A trajectory correction method, device, equipment and medium

Abstract

The application discloses a trajectory correction method, device, equipment and medium, and relates to the technical field of automatic driving, which comprises the following steps: when a target vehicle cuts into a self-lane, a first prediction trajectory of the target vehicle is acquired, and a trajectory endpoint position of the first prediction trajectory is determined; a target front vehicle closest to the self-vehicle in the self-lane is determined, and a second prediction trajectory of the target front vehicle is acquired, so as to determine a target endpoint position of the target front vehicle in the self-lane based on the second prediction trajectory; whether the longitudinal distance between the trajectory endpoint position and the self-vehicle position is greater than the longitudinal distance between the target endpoint position and the self-vehicle position is judged; if yes, the first prediction trajectory is corrected based on a preset polynomial fitting method to obtain a corrected trajectory, and a driving path of the self-vehicle is planned based on the corrected trajectory, otherwise, the driving path of the self-vehicle is directly planned based on the first prediction trajectory. The application can effectively correct the prediction trajectory and guarantee the safety of automatic driving.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a predicted trajectory correction method, device, equipment, and medium. Background Art

[0002] Trajectory prediction and speed planning are crucial components of autonomous driving technology, directly impacting the safety and smoothness of driving. Currently, the Apollo (an autonomous driving method) approach is widely used in planning, but it suffers from significant drawbacks. Its longitudinal speed planning relies heavily on predicted trajectories. In real-world traffic scenarios, predictions struggle to accurately describe the future state of moving objects, especially in complex traffic flows. Traffic flows involve numerous vehicles, pedestrians, and other obstacles, all of which influence and interact with each other. For example, multiple vehicles changing lanes simultaneously in close proximity, vehicles accelerating and decelerating suddenly, and pedestrians crossing the road at random are common occurrences. Accurately describing the interactions between these objects is extremely difficult.

[0003] Taking the cut-in scenario of other vehicles as an example, when a vehicle quickly cuts into the lane, due to its high speed, the existing prediction model is prone to deviation when predicting its motion trajectory. For example, the predicted trajectory is prone to cross the vehicle in front. Figure 1 In this case, if the autonomous driving system performs longitudinal velocity planning based on a flawed predicted trajectory, an anomaly will occur. Due to the inaccurate prediction, the ego vehicle will misjudge the impact of the cut-in vehicle on itself during longitudinal velocity planning, resulting in an underestimated impact of the cut-in vehicle on the ego vehicle's braking, which undoubtedly increases the risk of accidents such as rear-end collisions.

[0004] In summary, how to effectively correct the predicted trajectory to ensure that the predicted trajectory truly conforms to the actual scenario, thereby ensuring the safety and reliability of autonomous driving, is a problem that needs to be solved. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a predicted trajectory correction method, device, equipment and medium that can effectively correct the predicted trajectory to ensure that the predicted trajectory truly conforms to the actual scenario, thereby ensuring the safety and reliability of autonomous driving. The specific solution is as follows:

[0006] In a first aspect, the present application discloses a method for correcting a predicted trajectory, comprising:

[0007] When the target vehicle cuts into the own lane, obtaining a first predicted trajectory obtained by predicting the motion trajectory of the target vehicle, and determining a trajectory end position of the first predicted trajectory;

[0008] determining a target front vehicle closest to the ego vehicle in the ego lane, and obtaining a second predicted trajectory of the target front vehicle after predicting a motion trajectory of the target front vehicle, to determine a target end position of the target front vehicle in the ego lane based on the second predicted trajectory;

[0009] judging whether a longitudinal distance between the trajectory end position and the ego vehicle position is greater than a longitudinal distance between the target end position and the ego vehicle position;

[0010] if yes, correcting the first predicted trajectory based on a preset polynomial fitting method to obtain a corrected trajectory, and planning a driving path of the ego vehicle based on the corrected trajectory, otherwise directly planning the driving path of the ego vehicle based on the first predicted trajectory.

[0011] Optionally, the determining of the target end position of the target front vehicle in the ego lane based on the second predicted trajectory comprises:

[0012] judging whether a trajectory end of the second predicted trajectory is located in the ego lane;

[0013] if yes, determining a position where the trajectory end of the second predicted trajectory is located as the target end position of the target front vehicle in the ego lane;

[0014] if no, extending a lane line of the ego lane outward by a preset distance to obtain an extended lane line, and determining the target end position of the target front vehicle in the ego lane based on an intersection position of the second predicted trajectory and the extended lane line.

[0015] Optionally, the extending of the lane line of the ego lane outward by the preset distance to obtain the extended lane line comprises:

[0016] determining a target lane line intersecting the second predicted trajectory; the target lane line is any one of a left lane line or a right lane line of the ego lane;

[0017] determining a curve model corresponding to the target lane line, and determining a derivative equation corresponding to the curve model, to construct a corresponding tangent line model based on each lane line point on the target lane line and the derivative equation;

[0018] obtaining a corresponding tangent line based on the tangent line model, and translating each lane line point by a preset distance along a direction perpendicular to the corresponding tangent line to obtain an extended lane line.

[0019] Optionally, the second predicted trajectory comprises a plurality of first discrete points, the extended lane line comprises a plurality of second discrete points, and the determination of the intersection position of the second predicted trajectory and the extended lane line comprises:

[0020] constructing a first straight line based on any two adjacent first discrete points;

[0021] constructing a second straight line based on any two adjacent second discrete points;

[0022] if a target intersection point between the first straight line and the second straight line is located between the any two adjacent first discrete points and between the any two adjacent second discrete points, taking the target intersection point as an intersection position of the second predicted trajectory and the extended lane line.

[0023] Optionally, the determining the target end position of the target front vehicle in the self-lane based on the intersection position of the second predicted trajectory and the extended lane line comprises:

[0024] determining a target discrete point closer to the self-vehicle from the any two adjacent first discrete points based on the intersection position of the second predicted trajectory and the extended lane line;

[0025] obtaining position coordinate information of the target discrete point in a self-vehicle coordinate system, and determining a vehicle angle of the target front vehicle at the target discrete point; wherein the self-vehicle coordinate system is a coordinate system established based on a rear axle center of the self-vehicle;

[0026] calculating a tail center point position coordinate of the target front vehicle based on the position coordinate information, the vehicle angle and vehicle length information of the target front vehicle, and determining a position represented by the tail center point position coordinate as the target end position of the target front vehicle in the self-lane.

[0027] Optionally, the determining the vehicle angle of the target front vehicle at the target discrete point comprises:

[0028] determining a first slope of the first straight line corresponding to the intersection position;

[0029] determining a second slope of the second straight line corresponding to the intersection position;

[0030] determining the vehicle angle of the target front vehicle at the target discrete point based on the first slope and the second slope.

[0031] Optionally, the correcting the first predicted trajectory based on a preset polynomial fitting method to obtain a corrected trajectory comprises:

[0032] taking a current position and a current motion state of the target vehicle as a starting planning position and a starting motion state respectively; the motion state comprises lateral velocity, longitudinal velocity, lateral acceleration and longitudinal acceleration;

[0033] determine a predicted motion state of the target front vehicle at the target end position based on the second predicted trajectory, and take the target end position and the predicted motion state as an end point planning position and an end point motion state respectively;

[0034] determine a target model of a predicted motion trajectory of the target vehicle in advance, and solve the target model based on the start planning position, the start motion state, the end point planning position, the end point motion state and a preset driving time to obtain the predicted motion trajectory;

[0035] take the predicted motion trajectory as a corrected trajectory after correcting the first predicted trajectory.

[0036] In a second aspect, the present application discloses a predicted trajectory correction device, a first position determination module is configured to, when a target vehicle cuts into a self-lane, acquire a first predicted trajectory obtained by predicting a motion trajectory of the target vehicle, and determine a trajectory end position of the first predicted trajectory;

[0037] a second position determination module is configured to determine a target front vehicle closest to a self-vehicle position in the self-lane, and acquire a second predicted trajectory obtained by predicting a motion trajectory of the target front vehicle, so as to determine a target end position of the target front vehicle in the self-lane based on the second predicted trajectory;

[0038] a judgment module is configured to judge whether a longitudinal distance between the trajectory end position and the self-vehicle position is greater than a longitudinal distance between the target end position and the self-vehicle position;

[0039] a trajectory correction module is configured to, if yes, correct the first predicted trajectory based on a preset polynomial fitting method to obtain a corrected trajectory, and plan a driving path of the self-vehicle based on the corrected trajectory, or directly plan the driving path of the self-vehicle based on the first predicted trajectory.

[0040] In a third aspect, the present application discloses an electronic device, comprising:

[0041] a memory configured to save a computer program;

[0042] a processor configured to execute the computer program to realize steps of the predicted trajectory correction method disclosed above.

[0043] In a fourth aspect, the present application discloses a computer readable storage medium configured to store a computer program; wherein the computer program is executed by a processor to realize steps of the predicted trajectory correction method disclosed above.

[0044] It can be seen that, in the present application, when the target vehicle cuts into the self-lane, the first prediction trajectory obtained after predicting the motion trajectory of the target vehicle is acquired, and the trajectory endpoint position of the first prediction trajectory is determined; the target front vehicle closest to the self-vehicle position in the self-lane is determined, and the second prediction trajectory obtained after predicting the motion trajectory of the target front vehicle is acquired, so as to determine the target endpoint position of the target front vehicle in the self-lane based on the second prediction trajectory; it is judged whether the longitudinal distance between the trajectory endpoint position and the self-vehicle position is greater than the longitudinal distance between the target endpoint position and the self-vehicle position; if yes, the first prediction trajectory is corrected based on a preset polynomial fitting method to obtain a corrected trajectory, and the self-vehicle driving path is planned based on the corrected trajectory, otherwise the self-vehicle driving path is directly planned based on the first prediction trajectory.

[0045] Beneficial effects: When there is a target vehicle cutting into the self-lane, the first prediction trajectory obtained after predicting the motion trajectory of the target vehicle is acquired, and the trajectory endpoint position of the first prediction trajectory is determined. Further, the target front vehicle closest to the self-vehicle position in the self-lane is determined, and the second prediction trajectory obtained after predicting the motion trajectory of the target front vehicle is acquired, so as to determine the target endpoint position of the target front vehicle in the self-lane based on the second prediction trajectory. Then it is judged whether the longitudinal distance between the trajectory endpoint position and the self-vehicle position is greater than the longitudinal distance between the target endpoint position and the self-vehicle position, if not, it means that the current first prediction trajectory output is normal, and the self-vehicle driving path can be directly planned based on the first prediction trajectory; if yes, it means that the prediction trajectory of the target vehicle passes through the front vehicle, so the first prediction trajectory at this time is abnormal and does not conform to the actual scene, and therefore needs to be corrected. Specifically, the first prediction trajectory is corrected based on a preset polynomial fitting method to obtain a corrected trajectory, and the prediction trajectory of the target vehicle is corrected, so that the trajectory of the target vehicle will not pass through the front vehicle, avoiding rear-end collision and other accidents, so that the self-vehicle can effectively realize stop according to the corrected trajectory, ensuring the safety and reliability of automatic driving. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only embodiments of the present application, and those skilled in the art can obtain other drawings according to the provided drawings without creating any inventive labor.

[0047] Figure 1 An abnormal prediction trajectory disclosed in the present application is shown in the figure;

[0048] Figure 2A flow chart of a prediction trajectory correction method disclosed in the present application;

[0049] Figure 3 A flow chart of a specific prediction trajectory correction method disclosed in the present application;

[0050] Figure 4 A schematic diagram of a lane line extension process disclosed in the present application;

[0051] Figure 5 A schematic diagram of a lane line extension result disclosed in the present application;

[0052] Figure 6 A schematic diagram of a vehicle angle disclosed in the present application;

[0053] Figure 7 A schematic diagram of a prediction trajectory correction device disclosed in the present application;

[0054] Figure 8 A structural diagram of an electronic device disclosed in the present application. DETAILED DESCRIPTION

[0055] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0056] When a vehicle quickly cuts into the lane of the ego vehicle, due to its high speed, the final prediction trajectory is prone to deviation when the existing prediction model predicts the motion trajectory of the vehicle, such as the prediction trajectory crossing the front vehicle. In this case, if the automatic driving system performs longitudinal speed planning according to the problematic prediction trajectory, an abnormality will occur. Due to the inaccuracy of the prediction, the ego vehicle will make a wrong judgment on the influence of the cut-in vehicle on itself when performing longitudinal speed planning, so that the influence of the cut-in vehicle on the ego vehicle is underestimated, which undoubtedly increases the risk of rear-end collision and other accidents.

[0057] Therefore, the embodiments of the present application disclose a prediction trajectory correction method, device, equipment and medium, which can effectively correct the prediction trajectory and ensure that the prediction trajectory truly conforms to the actual scene, thereby ensuring the safety and reliability of automatic driving.

[0058] Referring to Figure 2 The embodiments of the present application disclose a prediction trajectory correction method, which comprises:

[0059] Step S11: When the target vehicle cuts into the self-lane, a first predicted trajectory obtained after predicting the motion trajectory of the target vehicle is acquired, and a trajectory endpoint position of the first predicted trajectory is determined.

[0060] In this embodiment, when there is a target vehicle cutting into the self-lane, a first predicted trajectory obtained after predicting the motion trajectory of the target vehicle is acquired, and a trajectory endpoint position of the first predicted trajectory is determined. It should be noted that the predicted trajectory is actually composed of a plurality of discrete points, and the present application determines the trajectory endpoint position according to the last discrete point of the predicted trajectory. In one specific embodiment, the position of the last discrete point can be directly taken as the trajectory endpoint position, and in another specific embodiment, the longitudinal position of the last discrete point can be reduced by half the vehicle length and then combined with the lateral position as the trajectory endpoint position.

[0061] It can be understood that the autonomous vehicle relies on a variety of sensors to work together, such as lidar, camera, millimeter wave radar, etc. Among them, the lidar obtains high-precision distance information of surrounding objects by emitting laser beams and measuring the time of reflected light, and generates point cloud data; the camera collects images around the vehicle and detects various target objects using image recognition technology; the millimeter wave radar monitors the distance, speed and angle change of the object in real time. The data collected by these sensors is transmitted to the sensor fusion module, which processes the data to accurately determine the position, speed, motion direction, etc. of the target object, provides basic data for the predicted trajectory, and then outputs the preliminary predicted trajectory information.

[0062] Step S12: A target front vehicle closest to the self-vehicle position in the self-lane is determined, a second predicted trajectory obtained after predicting the motion trajectory of the target front vehicle is acquired, and a target endpoint position of the target front vehicle in the self-lane is determined based on the second predicted trajectory.

[0063] In this embodiment, a target front vehicle closest to the self-vehicle position in the self-lane is also determined, a second predicted trajectory obtained after predicting the motion trajectory of the target front vehicle is acquired, and a target endpoint position of the target front vehicle in the self-lane is determined based on the second predicted trajectory. It should be noted that the target front vehicle can perform a lane change operation, and the present application mainly focuses on the trajectory of the target front vehicle in the self-lane, so it is necessary to determine the final position of the target front vehicle in the self-lane.

[0064] Among them, the following methods are usually adopted when finding the target front vehicle:

[0065] First, find and store the left and right lane line points of the self-lane detected by the self-vehicle intelligent camera;

[0066] Then the current position of each vehicle is traversed, and it is respectively judged whether the position of each vehicle is in the self-lane. The judgment method is as follows: first, filter the vehicles located behind the ego vehicle. Specifically, a coordinate system can be established with the rear axle center of the current position of the ego vehicle as the origin, so as to obtain the position coordinates of each vehicle. Assuming that the position of a vehicle is (x, y), if x is less than 0, it indicates that it is located behind the ego vehicle, so it is filtered out. Then, the specific position of each vehicle in the self-lane is determined according to (x, y). The lane line points x i When a certain x i <x and x i+1 > x, for the left lane line, x i corresponding y i is greater than y, and for the right lane line, x i corresponding y i is less than y, if the above conditions are all met, the corresponding vehicle is marked as the front vehicle in the self-lane. After traversal, a set of front vehicle data can be obtained. Finally, the smallest value of x is searched from the front vehicle data, and the vehicle is the target front vehicle closest to the position of the ego vehicle.

[0067] Step S13: judging whether the longitudinal distance between the trajectory endpoint position of the trajectory of the target vehicle and the position of the ego vehicle is greater than the longitudinal distance between the target endpoint position in the self-lane of the predicted trajectory of the target front vehicle and the position of the ego vehicle.

[0068] In this embodiment, after obtaining the trajectory endpoint position of the predicted trajectory of the target vehicle and the target endpoint position in the self-lane of the predicted trajectory of the target front vehicle, it is needed to judge whether the longitudinal distance between the trajectory endpoint position and the position of the ego vehicle is greater than the longitudinal distance between the target endpoint position and the position of the ego vehicle.

[0069] Step S14: if yes, correcting the first predicted trajectory based on a preset polynomial fitting method to obtain a corrected trajectory, and planning a driving path of the ego vehicle based on the corrected trajectory, otherwise directly planning a driving path of the ego vehicle based on the first predicted trajectory.

[0070] In a specific embodiment, if the longitudinal distance between the trajectory endpoint position and the position of the ego vehicle is greater than the longitudinal distance between the target endpoint position and the position of the ego vehicle, it indicates that the predicted trajectory of the target vehicle passes through the front vehicle, that is, the first predicted trajectory at this time is abnormal and does not conform to the actual scene, so it needs to be corrected. Specifically, the first predicted trajectory is corrected based on a preset polynomial fitting method to obtain a corrected trajectory, so as to ensure that the trajectory of the target vehicle will not pass through the front vehicle, so that the ego vehicle can effectively realize stop according to the corrected trajectory, and the safety and reliability of automatic driving are ensured.

[0071] In another specific embodiment, if the longitudinal distance between the trajectory end position and the ego vehicle position is not greater than the longitudinal distance between the target end position and the ego vehicle position, it indicates that the predicted trajectory of the target vehicle does not pass through the front vehicle, i.e., the current first predicted trajectory output is normal, and thus the ego vehicle driving path can be directly planned based on the first predicted trajectory

[0072] In the specific embodiment, the above-mentioned correction of the first predicted trajectory based on the preset polynomial fitting method includes: taking the current position and the current motion state of the target vehicle as a starting planning position and a starting motion state, respectively; the motion state includes a lateral velocity, a longitudinal velocity, a lateral acceleration, and a longitudinal acceleration; determining a predicted motion state of the target front vehicle at the target end position based on the second predicted trajectory, and taking the target end position and the predicted motion state as an end planning position and an end motion state, respectively; determining a target model about the predicted motion trajectory of the target vehicle that is pre-constructed, and solving the target model based on the starting planning position, the starting motion state, the end planning position, the end motion state, and a preset driving time to obtain the predicted motion trajectory; and taking the predicted motion trajectory as a corrected trajectory obtained by correcting the first predicted trajectory.

[0073] It can be understood that the process of correcting the first predicted trajectory, i.e., the process of re-predicting the motion trajectory of the target vehicle, first needs to confirm the starting point and the end point of the re-planning. Specifically, the current position and the current motion state of the target vehicle are determined, and the current position and the current motion state are taken as a starting planning position and a starting motion state, respectively. The position specifically refers to a lateral position and a longitudinal position, and the motion state specifically includes a lateral velocity, a longitudinal velocity, a lateral acceleration, and a longitudinal acceleration of the vehicle. Further, the farthest position that the target vehicle is allowed to drive and reach is re-planned in this embodiment, which is the target end position of the target front vehicle in the ego lane, i.e., the predicted motion state of the target front vehicle at the target end position is determined based on the second predicted trajectory, and the target end position and the motion state are taken as an end planning position and an end motion state, respectively. It should be pointed out that the lateral acceleration and the longitudinal acceleration at the end planning position are both 0, and the speed at the end planning position is the speed of the target front vehicle.

[0074] Further, a target model about the predicted motion trajectory of the target vehicle that is pre-constructed is determined. In this embodiment of the application, a 5th order polynomial is used to fit the trajectory of the target vehicle, and the trajectory equation is represented as:

[0075] x(t) = a0 + a1t + a2t 2 +a3t 3 +a4t 4 +a5t5 ;

[0076] y(t) = b0 + b1t + b2t 2 +b3t 3 +b4t 4 +b5t 5 ;

[0077] Then boundary conditions are set based on the start planning position, the start motion state, the end planning position and the end motion state, and the target model is solved based on the boundary conditions and the preset driving time to obtain the predicted motion trajectory.

[0078] Specifically, based on the boundary conditions, the following equation group can be obtained:

[0079] ;

[0080] ;

[0081] Wherein, the start planning position is represented as (x0, y0); wherein, x0 is the start longitudinal position, y0 is the start transverse position; the start longitudinal velocity is v x0 , the start transverse velocity is v y0 , the start longitudinal acceleration is a x0 , and the start transverse acceleration is a y0 ; the end planning position is (x T , y T ), wherein, x T is the end longitudinal position, y T is the end transverse position, the end longitudinal velocity is v xT , the end transverse velocity is v yT , the end longitudinal acceleration is a xT =0, and the end transverse acceleration is a yT =0; t0 and t f represent the start time and the end time respectively, the start time is taken as 0 in the embodiment, and the end time t f =2s / (v x0 +v xT ), wherein s is the difference between x T and x0.

[0082] The specific expressions of x(t) and y(t) can be obtained by solving the above equation group according to the above parameters. Then the preset driving time is determined in the application, assuming that the preset driving time is 6s, then the curve of the equation at t∈[0, 6] is approximately considered as the predicted trajectory of the target vehicle. It should be pointed out that the above preset driving time is a parameter that can be calibrated, but it cannot be too long, because for prediction, the longer the prediction time is, the lower the accuracy of the prediction result will be.

[0083] It can be seen that when the target vehicle cuts into the lane of the ego vehicle, the first predicted trajectory obtained by predicting the motion trajectory of the target vehicle is needed to be acquired, and the trajectory endpoint position of the first predicted trajectory is determined. Further, the target front vehicle closest to the ego vehicle in the lane of the ego vehicle is determined, and the second predicted trajectory obtained by predicting the motion trajectory of the target front vehicle is acquired, so as to determine the target endpoint position of the target front vehicle in the lane of the ego vehicle based on the second predicted trajectory. Then, it is judged whether the longitudinal distance between the trajectory endpoint position and the ego vehicle position is greater than the longitudinal distance between the target endpoint position and the ego vehicle position. If not, it indicates that the current first predicted trajectory is normal, and the driving path of the ego vehicle can be planned based on the first predicted trajectory directly. If yes, it indicates that the predicted trajectory of the target vehicle passes through the front vehicle, so the first predicted trajectory at this time is abnormal and does not conform to the actual scene, and therefore needs to be corrected. Specifically, the first predicted trajectory is corrected based on a preset polynomial fitting method to obtain a corrected trajectory, so as to ensure that the trajectory of the target vehicle will not pass through the front vehicle, avoid rear-end collision and other accidents, and enable the ego vehicle to effectively realize stop according to the corrected trajectory, thereby ensuring the safety and reliability of automatic driving.

[0084] Referring to Figure 3 The embodiment of the present application discloses a specific prediction trajectory correction method. Compared with the previous embodiment, the technical solution is further described and optimized. Specifically, it includes the following steps:

[0085] Step S21: When the target vehicle cuts into the lane of the ego vehicle, the first predicted trajectory obtained by predicting the motion trajectory of the target vehicle is acquired, and the trajectory endpoint position of the first predicted trajectory is determined.

[0086] Step S22: The target front vehicle closest to the ego vehicle in the lane of the ego vehicle is determined, and the second predicted trajectory obtained by predicting the motion trajectory of the target front vehicle is acquired, so as to judge whether the trajectory endpoint of the second predicted trajectory is located in the lane of the ego vehicle.

[0087] In this embodiment, considering that the front vehicle may perform lane changing, the present application needs to judge whether the trajectory endpoint of the second predicted trajectory of the target front vehicle is located in the lane of the ego vehicle. If yes, it indicates that the target front vehicle does not perform lane changing operation, otherwise, it performs lane changing operation.

[0088] Step S23: If the trajectory endpoint of the second predicted trajectory is located in the lane of the ego vehicle, the position of the trajectory endpoint of the second predicted trajectory is determined as the target endpoint position of the target front vehicle in the lane of the ego vehicle.

[0089] In the embodiment, if the trajectory end point of the second prediction trajectory is located in the self-lane, it indicates that the preceding vehicle does not change lanes, and the trajectory end point of the second prediction trajectory is the target end point position of the target preceding vehicle in the self-lane.

[0090] In a specific implementation, the point at the end of the second prediction trajectory can be directly taken as the target end point position in the self-lane; in another specific implementation, a point at the end of the second prediction trajectory (x final , y final ) can be taken first, which is the final trajectory position of the target preceding vehicle, and then the longitudinal position of the point is subtracted by half the length of the preceding vehicle, which is represented as xF≈x final -0.5L, L being the length of the target preceding vehicle.

[0091] It should be noted that the determination manners of the end point positions of the first prediction trajectory and the second prediction trajectory in the self-lane need to be consistent.

[0092] Step S24: If not located in the self-lane, the lane line of the self-lane is expanded outward by a preset distance to obtain an expanded lane line, and the target end point position of the target preceding vehicle in the self-lane is determined based on the intersection position of the second prediction trajectory and the expanded lane line.

[0093] In the embodiment, if the trajectory end point of the second prediction trajectory is not located in the self-lane, the target end point position of the target preceding vehicle in the self-lane needs to be further determined. The method specifically adopted in the embodiment is to expand the lane line of the self-lane outward by a preset distance to obtain an expanded lane line, and finally determine the target end point position of the target preceding vehicle in the self-lane based on the intersection position of the second prediction trajectory and the expanded lane line.

[0094] Specifically, the above expanding the lane line of the self-lane outward by a preset distance to obtain an expanded lane line includes: determining a target lane line intersecting the second prediction trajectory; the target lane line is any one of the left lane line or the right lane line of the self-lane; determining a curve model corresponding to the target lane line, and determining a derivative equation corresponding to the curve model, to construct a corresponding tangent line model based on each lane line point on the target lane line and the derivative equation; obtaining a corresponding tangent line based on the tangent line model, and translating each lane line point by a preset distance along a direction perpendicular to the corresponding tangent line to obtain an expanded lane line.

[0095] It can be understood that the target front vehicle will intersect with either the left lane line or the right lane line of the self-lane during the lane changing process, so the application first determines the target lane line intersecting with the second prediction trajectory, and then further fits the target lane line to obtain a curve model corresponding to the target lane line and a derivative equation corresponding to the curve model. Further, the target lane line can be considered to be composed of a plurality of lane line points, and the application needs to construct a corresponding tangent line model based on each lane line point on the target lane line and the derivative equation, so as to obtain a corresponding tangent line based on the tangent line model, and finally translate each lane line point by a preset distance along a direction perpendicular to the corresponding tangent line to obtain an expanded lane line.

[0096] As shown in Figure 4 , the application takes the left lane line as an example, and first uses the least square method to fit the left lane line to obtain a curve model f x . Then the derivative equation of the curve model is calculated as d(f x ), each lane line point (x i , y i ) is traversed in turn, and a tangent line model of each lane line point is obtained. It can be understood that with the slope of the tangent line (i.e. the value of the derivative equation at the tangent point) and the coordinates of the tangent point, the point-slope equation of a straight line y-y1=m(x-x1) can be used to obtain the tangent line model, wherein (x1, y1) is the coordinates of the tangent point, and m is the slope of the tangent line. Once the tangent line model is obtained, a ray perpendicular to the tangent line model (towards the left) can be made, and a new point (x ni and y ni ) at a preset distance d can be obtained. For each x i , y i , there is such a point, so the x ni and y ni combination forms an expanded lane line with an expansion distance of d, and the expanded lane line is shown by the green line in Figure 5 . The application needs to determine the intersection position between the yellow dashed line and the green straight line. Wherein, d is a parameter that can be calibrated, and can be taken as 0.2 times the lane width.

[0097] In the specific embodiment, the second predicted trajectory includes a plurality of first discrete points, the extended lane line includes a plurality of second discrete points, and thus the determination of the intersection position of the second predicted trajectory and the extended lane line includes: constructing a first straight line based on any two adjacent first discrete points; constructing a second straight line based on any two adjacent second discrete points; if a target intersection between the first straight line and the second straight line is located between the any two adjacent first discrete points and between the any two adjacent second discrete points, regarding the target intersection as the intersection position of the second predicted trajectory and the extended lane line.

[0098] That is, the second predicted trajectory can be considered to be composed of a plurality of first discrete points, and the extended lane line can be considered to be composed of a plurality of second discrete points, and thus, in the determination of the intersection position of the second predicted trajectory and the extended lane line, a first straight line is constructed based on any two adjacent first discrete points, and a second straight line is constructed based on any two adjacent second discrete points, if a target intersection between the first straight line and the second straight line is located between the any two adjacent first discrete points and between the any two adjacent second discrete points, the target intersection is the intersection position of the second predicted trajectory and the extended lane line.

[0099] The specific process is as follows: traversing two groups of discrete points, wherein in the ego vehicle coordinate system, the first discrete points corresponding to the extended lane line are denoted as x m , y m , the second discrete points corresponding to the extended lane are denoted as x c , y c , two straight lines are drawn, and the two straight lines are ((x m , y m ), (x m+1 , y m+1 )) and ((x c , y c ), (x c+1 , y c+1 )) respectively, wherein the corresponding slopes are k1 and k2, and after the drawing, the intersection is solved, when the x and y of the intersection are between (x m , y m ) and (x m+1 , y m+1 ) and (x c , y c ) and (x c+1 , y c+1 ), it is indicated that the intersection is the intersection position of the second predicted trajectory and the extended lane line.

[0100] Further, the above determining the target endpoint position of the target front vehicle in the self-lane based on the intersection position of the second predicted trajectory and the extended rear lane line comprises: determining a target discrete point closer to the self-vehicle from the two adjacent first discrete points based on the intersection position of the second predicted trajectory and the extended rear lane line; obtaining position coordinate information of the target discrete point in a self-vehicle coordinate system, and determining a vehicle angle of the target front vehicle at the target discrete point; wherein the self-vehicle coordinate system is a coordinate system established based on a rear axle center of the self-vehicle; calculating a tail center point position coordinate of the target front vehicle based on the position coordinate information, the vehicle angle and a vehicle length information of the target front vehicle, and determining a position represented by the tail center point position coordinate as the target endpoint position of the target front vehicle in the self-lane.

[0101] That is, the present application determines the target discrete point closer to the self-vehicle from the two adjacent first discrete points based on the intersection position, that is, takes (x c , y c ) as the point at which the target front vehicle reaches the extended rear lane line, and (x c , y c ) represents the position coordinate information of the target discrete point in the self-vehicle coordinate system. Further, when the target front vehicle reaches (x c , y c ), the vehicle has a certain angle, so the present application also needs to determine the vehicle angle of the target front vehicle at the target discrete point, so as to calculate the tail center point position coordinate of the target front vehicle based on the position coordinate information, the vehicle angle θ and the vehicle length information L of the target front vehicle. As shown in the example of Figure 6 , when the target front vehicle changes lanes from the left side, the tail center point position coordinate is represented as (x ), and the position represented by the tail center point position coordinate is the target endpoint position of the target front vehicle in the self-lane. In addition, when the target front vehicle changes lanes from the right side, the tail center point position coordinate is represented as (x .

[0102] Further, the above determining the vehicle angle of the target front vehicle at the target discrete point comprises: determining a first slope of the first straight line corresponding to the intersection position; determining a second slope of the second straight line corresponding to the intersection position; and determining the vehicle angle of the target front vehicle at the target discrete point based on the first slope and the second slope. That is, as known from the foregoing, the intersection position is determined by two straight lines, assuming that the slope of the first straight line corresponding to the intersection position is k1, and the slope of the second straight line corresponding to the intersection position is k2, then the calculation expression of the vehicle angle is:

[0103] .

[0104] wherein k1xk2=-1 means that θ is 90 degrees.

[0105] Step S25: judging whether a longitudinal distance between the trajectory endpoint position and the ego vehicle position is greater than a longitudinal distance between the target endpoint position and the ego vehicle position.

[0106] Step S26: if yes, correcting the first predicted trajectory based on a preset polynomial fitting method to obtain a corrected trajectory, and planning a driving path of the ego vehicle based on the corrected trajectory, otherwise directly planning the driving path of the ego vehicle based on the first predicted trajectory.

[0107] wherein more specific processing procedures about the above steps S21, S25 and S26 can refer to the corresponding contents disclosed in the foregoing embodiments, which will not be described here in detail.

[0108] It can be seen that, considering that the preceding vehicle may perform lane changing, the present application needs to judge whether the trajectory endpoint of the second predicted trajectory of the target preceding vehicle is located in the ego lane, if yes, it means that the target preceding vehicle does not perform lane changing, otherwise it performs lane changing. When the preceding vehicle does not change lanes, the trajectory endpoint of the second predicted trajectory is the target endpoint position of the target preceding vehicle in the ego lane, if it changes lanes, the lane line of the ego lane needs to be expanded outward by a preset distance to obtain an expanded lane line, and then the target endpoint position of the target preceding vehicle in the ego lane is finally determined based on the intersection position of the second predicted trajectory and the expanded lane line. It can be seen that, for the two cases of lane changing and not changing lanes of the preceding vehicle, the present application determines the target endpoint position of the target preceding vehicle in the ego lane, and then performs the process of comparing the longitudinal distance between the trajectory endpoint position and the ego vehicle position with the longitudinal distance between the target endpoint position and the ego vehicle position, so as to more accurately judge whether the predicted trajectory needs to be corrected.

[0109] Referring to Figure 7 the present application discloses a predicted trajectory correction device, which comprises:

[0110] a first position determining module 11, configured to acquire a first predicted trajectory obtained by predicting a motion trajectory of a target vehicle when the target vehicle cuts into an ego lane, and determine a trajectory endpoint position of the first predicted trajectory;

[0111] a second position determining module 12, configured to determine a target preceding vehicle closest to an ego vehicle position in the ego lane, and acquire a second predicted trajectory obtained by predicting a motion trajectory of the target preceding vehicle, so as to determine a target endpoint position of the target preceding vehicle in the ego lane based on the second predicted trajectory;

[0112] a judging module 13, configured to judge whether a longitudinal distance between the trajectory endpoint position and the ego vehicle position is greater than a longitudinal distance between the target endpoint position and the ego vehicle position;

[0113] a trajectory correction module 14, configured to, if yes, correct the first predicted trajectory based on a preset polynomial fitting method to obtain a corrected trajectory, and plan a driving path of the ego vehicle based on the corrected trajectory, or directly plan the driving path of the ego vehicle based on the first predicted trajectory if no.

[0114] It can be seen that when the target vehicle cuts into the ego lane, the first predicted trajectory obtained by predicting the motion trajectory of the target vehicle is needed, and the trajectory endpoint position of the first predicted trajectory is determined. Further, the target front vehicle closest to the ego vehicle position in the ego lane is determined, and the second predicted trajectory obtained by predicting the motion trajectory of the target front vehicle is obtained, so as to determine the target endpoint position of the target front vehicle in the ego lane based on the second predicted trajectory. Then, it is judged whether the longitudinal distance between the trajectory endpoint position and the ego vehicle position is greater than the longitudinal distance between the target endpoint position and the ego vehicle position. If no, it indicates that the current first predicted trajectory output is normal, and the driving path of the ego vehicle can be directly planned based on the first predicted trajectory. If yes, it indicates that the predicted trajectory of the target vehicle passes through the front vehicle, so the first predicted trajectory at this time is abnormal and does not conform to the actual scene, and therefore needs to be corrected. Specifically, the first predicted trajectory is corrected based on a preset polynomial fitting method to obtain a corrected trajectory, so as to correct the predicted trajectory of the target vehicle, so as to ensure that the trajectory of the target vehicle will not pass through the front vehicle, avoid rear-end collision and other accidents, and enable the ego vehicle to effectively stop according to the corrected trajectory, thereby ensuring the safety and reliability of automatic driving.

[0115] Figure 8 A structural schematic diagram of an electronic device provided by an embodiment of the present application is provided. Specifically, it can include at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is configured to store a computer program, and the processor 21 is configured to load and execute the computer program to implement the related steps in the trajectory prediction correction method performed by the electronic device disclosed in any of the preceding embodiments.

[0116] In this embodiment, the power supply 23 is configured to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 is configured to create a data transmission channel between the electronic device 20 and external devices, and the communication protocol followed by the communication interface 24 can be any communication protocol applicable to the technical solution of the present application, which will not be specifically limited herein; the input and output interface 25 is configured to obtain external input data or output data to the outside, and the specific interface type can be selected according to the specific application needs, which will not be specifically limited herein.

[0117] The processor 21 can include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 21 can be implemented in at least one of a hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), and a PLA (Programmable Logic Array). The processor 21 can also include a main processor and a coprocessor. The main processor is a processor for processing data in a wake-up state, also known as a CPU (Central Processing Unit). The coprocessor is a low-power processor for processing data in a standby state. In some embodiments, the processor 21 can be integrated with a GPU (Graphics Processing Unit) that is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 21 can also include an AI (Artificial Intelligence) processor configured to process machine learning-related computing operations.

[0118] In addition, the memory 22, as a carrier for storing resources, can be a read-only memory, a random access memory, a magnetic disk, or an optical disk, etc. The resources stored thereon include an operating system 221, a computer program 222, and data 223, etc. The storage mode can be temporary storage or permanent storage.

[0119] The operating system 221 is used to manage and control each hardware device on the electronic device 20 and the computer program 222, so as to realize the operation and processing of the processor 21 on the mass data 223 in the memory 22, and can be Windows, Unix, Linux, etc. In addition to the computer program capable of completing the prediction trajectory correction method disclosed in any of the preceding embodiments executed by the electronic device 20, the computer program 222 can further include a computer program capable of completing other specific work. In addition to the data received by the electronic device from the external device, the data 223 can also include the data collected by the self input / output interface 25, etc.

[0120] Further, the embodiment of the present application further discloses a computer readable storage medium, the storage medium stores a computer program, and the computer program is loaded and executed by the processor to realize the prediction trajectory correction method steps disclosed in any of the preceding embodiments.

[0121] Each embodiment in the specification is described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the related parts can be referred to the method part.

[0122] Those skilled in the art can further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in the present application can be realized by electronic hardware, computer software or combination of the two. In order to clearly show the interchangeability of hardware and software, the composition and steps of each example have been described in the above description. Whether the functions are realized by hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0123] The steps of the method or algorithm described in combination with the embodiments disclosed in the present application can be directly implemented by hardware, software module executed by the processor, or combination of the two. The software module can be placed in random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, hard disk, removable disk, compact disc read-only memory (CD-ROM), or any other form of storage medium known in the art.

[0124] Finally, it needs to be pointed out that, in this article, the relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. Moreover, the term "includes", "contains" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the statement "includes a" does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0125] The above describes in detail the trajectory prediction correction method, device, equipment and storage medium provided by the present application. The principles and implementation manners of the present application are described by applying specific examples. The above description of the embodiments is only used to help understand the method of the present application and its core idea. Meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation manner and application range will be changed. In view of the above, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A method for correcting a predicted trajectory, characterized in that: include: When the target vehicle cuts into the own lane, obtaining a first predicted trajectory obtained by predicting the motion trajectory of the target vehicle, and determining a trajectory end position of the first predicted trajectory; Determining a target preceding vehicle closest to the ego vehicle in the ego vehicle lane, and obtaining a second predicted trajectory obtained by predicting the motion trajectory of the target preceding vehicle, so as to determine a target terminal position of the target preceding vehicle in the ego vehicle lane based on the second predicted trajectory; Determining whether the longitudinal distance between the trajectory endpoint and the vehicle position is greater than the longitudinal distance between the target endpoint and the vehicle position; If so, the first predicted trajectory is corrected based on a preset polynomial fitting method to obtain a corrected trajectory, and the vehicle's driving path is planned based on the corrected trajectory; otherwise, the vehicle's driving path is planned directly based on the first predicted trajectory.

2. The predicted trajectory correction method according to claim 1, characterized in that: Determining a target terminal position of the target preceding vehicle in the lane based on the second predicted trajectory includes: determining whether an end point of the second predicted trajectory is within the own lane; If the target vehicle is located in the own lane, determining the location of the trajectory end point of the second predicted trajectory as the target end point position of the target vehicle in the own lane; If the target vehicle is not located in the self-lane, the lane line of the self-lane is extended outward by a preset distance to obtain an extended lane line, and the target terminal position of the target front vehicle in the self-lane is determined based on the intersection position of the second predicted trajectory and the extended lane line.

3. The predicted trajectory correction method according to claim 2, characterized in that: The step of extending the lane line of the own lane outward by a preset distance to obtain an extended lane line includes: Determining a target lane line that intersects with the second predicted trajectory; the target lane line is either a left lane line or a right lane line of the ego lane; Determining a curve model corresponding to the target lane line, and determining a derivative equation corresponding to the curve model, so as to construct a corresponding tangent model based on each lane line point on the target lane line and the derivative equation; A corresponding tangent line is obtained based on the tangent line model, and each lane line point is translated by a preset distance in a direction perpendicular to the corresponding tangent line to obtain an extended lane line.

4. The predicted trajectory correction method according to claim 2, characterized in that: The second predicted trajectory includes a plurality of first discrete points, the extended lane line includes a plurality of second discrete points, and a process for determining an intersection position of the second predicted trajectory and the extended lane line includes: Construct a first straight line based on any two adjacent first discrete points; Constructing a second straight line based on any two adjacent second discrete points; If the target intersection point between the first straight line and the second straight line is located between any two adjacent first discrete points and between any two adjacent second discrete points, the target intersection point is used as the intersection position of the second predicted trajectory and the extended lane line.

5. The predicted trajectory correction method according to claim 4, characterized in that: The determining of a target terminal position of the target preceding vehicle in the own lane based on the intersection position of the second predicted trajectory and the extended lane line includes: Determine a target discrete point closer to the vehicle from any two adjacent first discrete points based on the intersection position of the second predicted trajectory and the extended lane line; Obtaining position coordinate information of the target discrete point in the ego vehicle coordinate system, and determining the vehicle angle of the target preceding vehicle at the target discrete point; wherein the ego vehicle coordinate system is a coordinate system established based on the center of the ego vehicle's rear axle; The rear center point position coordinates of the target preceding vehicle are calculated based on the position coordinate information, the vehicle angle, and the vehicle length information of the target preceding vehicle, and the position represented by the rear center point position coordinates is determined as the target terminal position of the target preceding vehicle in the self-lane.

6. The predicted trajectory correction method according to claim 5, characterized in that: Determining the vehicle angle of the target preceding vehicle at the target discrete point includes: determining a first slope of the first straight line corresponding to the intersection position; determining a second slope of the second straight line corresponding to the intersection position; A vehicle angle of the target preceding vehicle at the target discrete point is determined based on the first slope and the second slope.

7. The predicted trajectory correction method according to any one of claims 1 to 6, characterized in that: The step of correcting the first predicted trajectory based on a preset polynomial fitting method to obtain a corrected trajectory includes: The current position and current motion state of the target vehicle are respectively used as the starting planning position and the starting motion state; the motion state includes lateral velocity, longitudinal velocity, lateral acceleration and longitudinal acceleration; Determining a predicted motion state of the target preceding vehicle at the target terminal position based on the second predicted trajectory, and using the target terminal position and the predicted motion state as the terminal planning position and the terminal motion state, respectively; Determining a pre-built target model for the predicted motion trajectory of the target vehicle, and solving the target model based on the starting planned position, the starting motion state, the ending planned position, the ending motion state, and a preset travel time to obtain the predicted motion trajectory; The predicted motion trajectory is used as a corrected trajectory obtained by correcting the first predicted trajectory.

8. A predicted trajectory correction device, characterized in that: include: a first position determination module, configured to obtain a first predicted trajectory obtained by predicting the motion trajectory of the target vehicle when the target vehicle cuts into the own lane, and determine a trajectory end position of the first predicted trajectory; a second position determination module, configured to determine a target preceding vehicle closest to the ego vehicle in the ego vehicle lane, and obtain a second predicted trajectory obtained by predicting the motion trajectory of the target preceding vehicle, so as to determine a target terminal position of the target preceding vehicle in the ego vehicle lane based on the second predicted trajectory; a judgment module, configured to judge whether the longitudinal distance between the trajectory end position and the vehicle position is greater than the longitudinal distance between the target end position and the vehicle position; A trajectory correction module is used to, if so, correct the first predicted trajectory based on a preset polynomial fitting method to obtain a corrected trajectory, and plan the vehicle's driving path based on the corrected trajectory; otherwise, directly plan the vehicle's driving path based on the first predicted trajectory.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the predicted trajectory correction method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that Used to store a computer program; wherein, when the computer program is executed by a processor, the steps of the predicted trajectory correction method according to any one of claims 1 to 7 are implemented.

Images