A method for selecting a vehicle-following target based on the vehicle's driving trajectory

Abstract

The present invention discloses a method for selecting a following target based on the driving trajectory of the vehicle. The method combines lane lines, guardrails, traffic flow driving trajectory and high-precision maps to fit the driving trajectory of the vehicle; combines the driver's operating behavior to determine whether the driver intends to change lanes; if the driver has the intention to change lanes, the lane change trajectory line is determined according to the lane change intention; the interval of interest is determined according to the driving trajectory of the vehicle; the risk assessment of the following target in the interval of interest is performed; the following target selection; and the vehicle behavior decision. The present invention effectively overcomes the problems of misbraking and missed braking caused by poor quality of the far lane line or the jumping of the fitted lane line on the curve; and takes into account the lane changing intention of the driver of the vehicle, reduces the takeover rate during the lane changing process, and improves the driving experience; and based on the risk assessment and target selection strategy of the target TrackID and the driving trajectory, it effectively avoids the problem of discontinuous braking experience caused by the jumping of the target area ID.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a method for selecting a vehicle-following target based on a vehicle's driving trajectory. Background Art

[0002] In the current engineering application field, the target selection algorithm for autonomous driving following vehicles generally relies heavily on lane lines. These algorithms generally assume that the target vehicle is driving in the self-lane and the driver does not frequently turn the steering wheel. Algorithms that rely heavily on lane lines are suitable for L2 or L2+ scenarios where the lane line quality is good and the vehicle steering wheel is mainly controlled by the system. The lane lines in the industry are generally obtained by cubic curve fitting based on visual sampling points. Such curves have a natural weakness, that is, the quality of the far-end lane line is greatly affected by the actual curvature. Using such lane lines to select targets is prone to misbraking or missed braking. In addition, using the existing strategy, when the driver actively takes over the steering wheel, the target selected by the lane line is not easy to release in time, and the target of the following vehicle in the adjacent lane cannot be selected in time, resulting in the actual experience not meeting the driver's expectations. And when the driver takes over the steering wheel, the industry generally practices repeated confirmation of the target to reduce the misselection of the target, but there will also be the problem of the driver braking the target of the target lane late during the lane change process, resulting in a high takeover rate and poor braking experience during the early lane change process when only the longitudinal activation is activated. In addition, the driving assistance systems in the industry generally only brake the selected following targets. The timing of braking for following targets that may pose a risk of collision is generally calibrated to be late and the minimum deceleration is limited to a very conservative level. Summary of the invention

[0003] In view of the above-mentioned deficiencies of the prior art, the technical problem to be solved by the present invention is: how to provide a method for selecting a following target based on the driving trajectory of the vehicle, which improves the scene adaptability of the unmanned vehicle and takes into account the driver's riding experience and driving safety.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0005] A method for selecting a vehicle-following target based on a vehicle's driving trajectory comprises the following steps:

[0006] (1) Fit the vehicle's driving trajectory by combining lane lines, guardrails, traffic trajectories, and high-precision maps;

[0007] (2) Based on the driver’s operating behavior, determine whether the driver intends to change lanes;

[0008] (3) If the driver intends to change lanes, the lane-changing trajectory is determined based on the lane-changing intention;

[0009] (4) Determine the interval of interest based on the vehicle’s driving trajectory;

[0010] (5) Conduct risk assessment on the vehicle-following target in the area of ​​interest;

[0011] (6) Selection of vehicle-following targets;

[0012] (7) Vehicle behavior decision.

[0013] As an optimization, in step (1), when fitting the driving trajectory of the vehicle, the center line equation of the lane line is determined according to the lane line equation, which is used as the reference line 1 of the driving trajectory of the vehicle; the reference points output by the high-precision map are fitted into a cubic curve, which is used as the reference line 2 of the driving trajectory of the vehicle; the guardrail points are merged with the trajectory points of the vehicles around the vehicle, and the obtained cubic curve is used as the reference line 3 of the driving trajectory of the vehicle; the lane line equation coefficients beyond 2 times the time headway are corrected using reference lines 2 and 3; the corrected reference line is used as the reference line 0 of the driving trajectory of the vehicle, and its coefficients A0, A1, A2, and A3 are given.

[0014] As an optimization, in step (2), when judging whether the driver intends to change lanes, the continuous moving distance S1 of the vehicle relative to the reference line 0 is calculated; the lateral moving speed V1 of the vehicle is calculated based on the continuous moving distance; and the distance D1 of the vehicle relative to the lane line is calculated; if S1 is greater than the calibration value, and the absolute value of V1 is greater than the calibration value, and D1 is less than the calibration value, it is considered that the driver intends to change lanes.

[0015] As an optimization, in step (3), when the lane changing trajectory is determined according to the driver's lane changing intention, the coefficients of the quadratic term and the cubic term of the equation of the lane changing trajectory are A2 and A3 respectively, and the coefficients of the zero-order term and the linear term are both 0.

[0016] As an optimization, in step (4), when determining the interval of interest based on the vehicle's driving trajectory, the reference line of the vehicle's driving trajectory is translated to both sides of the vehicle by half a lane width to obtain the original pipeline No. 0 in the area of ​​interest; the reference line on the left side of the original pipeline No. 0 is translated to the left by one lane width to obtain the original pipeline No. 1 in the area of ​​interest; the reference line on the right side of the pipeline No. 0 is translated to the right by one lane width to obtain the original pipeline No. 2 in the area of ​​interest; the interval consisting of the original pipeline No. 1 and the original pipeline No. 2 within 3 times the time interval and the original pipeline No. 0 within 5 times the time interval is set as the interval of interest.

[0017] As an optimization, in step (5), when risk assessment is performed on the following target in the interval of interest, the following target appearing in the interval of interest is tracked and the historical trajectory of the following target is recorded; based on the historical trajectory of the following target, the approach trend of the following target and the corresponding lateral speed are determined; based on the lateral speed of the following target, the time TCL required for the following target to reach the driving trajectory of the host vehicle is determined; based on the longitudinal distance of the following target and the longitudinal speed difference of the host vehicle, the longitudinal collision time TTC of the following target is determined; based on the current position and lateral and longitudinal speeds of the following target, the position P0 of the following target after 1 second is predicted; based on P0, the intrusion amount of the following target into the No. 0 driving channel of the host vehicle is determined, and the cutting probability P of the following target is determined based on the intrusion amount.

[0018] As an optimization, in step (6), when selecting a following target, if the following target has a tendency to approach and its longitudinal speed is less than the longitudinal speed of the vehicle, if the time TCL required for the following target to reach the driving trajectory of the vehicle minus the longitudinal collision time TTC of the following target is less than a fixed value and the cut-in probability P is greater than a calibrated value, then the following target is considered to be a potential collision target; from the potential collision targets, the one with the smallest longitudinal collision time TTC is selected as the following target.

[0019] As an optimization, in step (7), when making the behavior decision of the vehicle, the PID algorithm is used to plan the required deceleration a1 according to the longitudinal relative distance, longitudinal relative speed and longitudinal acceleration of the following target and the vehicle; the PID algorithm is used to plan the required deceleration a2 according to the cut-in probability of the following target, as well as the longitudinal relative distance, longitudinal relative speed and longitudinal acceleration of the following target and the vehicle; a2 is the minimum deceleration planned by PID for the following targets in all areas of interest; at each specific moment, the deceleration of the vehicle is decided to be min(a1,a2).

[0020] In summary, the beneficial effects of the present invention are:

[0021] (1) The expected driving trajectory of the vehicle is obtained by integrating lane lines, guardrails, vehicle flow trajectories, and high-precision map information, effectively overcoming the problems of misbraking and missed braking caused by poor quality of the far lane lines or jittery lane lines fitted on curves;

[0022] (2) The system takes into account the driver's lane-changing intention. When the driver actively changes lanes, the system switches the driving trajectory in advance, selects the following vehicle target earlier, reduces the takeover rate during the lane-changing process, and improves the driving experience.

[0023] (3) The risk assessment and target selection strategy based on the target TrackID and driving trajectory can effectively avoid the problem of discontinuous braking experience caused by the ID jump of the target area. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to make the purpose, technical solution and advantages of the invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0025] Figure 1 It is a schematic diagram of the process of the present invention;

[0026] Figure 2 It is a schematic diagram of the vehicle driving trajectory fitting process in the present invention. DETAILED DESCRIPTION

[0027] The present invention will be further described in detail below in conjunction with the accompanying drawings.

[0028] according to Figure 1 and Figure 2 As shown, in this specific implementation, a method for selecting a vehicle-following target based on the vehicle's driving trajectory includes the following steps:

[0029] (1) Fit the vehicle's driving trajectory by combining lane lines, guardrails, traffic trajectories, and high-precision maps;

[0030] (2) Based on the driver’s operating behavior, determine whether the driver intends to change lanes;

[0031] (3) If the driver intends to change lanes, the lane-changing trajectory is determined based on the lane-changing intention;

[0032] (4) Determine the interval of interest based on the vehicle’s driving trajectory;

[0033] (5) Conduct risk assessment on the vehicle-following target in the area of ​​interest;

[0034] (6) Selection of vehicle-following targets;

[0035] (7) Vehicle behavior decision.

[0036] In this specific implementation, in step (1), when fitting the driving trajectory of the vehicle, the center line equation of the lane line is determined according to the lane line equation as the reference line 1 of the driving trajectory of the vehicle; the reference points output by the high-precision map are fitted into a cubic curve as the reference line 2 of the driving trajectory of the vehicle; the guardrail points are merged with the trajectory points of the vehicles around the vehicle, and the obtained cubic curve is used as the reference line 3 of the driving trajectory of the vehicle; the lane line equation coefficients beyond 2 times the time headway are corrected using reference lines 2 and 3; the corrected reference line is used as the reference line 0 of the driving trajectory of the vehicle, and its coefficients A0, A1, A2, and A3 are given.

[0037] In this specific implementation, in step (2), when determining whether the driver has the intention to change lanes, the continuous moving distance S1 of the vehicle relative to the reference line 0 is calculated; the lateral moving speed V1 of the vehicle is calculated based on the continuous moving distance; and the distance D1 of the vehicle relative to the lane line is calculated; if S1 is greater than a calibration value, and the absolute value of V1 is greater than the calibration value, and D1 is less than the calibration value, it is considered that the driver has the intention to change lanes.

[0038] In this specific implementation, in step (3), when the lane changing trajectory is determined according to the driver's lane changing intention, the quadratic term and cubic term coefficients of the lane changing trajectory equation are A2 and A3 respectively, and the zero-order term coefficient and the linear term coefficient are both 0.

[0039] In this specific implementation, in step (4), when determining the interval of interest based on the vehicle's driving trajectory, the reference line of the vehicle's driving trajectory is translated to both sides of the vehicle by half a lane width to obtain the original pipeline No. 0 in the area of ​​interest; the reference line on the left side of the original pipeline No. 0 is translated to the left by one lane width to obtain the original pipeline No. 1 in the area of ​​interest; the reference line on the right side of the pipeline No. 0 is translated to the right by one lane width to obtain the original pipeline No. 2 in the area of ​​interest; and the interval consisting of the original pipeline No. 1 and the original pipeline No. 2 within three times the time interval and the original pipeline No. 0 within five times the time interval is set as the interval of interest.

[0040] In this specific implementation, in step (5), when risk assessment is performed on the following target in the interval of interest, the following target appearing in the interval of interest is tracked and the historical trajectory of the following target is recorded; based on the historical trajectory of the following target, the approach trend of the following target and the corresponding lateral speed are determined; based on the lateral speed of the following target, the time TCL required for the following target to reach the driving trajectory of the host vehicle is determined; based on the longitudinal distance of the following target and the longitudinal speed difference of the host vehicle, the longitudinal collision time TTC of the following target is determined; based on the current position and the lateral and longitudinal speeds of the following target, the position P0 of the following target after 1 second is predicted; based on P0, the intrusion amount of the following target into the No. 0 driving channel of the host vehicle is determined, and the cutting-in probability P of the following target is determined based on the intrusion amount.

[0041] In this specific implementation, in step (6), when selecting a following target, if the following target has a tendency to approach and its longitudinal speed is less than the longitudinal speed of the vehicle, if the time TCL required for the following target to reach the driving trajectory of the vehicle minus the longitudinal collision time TTC of the following target is less than a fixed value and the cut-in probability P is greater than a calibrated value, then the following target is considered to be a potential collision target; from the potential collision targets, the one with the smallest longitudinal collision time TTC is selected as the following target.

[0042] In this specific implementation, in step (7), when making a decision on the behavior of the vehicle, the PID algorithm is used to plan the required deceleration a1 according to the longitudinal relative distance, longitudinal relative speed and longitudinal acceleration of the following target and the vehicle; the PID algorithm is used to plan the required deceleration a2 according to the cut-in probability of the following target, as well as the longitudinal relative distance, longitudinal relative speed and longitudinal acceleration of the following target and the vehicle; a2 is the minimum deceleration planned by PID for the following targets in all areas of interest; at each specific moment, the deceleration of the vehicle is decided to be min(a1, a2).

[0043] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described with reference to the preferred embodiments of the present invention, it should be understood by those skilled in the art that various changes may be made in form and details without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A method for selecting a vehicle-following target based on the vehicle's driving trajectory, characterized in that: The following steps are involved: (1) Fit the vehicle's driving trajectory by combining lane lines, guardrails, traffic trajectories, and high-precision maps; (2) Based on the driver’s operating behavior, determine whether the driver intends to change lanes; (3) If the driver intends to change lanes, the lane-changing trajectory is determined based on the lane-changing intention; (4) Determine the interval of interest based on the vehicle's driving trajectory. Shift the reference line of the vehicle's driving trajectory to both sides of the vehicle by half a lane width to obtain the original pipeline No. 0 in the area of ​​interest. Shift the reference line on the left side of the original pipeline No. 0 to the left by one lane width to obtain the original pipeline No. 1 in the area of ​​interest. Shift the reference line on the right side of the pipeline No. 0 to the right by one lane width to obtain the original pipeline No. 2 in the area of ​​interest. Set the interval consisting of the original pipeline No. 1 and the original pipeline No. 2 within three times the time interval and the original pipeline No. 0 within five times the time interval as the interval of interest. (5) Conduct risk assessment on the vehicle-following target in the area of ​​interest; (6) Selection of vehicle-following targets; (7) Vehicle behavior decision.

2. A method for selecting a vehicle target based on the vehicle's driving trajectory according to claim 1, characterized in that: In step (1), when fitting the driving trajectory of the vehicle, the center line equation of the lane line is determined according to the lane line equation, which is used as the reference line 1 of the driving trajectory of the vehicle; the reference points output by the high-precision map are fitted into a cubic curve, which is used as the reference line 2 of the driving trajectory of the vehicle; the guardrail points are merged with the trajectory points of the vehicles around the vehicle, and the obtained cubic curve is used as the reference line 3 of the driving trajectory of the vehicle; the lane line equation coefficients beyond 2 times the time headway are corrected using reference lines 2 and 3; the corrected reference line is used as the reference line 0 of the driving trajectory of the vehicle, and its coefficients A0, A1, A2, and A3 are given.

3. The method for selecting a vehicle target based on the vehicle's driving trajectory according to claim 2, characterized in that: In step (2), when judging whether the driver intends to change lanes, the continuous moving distance S1 of the vehicle relative to the reference line 0 is calculated; the lateral moving speed V1 of the vehicle is calculated based on the continuous moving distance; and the distance D1 of the vehicle relative to the lane line is calculated; if S1 is greater than the calibration value, and the absolute value of V1 is greater than the calibration value, and D1 is less than the calibration value, it is considered that the driver intends to change lanes.

4. The method for selecting a vehicle target based on the vehicle's driving trajectory according to claim 1, characterized in that: In step (3), when the lane changing trajectory is determined according to the driver's lane changing intention, the coefficients of the quadratic term and the cubic term of the equation of the lane changing trajectory are A2 and A3 respectively, and the coefficients of the zero-order term and the linear term are both 0.

5. The method for selecting a vehicle-following target based on the vehicle's driving trajectory according to claim 1, characterized in that: In step (5), when performing risk assessment on the following target in the interval of interest, the following target appearing in the interval of interest is tracked and the historical trajectory of the following target is recorded; based on the historical trajectory of the following target, the approach trend of the following target and the corresponding lateral speed are determined; based on the lateral speed of the following target, the time TCL required for the following target to reach the driving trajectory of the host vehicle is determined; based on the longitudinal distance of the following target and the longitudinal speed difference of the host vehicle, the longitudinal collision time TTC of the following target is determined; based on the current position and lateral and longitudinal speeds of the following target, the position P0 of the following target 1 second later is predicted; The intrusion amount of the following target into the No. 0 driving channel of the vehicle is determined based on P0, and the cutting-in probability P of the following target is determined based on the intrusion amount.

6. A method for selecting a vehicle-following target based on the vehicle's driving trajectory according to claim 5, characterized in that: In step (6), when selecting a following target, if the following target has a tendency to approach and its longitudinal speed is less than the longitudinal speed of the vehicle, if the time TCL required for the following target to reach the driving trajectory of the vehicle minus the longitudinal collision time TTC of the following target is less than a fixed value and the cut-in probability P is greater than a calibrated value, then the following target is considered to be a potential collision target; from the potential collision targets, the one with the smallest longitudinal collision time TTC is selected as the following target.

7. The method for selecting a vehicle-following target based on the vehicle's driving trajectory according to claim 1, characterized in that: In step (7), when making the behavior decision of the vehicle, the PID algorithm is used to plan the required deceleration a1 according to the longitudinal relative distance, longitudinal relative speed and longitudinal acceleration of the following target and the vehicle; the PID algorithm is used to plan the required deceleration a2 according to the cut-in probability of the following target, as well as the longitudinal relative distance, longitudinal relative speed and longitudinal acceleration of the following target and the vehicle; a2 is the minimum deceleration planned by PID for the following targets in all the areas of interest; at each specific moment, the deceleration of the vehicle is decided to be min(a1,a2).

Images