane change. If the conditions are met, it enters the lane change execution state and performs the lane change operation. After the vehicle has approached the virtual reference line of the target lane, it enters the lane change completion state and then jumps to the lane keeping state.

[0030] Accordingly, another aspect of the present invention also provides a system for lane change control in autonomous driving, comprising at least:

[0031] The lane change position information acquisition unit is used to acquire the current position information of the vehicle when a lane change cancellation command is received during the lane change execution process. The vehicle position information includes at least: the lateral offset δl of the vehicle relative to the original lane virtual reference line and the angular value δθ between the vehicle and the original lane virtual reference line.

[0032] The cancellation level determination unit is used to determine the current lane change cancellation level based on the vehicle's position information;

[0033] The target state switching unit is used to obtain the target state and target pose of the vehicle corresponding to the current lane change cancellation level according to the current lane change cancellation level, switch the vehicle decision state machine to the target state, and control the vehicle to reach the target pose; wherein each lane change cancellation level corresponds to a target state, and the target state includes: lane keeping, lane change proposal, lane change preparation, or target lane keeping.

[0034] The cancellation level determination unit further includes:

[0035] The first level determination unit is used to determine the current lane change cancellation level as the first level when the δl is less than or equal to the first offset threshold and the δθ is less than or equal to the first radian value threshold.

[0036] The second level determination unit is used to determine the current lane change cancellation level as the second level when the δl is less than or equal to the second offset threshold and the δθ is greater than the first radian value threshold.

[0037] The third-level determination unit is used to determine the current lane change cancellation level as the third level when the δl is greater than the second offset threshold and the δθ is less than or equal to the first radian value threshold.

[0038] The fourth level determination unit is used to determine the current lane change cancellation level as the fourth level when the δl is greater than the second offset threshold and the δθ is greater than the first radian value threshold.

[0039] Wherein, the second offset threshold is greater than the first offset threshold.

[0040] The target state switching unit further includes:

[0041] The first switching processing unit is used to switch the state to the lane keeping state when the current lane change cancellation level is the first level and the target state is lane keeping. The target pose is: δl=0 and δθ=0.

[0042] The second switching processing unit is used to switch the state to the lane change initiated state when the current lane change cancellation level is the second level and the target state is lane change initiated. It will not continue to enter the target lane and will straighten the front of the vehicle in the current lane. The target position is: δθ=0.

[0043] The third switching processing unit is used to switch the state to the lane change preparation state when the current lane change cancellation level is the third level and the target state is lane change preparation. It then continues to enter the target lane, straightens the front of the vehicle in the target lane, and the target pose is: δl is greater than half the width of the lane and δθ = 0.

[0044] The third switching processing unit is used to switch the state to target lane keeping when the current lane change cancellation level is level four and the target state is target lane keeping. It maintains the current pose and when a suitable space appears, it moves towards the virtual reference line of the target lane. The target pose is: δθ=0.

[0045] This further includes:

[0046] The switching post-processing unit is used to trigger the judgment logic corresponding to the target state for subsequent control after the vehicle jumps into the target state;

[0047] The post-switching processing unit further includes:

[0048] The lane keeping processing unit is used to trigger the longitudinal speed decision module of the vehicle when the vehicle enters the lane keeping or target lane keeping state, and then decide whether to accelerate or decelerate by judging the movement of the vehicles in front and behind in the lane.

[0049] The lane change proposal processing unit is used to determine whether the driving environment continuously meets the necessity and legality of lane changing when the vehicle enters the lane change proposal state. If it continues to meet the requirements, it enters the next lane change preparation state; otherwise, it returns to the lane keeping state.

[0050] The lane change preparation processing unit is used to determine whether there is a suitable opportunity and duration for lane change in each frame when the vehicle enters the lane change preparation state. If the conditions are met, the system enters the lane change execution state and performs the lane change operation. After the vehicle has approached the virtual reference line of the target lane, it enters the lane change completion state and then jumps to the lane keeping state.

[0051] In another aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0052] In another aspect, the present invention provides a computing device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the steps of the method described above.

[0053] In another aspect, the present invention provides a vehicle having at least a lane change decision device thereon, wherein the lane change decision device is equipped with the system described above.

[0054] Implementing the embodiments of the present invention has the following beneficial effects:

[0055] This invention provides a method, system, device, storage medium, and vehicle for lane change control in autonomous driving. In an embodiment of this invention, during the execution of a lane change, when the lane change state of the vehicle changes from lane change execution to lane change cancellation, the current position information of the vehicle is first obtained: the angle between the vehicle and the original lane virtual reference line and the lateral offset of the vehicle relative to the original lane virtual reference line, to determine the degree of execution of the vehicle's lane change behavior, determine the current lane change cancellation level, and intelligently decide on one of the three states of lane keeping, lane change proposal, or lane change preparation based on the lane change cancellation level. Then, based on the judgment rules in that state, the normal transition process of the state machine is executed.

[0056] In this embodiment, based on different cancellation levels, the state machine will intelligently jump to different states, thereby triggering the corresponding judgment logic in advance, making the decision-making of the autonomous driving system more flexible.

[0057] In this embodiment, when the state machine switches to the lane change proposal or lane change preparation state, compared with the prior art, its lane change-related judgment logic will be triggered in advance, so that the autonomous driving system will not miss the fleeting opportunity and space for lane change.

[0058] In this embodiment, when the state machine switches to the lane change preparation state, the vehicle will not return to the center of the original lane, but will straighten its front end on the spot and directly look for the next lane change space. By adopting this intelligent transition method, the mechanical state transition in the original solution can be avoided, thereby triggering the judgment logic modules such as finding lane change space as soon as possible, ensuring that the autonomous driving system can effectively seize the fleeting lane change opportunity.

[0059] Furthermore, implementing this embodiment of the invention can reduce the degree of steering wheel return when lane change is cancelled, lessen the fluctuation of the vehicle's heading angle when the cancellation occurs, and improve driving comfort during lane changes. It also reduces the control tracking pressure on the downstream control unit in such scenarios, making the lane change decisions and behaviors of the autonomous driving system more decisive, efficient, smooth, and comfortable, and more in line with human driving habits under high traffic density. Attached Figure Description

[0060] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, obtaining other drawings based on these drawings without creative effort still falls within the scope of the present invention.

[0061] Figure 1 This is a schematic diagram of the main flow of an embodiment of a method for lane change control in autonomous driving provided by the present invention.

[0062] Figure 2 This is a more detailed schematic diagram illustrating the application principle of the present invention;

[0063] Figure 3 This is a schematic diagram of the vehicle location information involved in the present invention;

[0064] Figure 4 This is a schematic diagram showing the correspondence between the vehicle position information corresponding to the four lane change cancellation levels involved in the present invention.

[0065] Figure 5 This is a schematic diagram of the processing procedure after switching to lane keeping mode, as per the present invention.

[0066] Figure 6 This is a schematic diagram of the processing procedure after switching to the lane change request state, as per the present invention.

[0067] Figure 7 This is a schematic diagram illustrating the processing procedure following switching to lane change preparation mode, as per the present invention.

[0068] Figure 8A schematic diagram of a system for lane change control in autonomous driving provided by the present invention;

[0069] Figure 9 for Figure 8 A schematic diagram of the structure of the cancellation level determination unit;

[0070] Figure 10 for Figure 8 A schematic diagram of the target state switching unit;

[0071] Figure 11 for Figure 8 A schematic diagram of the structure of the post-processing unit after switching. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings.

[0073] like Figure 1 The diagram shown illustrates the main flow of an embodiment of a lane change control method for autonomous driving provided by the present invention; in conjunction with... Figures 2 to 7 As shown, in this embodiment, the method for lane change control in autonomous driving includes at least the following steps:

[0074] Step S10: During the lane change process, if a lane change cancellation command is received, the current vehicle position information is obtained. The vehicle position information includes at least: the lateral offset δl of the vehicle relative to the original lane virtual reference line and the angular value δθ between the vehicle and the original lane virtual reference line.

[0075] Understandably, in an autonomous driving system, the state machine's pre-processing state for lane change decision-making is the lane change execution state. During lane change execution, the system comprehensively considers the traffic conditions in the current target lane, the predicted trajectories of surrounding dynamic objects, and the results of downstream motion planning to assess lane change safety in real time. When the lane change safety conditions are not met, the system enters the lane change cancellation state. At this point, the autonomous driving system's decision module needs to record the lateral offset δl of the vehicle relative to the original lane virtual reference line (such as the original lane centerline) and the angle δθ between the vehicle and the original lane virtual reference line. Figure 3 The diagram shows the lateral offset δl and the included angle in radians δθ during lane changing. The lateral offset δl can be the distance between the vehicle's center point O and the original lane virtual reference line.

[0076] Step S11: Based on the vehicle's location information, determine the current lane change cancellation level to ascertain the degree to which the current lane change decision is executed;

[0077] In a specific example, step S11 further includes:

[0078] When δl is less than or equal to the first offset threshold and δθ is less than or equal to the first radian value threshold, the current lane change cancellation level is determined to be the first level. In this case, both δl and δθ are very small, indicating that no lane change action has actually started at this time.

[0079] When δl is less than or equal to the second offset threshold and δθ is greater than the first radian value threshold, the current lane change cancellation level is determined to be the second level. In this case, δθ is larger and δl is smaller, indicating that the front of the car swings more, but the speed is slower and the forward distance is shorter, and the lane change action is only performed to a small extent.

[0080] When δl is greater than the second offset threshold and δθ is less than or equal to the first radian value threshold, the current lane change cancellation level is determined to be the third level. In this case, δθ is smaller and δl is larger, indicating that the vehicle speed is faster and a small steering wheel angle is enough to make the vehicle quickly deviate from the original reference line, and the lane change action is executed to a greater extent.

[0081] When δl is greater than the second offset threshold and δθ is greater than the first radian value threshold, the current lane change cancellation level is determined to be the fourth level. In this case, both δl and δθ are very large, indicating that the lane change action has been deeply executed.

[0082] Wherein, the second offset threshold is greater than the first offset threshold; in a specific example, the first offset threshold is 0.3m, the second offset threshold is 0.6m, and the first radian threshold is 0.06 degrees.

[0083] Step S12: Based on the current lane change cancellation level, obtain the target state of the vehicle corresponding to the current lane change cancellation level and the target pose that needs to be adjusted, jump the vehicle decision state machine to the target state, and control the vehicle to reach the target pose; wherein each lane change cancellation level corresponds to a target state, and the target state includes: ego lane keeping, lane change proposal, lane change ready, or target lane keeping.

[0084] In a specific example, step S12 further includes:

[0085] When the current lane change cancellation level is the first level, the target state is lane keeping. The state is switched to lane keeping state, and the tracking forward movement based on the original lane virtual reference line is executed. The target pose is: δl=0 and δθ=0.

[0086] When the current lane change cancellation level is the second level, the target state is lane change initiated. The state will be switched to the lane change initiated state, and the vehicle will no longer enter the target lane. The vehicle will straighten its front end in the current lane. The target position is: δθ=0.

[0087] When the current lane change cancellation level is level 3, the target state is lane change preparation. The state is switched to lane change preparation state, and the vehicle continues to enter the target lane. The vehicle is straightened in the target lane. The target position is: δl is greater than half the width of the lane and δθ = 0.

[0088] When the current lane change cancellation level is level four, the target state is target lane keeping. The state is switched to target lane keeping, and the current pose is maintained. When a suitable space appears, the lane moves towards the target lane virtual reference line. The target pose is: δθ=0.

[0089] In a specific example, the method further includes:

[0090] Step S13: After the vehicle transitions to the target state, the judgment logic corresponding to the target state is triggered for subsequent control. Step S13 further includes:

[0091] Step S130: When the vehicle enters the ego lane keeping or target lane keeping state, the longitudinal speed decision module of the vehicle is triggered, and then the vehicle determines whether to accelerate or decelerate by judging the movement of the vehicles in front and behind in the lane.

[0092] Specifically, when the vehicle enters ego lane keeping, it follows the path of the original lane's virtual reference line. When it enters target lane keeping, it maintains its current position, blocks the existing lane-changing space, and attempts to align itself with the target lane's virtual reference line. A diagram illustrating the state machine transitions in this scenario can be found here. Figure 5 As shown;

[0093] Step S131: When the vehicle enters the lane change proposal state, it determines whether the driving environment continuously satisfies the necessity and legality of changing lanes. If it continuously satisfies these conditions, it enters the next lane change preparation state, executes the logic for finding lane change space, and then performs the lane change operation when the opportunity arises. If the conditions are not met, it returns to the lane holding state, meaning the vehicle drives close to or along the lane edge for a period of time before returning to the center of the original lane. A state machine transition diagram for this scenario can be found in [reference needed]. Figure 6 As shown;

[0094] In specific examples, the following factors generally need to be considered when determining the necessity and legality of a lane change:

[0095] Regarding the necessity of lane changing: When changing lanes via navigation, it is necessary to consider the distance to the next intersection, whether there are lane markings, and whether the attributes of the current lane and the target lane meet the requirements for straight driving, left turn, right turn, or U-turn. For overtaking lane changes, it is necessary to consider the relative speed and distance to the vehicle in front, as well as the speed of traffic in the current lane and the speed of traffic in the adjacent lane.

[0096] The legality of lane changes needs to take into account factors such as speed limit information, lane edge information (such as dashed or solid lines), and right-of-way.

[0097] In step S132, when the vehicle enters the lane change preparation (ready) state, since the vehicle has already significantly entered the target lane, reversing into the original lane to maintain lane keeping would cause a significant sway of the front of the vehicle. Therefore, the vehicle will temporarily maintain its position on the spot, keeping its front end within the target lane and moving forward while keeping it close to the lane edge line.

[0098] The decision-making system directly triggers the logic of finding suitable lane-changing space. That is, in each frame, it determines whether there is a suitable opportunity and duration for a lane change. If the conditions are met, it means that performing a lane change is safe, and the system enters the lane-changing execution state to perform the lane change operation. Specifically, in this state, the vehicle will steer away from its original lane based on the locked lane-changing space, and accelerate or decelerate according to its relative speed with the vehicles in front and behind. This state continues until the vehicle is fully inside the target lane and close to the virtual reference line of the target lane. Understandably, since it has already occupied part of the right-of-way in the target lane, the vehicles behind will give way to the vehicle, which also increases the probability of finding suitable lane-changing space. Furthermore, the vehicle is constantly searching for suitable lane-changing space, thus increasing the success rate of the lane change.

[0099] Once the vehicle has approached the virtual reference line of the target lane, it indicates that the lane change is about to be completed, and the vehicle enters the lane change completed state. This state will not last long; when the vehicle's orientation aligns with the target lane's orientation, the lane change is completely completed, and the vehicle enters the lane-keeping state. The state machine transitions to "lane-keeping." A diagram illustrating this state machine transition can be found in [link to diagram]. Figure 7 As shown.

[0100] In specific examples, determining whether there is a suitable time to change lanes generally considers the following conditions: whether there is suitable space for changing lanes within the target lane, the gap between the vehicle in front and behind the vehicle currently relative to the vehicle, and the gap between the vehicle in front and behind the vehicle two seconds from now, calculated using relative speed and fixed time intervals; a suitable gap, where the longitudinal distance is greater than the length of the vehicle body, etc.

[0101] In a specific example, the lane change timing duration can be set to 10 frames; the "lane change preparation" timeout timeout time is set to 8 seconds; the "lane change complete" judgment condition is set to: δl is less than 0.2m and δθ is less than 0.02; where traffic density is the value obtained by dividing the number of vehicles within a distance by that distance.

[0102] Understandably, this invention combines the vehicle's own position information (i.e., pose) after lane change cancellation and subdivides the lane change cancellation state into four levels based on the degree of execution of the lane change. Based on different cancellation levels, the state machine will intelligently transition to different states, thereby triggering corresponding judgment logic in advance, making the autonomous driving system's decision-making more flexible and efficient. In particular, it will directly transition the state machine to the ready state, in which the autonomous driving decision-making module will immediately begin searching for a suitable lane change opportunity. This type of transition can avoid missing suitable lane change opportunities and improve the success rate of lane changes.

[0103] Meanwhile, in this embodiment of the invention, based on the four levels of lane change cancellation status and combined with target pose information, the vehicle's pose can be intelligently adjusted to avoid large-scale front-end swaying. When lane change cancellation occurs and the vehicle is already on the lane edge, in this embodiment, the vehicle will straighten its front end on the spot, thus giving following vehicles the impression that it has not abandoned the lane change attempt. If the vehicle has almost completely entered the target lane at this time, in this embodiment, the vehicle will maintain that pose, specifically by having its front end directly enter the fleeting lane change space, thus giving following vehicles the impression that it is still performing a lane change. Therefore, in this solution, when lane change is cancelled, the front-end swaying amplitude of the vehicle will be significantly reduced, which can improve driving comfort and also allow it to occupy a certain degree of right-of-way, thereby helping to improve the success rate of capturing lane change space and timing.

[0104] In this embodiment, by utilizing the posture of the vehicle after canceling the lane change, the state machine is intelligently selected to enter the state that follows, thereby triggering the lane change timing judgment module in a timely manner, seizing the appropriate lane change opportunity as soon as possible, and allowing the vehicle to quickly adjust to the appropriate posture, reducing the degree of vehicle sway, making the lane change decision and behavior of the autonomous driving system more decisive, efficient, smooth, and comfortable, and more in line with human driving habits under high traffic density.

[0105] like Figure 8 The diagram shown illustrates a structural schematic of an embodiment of a lane change control system for autonomous driving provided by the present invention. (In conjunction with...) Figures 9 to 11 As shown, in this embodiment, the system includes at least:

[0106] The lane change position information acquisition unit 10 is used to acquire the current position information of the vehicle when a lane change cancellation command is received during the lane change execution process. The vehicle position information includes at least: the lateral offset δl of the vehicle relative to the original lane virtual reference line and the angular value δθ between the vehicle and the original lane virtual reference line.

[0107] Cancellation level determination unit 11 is used to determine the current lane change cancellation level based on the vehicle position information;

[0108] The target state switching unit 12 is used to obtain the target state and target pose of the vehicle corresponding to the current lane change cancellation level according to the current lane change cancellation level, switch the vehicle decision state machine to the target state, and control the vehicle to reach the target pose; wherein each lane change cancellation level corresponds to a target state, and the target state includes: lane keeping, lane change proposal, lane change preparation, or target lane keeping.

[0109] The switching post-processing unit 13 is used to trigger the judgment logic corresponding to the target state for subsequent control after the vehicle jumps into the target state.

[0110] like Figure 9 As shown, in a specific example, the cancellation level determination unit 11 further includes:

[0111] The first level determination unit 110 is used to determine the current lane change cancellation level as the first level when the δl is less than or equal to the first offset threshold and the δθ is less than or equal to the first radian value threshold.

[0112] The second level determination unit 111 is used to determine the current lane change cancellation level as the second level when the δl is less than or equal to the second offset threshold and the δθ is greater than the first radian value threshold.

[0113] The third level determination unit 112 is used to determine the current lane change cancellation level as the third level when the δl is greater than the second offset threshold and the δθ is less than or equal to the first radian value threshold.

[0114] The fourth level determination unit 113 is used to determine the current lane change cancellation level as the fourth level when the δl is greater than the second offset threshold and the δθ is greater than the first radian value threshold.

[0115] Wherein, the second offset threshold is greater than the first offset threshold.

[0116] like Figure 10 As shown, in a specific example, the target state switching unit 12 further includes:

[0117] The first switching processing unit 120 is used to switch the state to the lane keeping state when the current lane change cancellation level is the first level and the target state is lane keeping. The target pose is: δl=0 and δθ=0.

[0118] The second switching processing unit 121 is used to switch the state to the lane change initiated state when the current lane change cancellation level is the second level and the target state is lane change initiated. It will not continue to enter the target lane and will straighten the front of the vehicle in the current lane. The target position is: δθ=0.

[0119] The third switching processing unit 122 is used to switch the state to the lane change preparation state when the current lane change cancellation level is the third level and the target state is lane change preparation, continue to enter the target lane, straighten the front of the vehicle in the target lane, and the target pose is: δl is greater than half the width of the lane and δθ = 0.

[0120] The third switching processing unit 123 is used to switch the state to target lane keeping when the current lane change cancellation level is the fourth level and the target state is target lane keeping. It maintains the current pose and when a suitable space appears, it attaches to the virtual reference line of the target lane. The target pose is: δθ=0.

[0121] like Figure 11 As shown, in a specific example, the post-switching processing unit 13 further includes:

[0122] The lane keeping processing unit 130 is used to trigger the longitudinal speed decision module of the vehicle when the vehicle enters the lane keeping or target lane keeping state, and then determine whether to accelerate or decelerate by judging the movement of the vehicles in front and behind in the lane.

[0123] The lane change proposal processing unit 131 is used to determine whether the driving environment continuously meets the necessity and legality of lane changing when the vehicle enters the lane change proposal state. If it continues to meet the requirements, it enters the next lane change preparation state; otherwise, it returns to the lane keeping state.

[0124] The lane change preparation processing unit 132 is used to determine whether there is a suitable opportunity and duration for lane change in each frame when the vehicle enters the lane change preparation state. If the conditions are met, the system enters the lane change execution state and performs the lane change operation. After the vehicle has approached the virtual reference line of the target lane, it enters the lane change completion state and then jumps to the lane keeping state.

[0125] In another aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned... Figures 1 to 7 The steps of the described method. For more details, please refer to and combine with the foregoing explanation. Figures 1 to 7 The description of that will not be repeated here.

[0126] In another aspect, the present invention provides a computing device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the aforementioned... Figures 1 to 7 The steps of the described method. For more details, please refer to and combine with the foregoing explanation. Figures 1 to 7 The description of that will not be repeated here.

[0127] In another aspect, the present invention provides a vehicle having at least a lane change decision device thereon, wherein the lane change decision device is configured with the aforementioned... Figures 8 to 11 The system described. For more details, please refer to and combine with the foregoing descriptions. Figures 8 to 11 The description of that will not be repeated here.

[0128] Implementing the embodiments of the present invention has the following beneficial effects:

[0129] This invention provides a method, system, device, storage medium, and vehicle for lane change control in autonomous driving. In an embodiment of this invention, during the execution of a lane change, when the lane change state of the vehicle changes from lane change execution to lane change cancellation, the current position information of the vehicle is first obtained: the angle between the vehicle and the original lane virtual reference line and the lateral offset of the vehicle relative to the original lane virtual reference line, to determine the degree of execution of the vehicle's lane change behavior, determine the current lane change cancellation level, and intelligently decide on one of the three states of lane keeping, lane change proposal, or lane change preparation based on the lane change cancellation level. Then, based on the judgment rules in that state, the normal transition process of the state machine is executed.

[0130] In this embodiment, based on different cancellation levels, the state machine will intelligently jump to different states, thereby triggering the corresponding judgment logic in advance, making the decision-making of the autonomous driving system more flexible.

[0131] In this embodiment, when the state machine switches to the lane change proposal or lane change preparation state, compared with the prior art, its lane change-related judgment logic will be triggered in advance, so that the autonomous driving system will not miss the fleeting opportunity and space for lane change.

[0132] In this embodiment, when the state machine switches to the lane change preparation state, the vehicle will not return to the center of the original lane, but will straighten its front end on the spot and directly look for the next lane change space. By adopting this intelligent transition method, the mechanical state transition in the original solution can be avoided, thereby triggering the judgment logic modules such as finding lane change space as soon as possible, ensuring that the autonomous driving system can effectively seize the fleeting lane change opportunity.

[0133] Furthermore, implementing this embodiment of the invention can reduce the degree of steering wheel return when lane change is cancelled, lessen the fluctuation of the vehicle's heading angle when the cancellation occurs, and improve driving comfort during lane changes. It also reduces the control tracking pressure on the downstream control unit in such scenarios, making the lane change decisions and behaviors of the autonomous driving system more decisive, efficient, smooth, and comfortable, and more in line with human driving habits under high traffic density.

[0134] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0135] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0136] The above description is merely a preferred embodiment of the present invention and should not be construed as limiting the scope of the invention. Therefore, any equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for lane change control in autonomous driving, characterized in that, At least comprising the following steps: In the lane changing execution process, if a lane changing cancellation command is received, current ego vehicle position information is obtained, the ego vehicle position information at least including: a lateral offset amount of the ego vehicle relative to a virtual reference line of an original lane And an included angle radian value of the ego vehicle and the virtual reference line of the original lane ; According to the self-vehicle position information, determine the current lane change cancellation level; According to the current lane change cancellation level, obtain the target state and target pose of the self-vehicle corresponding to the current lane change cancellation level, jump the vehicle decision state machine to the target state, and control the self-vehicle to reach the target pose; wherein each lane change cancellation level corresponds to a target state, and the target state includes: self-lane keeping, lane change proposal, lane change preparation, or target lane keeping; Wherein, the second offset threshold is greater than the first offset threshold. When the first offset threshold is less than or equal to the first offset threshold, and the first curvature value is less than or equal to the first curvature value threshold, it is determined that the current lane-changing cancellation level is a first level. When the first offset threshold is less than or equal to the first offset threshold, and the first curvature value is less than or equal to the first curvature value threshold, it is determined that the current lane-changing cancellation level is a first level. When the first offset threshold is less than When the first offset value is less than or equal to a second offset value threshold, and the second offset value is greater than the first offset value threshold, it is determined that the current lane-changing cancellation level is a second level. When the first offset value is less than or equal to a second offset value threshold, and the second offset value is greater than the first offset value threshold, it is determined that the current lane-changing cancellation level is a second level. When the first offset value is less than When the Greater than the second offset threshold, and the If the value is less than or equal to the first radian threshold, the current lane change cancellation level is determined to be level three. When the first offset value is greater than the second offset value threshold, and the first curvature value is greater than the first curvature value threshold, it is determined that the current lane-changing cancellation level is a fourth level. When the first offset value is greater than the second offset value threshold, and the first curvature value is greater than the first curvature value threshold, it is determined that the current lane-changing cancellation level is a fourth level. When the first offset value is greater than the second According to the current lane change cancellation level, obtain the target state and target pose of the self-vehicle corresponding to the current lane change cancellation level, jump the vehicle decision state machine to the target state, and control the self-vehicle to reach the target pose, further comprising:

2. The method of claim 1, wherein, Further comprising: When the current lane-changing cancel level is the first level, the state is jumped to the self-lane keeping state, and the target pose thereof is: and ; When the current lane-changing cancel level is the second level, the state is jumped to the lane-changing proposing state, and the vehicle head is straightened in the current lane without entering the target lane, and the target pose is: ; When the current lane-changing cancellation level is the third level, the state jumps to the lane-changing preparation state, continues to enter the target lane, and aligns the vehicle head in the target lane, and the target pose is: greater than half the width of the lane, and ; When the current lane-changing cancellation level is the fourth level, the state is jumped to the target lane keeping, the current pose is kept, and when a suitable space appears, the vehicle is pasted to the target lane virtual reference line, and the target pose is: .

3. The method of claim 2, wherein, After the self-vehicle jumps into the target state, trigger the judgment logic corresponding to the target state for subsequent control. After the self-vehicle jumps into the target state, and triggers the judgment logic corresponding to the target state for subsequent control, further comprising:

4. The method of claim 3, wherein, When the self-vehicle enters the self-lane keeping or target lane keeping state, trigger the longitudinal speed decision module of the self-vehicle, and then determine whether to accelerate or decelerate currently by judging the motion of the front and rear vehicles in the lane; When the self-vehicle enters the lane change proposal state, determine whether the driving environment continues to meet the necessity and legality of lane changing, if it continues to meet, enter the next lane change preparation state; if it does not meet, return to the lane keeping state; When the self-vehicle enters the lane change preparation state, the decision system determines whether there is a suitable lane change opportunity and duration at each frame, if it meets, enter the lane change execution state to perform lane change operation; and after the self-vehicle has approached the target lane virtual reference line, enter the lane change completion state, and then jump to the lane keeping state. At least comprising: 5.A system for performing a lane change control in autonomous driving, the system comprising: The cancellation level determination unit is configured to determine the current lane change cancellation level according to the self-vehicle position information; The lane-changing position information acquisition unit is configured to acquire current ego vehicle position information during lane-changing execution, if a lane-changing cancellation command is received, wherein the ego vehicle position information at least includes: a lateral offset of the ego vehicle relative to a virtual reference line of the original lane and an included angle radian value of the ego vehicle and the virtual reference line of the original lane ​ The target state switching unit is configured to obtain the target state and target pose of the self-vehicle corresponding to the current lane change cancellation level according to the current lane change cancellation level, jump the vehicle decision state machine to the target state, and control the self-vehicle to reach the target pose; wherein each lane change cancellation level corresponds to a target state, and the target state includes: self-lane keeping, lane change proposal, lane change preparation, or target lane keeping; Wherein, the second offset threshold is greater than the first offset threshold. The target state switching unit further comprises: The first grade determining unit is configured to determine the current lane-changing canceling grade as a first grade when the first offset value is less than or equal to a first offset threshold value, and the second offset value is less than or equal to a second offset threshold value. The first grade determining unit is configured to determine the current lane-changing canceling grade as a first grade when the first offset value is less than or equal to a first offset threshold value, and the second offset value is less than or equal to a second offset threshold value. The first grade determining unit is configured to determine the current lane The second grade determining unit is configured to determine the current lane-changing canceling grade as a second grade when the first grade determining unit determines that the first grade is not met, the second grade is met, and the first grade is met. The second grade determining unit is configured to determine the current lane-changing canceling grade as a second grade when the first grade determining unit determines that the first grade is not met, the second grade is met, and the first grade is met. The second grade determining unit is configured to determine the current lane-changing canceling grade as a second grade The third grade determining unit is configured to determine the current lane-changing canceling grade as a third grade when the first offset value is greater than a second offset value threshold and the second offset value is less than or equal to a first radian value threshold. The third grade determining unit is configured to determine the current lane-changing canceling grade as a third grade when the first offset value is greater than a second offset value threshold and the second offset value is less than or equal to a first radian value threshold. The third grade determining unit is configured to determine the current lane-changing cancel The fourth grade determining unit is configured to determine that the current lane-changing canceling grade is the fourth grade when the first offset value is greater than a second offset value threshold and the second offset value is greater than a first offset value threshold. The fourth grade determining unit is configured to determine that the current lane-changing canceling grade is the fourth grade when the first offset value is greater than a second offset value threshold and the second offset value is greater than a first offset value threshold. The fourth grade determining unit is configured to determine that the current lane-changing canceling Further comprising:

6. The system of claim 5, wherein, The post-switching processing unit is configured to trigger the judgment logic corresponding to the target state for subsequent control after the self-vehicle jumps into the target state; The first switching processing unit is configured to jump to the self-lane keeping state when the current lane-changing cancel level is the first level, and the target pose of the self-lane keeping state is: and ; The second switching processing unit is configured to, when the current lane-changing cancel level is a second level, jump the state to a lane-changing proposal state, no longer continue to enter the target lane, and in-place straighten the vehicle head in the current lane, with a target pose being: ; The third switching processing unit is configured to jump to the lane-changing preparation state when the current lane-changing cancel level is the third level, continue to enter the target lane, and align the vehicle head in the target lane, with a target pose being: greater than half the width of the lane, and ; The third switching processing unit is configured to, when the current lane-changing cancel level is the fourth level, jump the state to the target lane keeping, keep the current pose, and when a suitable space appears, stick to the target lane virtual reference line, and the target pose is: .

7. The system of claim 6, wherein, The post-switching processing unit further comprises: The lane keeping processing unit is configured to trigger the longitudinal speed decision module of the self-vehicle when the self-vehicle enters the self-lane keeping or target lane keeping state, and then determine whether to accelerate or decelerate currently by judging the motion of the front and rear vehicles in the lane; ​ ​ The lane changing proposal processing unit is configured to determine whether the driving environment continues to satisfy the necessity and legality of lane changing when the ego vehicle enters the lane changing proposal state, and if so, enter the next lane changing preparation state; if not, return to the lane keeping state. The lane changing preparation processing unit is configured to determine whether there is a suitable lane changing opportunity and duration when the ego vehicle enters the lane changing preparation state, and if so, enter the lane changing execution state to perform the lane changing operation; and enter the lane changing completion state after the ego vehicle has approached the target lane virtual reference line, and then jump to the lane keeping state.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the steps of the method of any one of claims 1 to 4.

9. A computing device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein, The processor, when executing the computer program, implements the steps of the method of any one of claims 1 to 4. 10.A vehicle, wherein at least a lane changing decision device is arranged on the vehicle, and the lane changing decision device is provided with the system of any one of claims 5 to 7.

Images