Target screening optimization algorithm during lane changing of own vehicle

Abstract

The invention discloses a target screening optimization algorithm during lane changing of an own vehicle, which comprises the following steps of: firstly, activating an ACC longitudinal control function of the own vehicle, acquiring own vehicle parameters, front target vehicle parameters, own lane parameters, left lane parameters and right lane parameters, and calculating a center line equation of the own lane, the left lane and the right lane; judging whether to change a lane or not and which lane changes to the left and right; when a lane to be switched is determined and entered, the current lane center line of the vehicle is calculated; when the vehicle just enters the lane to be switched, no lane change judgment is carried out within 3s; and after the vehicle enters the lane to be switched, no lane changing judgment is carried out within 3 seconds, and the lane changing work of the vehicle is completed. The target screening optimization algorithm for the self-vehicle during lane changing has the advantages that the selection of the vehicle following target is more reliable and effective, ACC control is smoother, and the safety and reliability of lane changing operation are improved.

Description

Technical Field

[0001] The present invention relates to a vehicle control technology for intelligent vehicle driving, and in particular to a target screening optimization algorithm for longitudinal control of a vehicle when the vehicle changes lanes. Background Art

[0002] With the rapid development of automobile technology, the level of automobile intelligence is getting higher and higher. At present, autonomous driving can be divided into six levels from L0 to L5: L0 is a completely manual driving vehicle. Although it may have auxiliary protection systems such as brake assist and blind spot monitoring, the driver needs to control it throughout the process; L1 is an assisted driving level that can assist the driver to complete simple repetitive operations such as adaptive cruise control speed, lane keeping assist control direction, etc. The driver still participates in the whole process; L2 is a partially autonomous driving vehicle that can control the speed and direction at the same time, such as adaptive cruise with lane keeping, automatic parking and automatic lane change, but the driver needs to monitor the environment and take over at any time; L3 is conditional autonomous driving. Under limited conditions, the driver can give up control, but if there is a problem with the system, the driver needs to take over and the takeover time is short; L4 is highly automated driving that can complete driving tasks and monitor the environment under specific environments and conditions without driver intervention; L5 is fully automated and can complete all driving tasks under any conditions.

[0003] With the improvement of the level of autonomous driving, the assisted driving functions of automobiles are becoming more and more complete. Among the assisted driving functions of automobiles, the longitudinal control of the vehicle mainly assists the driver in controlling the throttle and brake of the vehicle, improves the longitudinal comfort of the vehicle, reduces the operating burden of the driver, and can follow the vehicle in front to drive stably and improve the safety performance of the vehicle. The basic principle of longitudinal control of the vehicle: the size of the longitudinal driving distance depends on the driving speed, the driver's concentration, driving skills, energy level, the working performance of the vehicle's braking device, and operating conditions (road, climate).

[0004] The longitudinal control in the intelligent driving system mainly has the function of adaptive cruise control (ACC). Intelligent cars equipped with adaptive cruise control systems use radar and computers to identify whether the approaching vehicle is a bicycle, a car or a pedestrian, and control the vehicle's driving state according to the road conditions, completely or partially replacing the driver's operation. When the driver turns on the ACC function, the auxiliary driving controller controls the vehicle's brakes and throttle according to the cruising speed set by the driver and the information of the target vehicle ahead detected by the camera or millimeter-wave radar, and controls the longitudinal speed of the vehicle according to the cruising speed set by the driver and the information of the target vehicle ahead. When there is no target vehicle ahead, ACC controls the vehicle to drive at the cruising speed set by the driver. When a target vehicle ahead is detected, and the cruising speed set by the driver is less than the target vehicle ahead, ACC controls the vehicle to drive at the cruising speed set by the driver. When the cruising speed set by the driver is greater than the speed of the target vehicle ahead, ACC controls the vehicle to drive at the speed of the target vehicle ahead.

[0005] ACC longitudinal control, in order to make the vehicle more comfortable and safe when driving longitudinally, ACC generally uses a state machine for control. The state machine of ACC in its algorithm is as follows Figure 5 shown.

[0006] like Figure 5 , the main states of ACC control are introduced as follows: Override means that when the ACC function is turned on, the driver operates the accelerator pedal, and the depth of the accelerator pedal exceeds the accelerator pedal depth value requested by ACC, and the vehicle responds to the driver's accelerator pedal operation; ON state means that when there is a target vehicle in front of the vehicle and the speed set by the driver is greater than the target speed in front, ACC controls the vehicle to travel at the speed in front; SpeedControl state means that when there is no target vehicle in front, or there is a target vehicle in front but the speed of the target vehicle in front is greater than the cruising speed set by the driver, ACC controls the vehicle to travel at the set cruising speed; tempStop state means that ACC controls the vehicle to follow the target in front to a stop state, and when the vehicle in front starts again, ACC controls the vehicle to start again and follow the vehicle, and the duration of tempStop is generally about 3s. If this time is exceeded, ACC will enter the Stop state; Stop state means that ACC controls the vehicle to stop following the vehicle in front. After the vehicle leaves, the driver can step on the accelerator or operate the ACC system switch, and the ACC system controls the vehicle to follow the vehicle in front again.

[0007] When the ACC function is turned on, the target screening algorithm will screen out the target of the vehicle ahead based on the target information ahead, lane line information and the movement of the vehicle detected by the sensor. The accuracy of target screening has a great impact on the control of ACC: the wrong selection of the target ahead often causes unnecessary deceleration of ACC, and the target selection too late will cause the vehicle to have a relatively large deceleration after the target is selected. The target screening algorithm is also critical when the vehicle changes lanes. When the driver is preparing to change lanes, if the target ahead of the lane is released too late, and the distance to the target ahead of the lane is relatively close, this will often cause ACC to control the vehicle to accelerate too slowly, resulting in a long lane change time. When the vehicle is changing lanes, the wrong selection of the adjacent lane target or the adjacent adjacent lane target often causes the vehicle to decelerate or accelerate unexpectedly during the lane change process, which often affects the driver's experience of ACC control. During the lane change process, if the steering wheel angle is relatively large, after the vehicle changes lanes, the steering wheel is straightened, and the target in the lane before the lane change is often selected, which is often not the vehicle that the driver wants to select.

[0008] In the prior art, there are two options that are often selected by the target screening algorithm when a conventional vehicle changes lanes.

[0009] Solution 1: When changing lanes or turning, the lane line is used as the reference, and the vehicle selects the nearest target in the lane where the vehicle is located; when the vehicle starts changing lanes or turning, the front of the vehicle has not crossed the lane, so the target selected is still the nearest target in the lane; when the front of the vehicle crosses the adjacent lane, the lane line will also switch at this time, and the target selection will be changed to the nearest target in the adjacent lane. In Solution 1, the lane line is used as the reference, because the camera needs a certain threshold after the middle of the front of the vehicle crosses the lane line before switching the lane line. As a result, the target in the lane is often released too late, affecting the continuity of ACC control. And if the distance between the vehicle and the vehicle in front is relatively close, when the vehicle starts turning, since ACC has not released the front target, the ACC algorithm will not be able to accelerate or accelerate slowly because the front target is close, and it will continue until the center line of the front of the vehicle crosses the lane line before releasing the target in the lane, and ACC will control the vehicle to accelerate faster. Therefore, Solution 1 will cause the vehicle to change lanes for a long time, affecting the control effect of ACC.

[0010] Solution 2: When changing lanes or turning, the vehicle's motion state is determined. Once the vehicle is detected to be changing lanes or turning, the following target in the current lane is abandoned, and the vehicle's front direction is used to select the following target in front of the vehicle. In Solution 2, the target in front of the vehicle is selected by the vehicle's front direction. Since the vehicle's front direction will change during the lane change process. If the target is selected completely according to the vehicle's front direction, the following target will often switch frequently during the lane change or turning process, and the target in front of the vehicle is not the following target that the driver wants during the turning process or lane change process. During the lane change process, if there is a target in the adjacent lane, the vehicle's front direction will also select the following target when it passes by. This misselection will affect the control effect of ACC. During the turning process, the vehicle's front direction will change greatly, and the surrounding vehicles will be relatively close during the turning process. If the front target is selected completely according to the direction, it will often lead to the misselection of the surrounding relatively close targets, causing ACC to control the vehicle to decelerate more greatly during the turning process or affect the acceleration effect. During the turning process, the following target will also switch frequently, affecting the continuity of longitudinal control. Summary of the invention

[0011] The present invention aims to avoid the deficiencies in the above-mentioned prior art and to provide a target screening optimization algorithm for a vehicle when changing lanes, so that the driver can select the following target more reliably and effectively during the lane changing process and make ACC control smoother.

[0012] The present invention adopts the following technical solutions to solve the technical problems.

[0013] The present invention provides a target screening optimization algorithm for a vehicle when changing lanes, comprising the following steps: Step 1: Activate the ACC longitudinal control function of the ego vehicle, obtain the ego vehicle parameters through the ego vehicle control system, obtain the parameters of the target vehicle in front through the ego vehicle camera, and obtain the ego lane parameters, left lane parameters, and right lane parameters through the ego vehicle camera; Step 2: Calculate the centerline equation of the self-lane, the centerline equation of the left lane and the centerline equation of the right lane according to the parameters of the self-lane, the parameters of the left lane and the parameters of the right lane; Step 3: According to the vehicle parameters, the centerline equation of the lane, the centerline equation of the left lane and the centerline equation of the right lane, determine whether the vehicle needs to change lanes, change lanes to the left lane or change lanes to the right lane; Step 4: When determining and entering the lane to be switched, calculate the current lane centerline of the vehicle; Step 5: When the vehicle just enters the lane to be switched, no lane change judgment will be performed within 3 seconds; Step 6: After the vehicle enters the lane to be switched, no lane change judgment will be made within 3 seconds, and the lane change of the vehicle is completed.

[0014] The structural features of the target screening optimization algorithm of the vehicle when changing lanes of the present invention are also as follows: Furthermore, in step 1, the vehicle parameters include the lateral speed, longitudinal speed, lateral acceleration, longitudinal acceleration and steering wheel angle of the vehicle.

[0015] Furthermore, in step 1, the front target vehicle parameters include a lateral speed, a longitudinal speed, a lateral acceleration, a longitudinal acceleration, a lateral position and a longitudinal position of the front vehicle.

[0016] Furthermore, in step 1, the self-lane parameters include a cubic polynomial coefficient of a left lane line, a cubic polynomial coefficient of a right lane line, a lane line credibility, and a lane line length of the self-lane.

[0017] Furthermore, in step 1, the left lane parameters include a cubic polynomial coefficient of a left lane line of the left lane, a cubic polynomial coefficient of a right lane line, lane line credibility and lane line length.

[0018] Furthermore, in step 1, the right lane parameters include the cubic polynomial coefficients of the left lane line of the right lane, the cubic polynomial coefficients of the right lane line, the lane line credibility and the lane line length.

[0019] Furthermore, in step 2, the centerline equation and lane width of the self-lane are obtained through the self-lane camera. ; The lane centerline equation is shown in the following formula (1); (1); In formula (1), y0 is the dependent variable of the lane centerline equation, x is the independent variable, a0 is the constant term of the lane centerline equation, and a1 to a3 are the coefficients of the linear to cubic terms of the lane centerline equation, respectively.

[0020] Furthermore, in step 3, when judging whether the vehicle needs to change lanes, change lanes to the left lane or change lanes to the right lane, when the vehicle is in the initial state, the distance △ between the center position of the vehicle at the starting position and the lane line y 0 <0, the vehicle at the preview point is within a certain range to the left of the lane line, and the distance between the vehicle position and the lane line is △ y preview <-1.0, and the absolute value of the steering wheel angle A ego If the angle is less than 30° and the time from exiting the last lane change to the left or right exceeds 3 seconds, the vehicle is considered to be changing lanes on the left.

[0021] Furthermore, in step 3, when judging whether the vehicle needs to change lanes, change lanes to the left lane or change lanes to the right lane, when the vehicle is in the initial state, the distance △ between the center position of the vehicle at the starting position and the lane line y 0 >0, and at the preview point, the vehicle is within a certain range to the right of the lane line, and the distance between the vehicle position and the lane line is △ y preview >1.0, and the steering wheel angle is the absolute value A ego If the angle is less than 30° and the time from exiting the last lane change to the left or right exceeds 3 seconds, the vehicle is considered to be changing lanes to the right.

[0022] Furthermore, in step 4, the current lane centerline of the vehicle is expressed by a fifth-order polynomial using the following formula (9): (9) In formula (9), C0 is the constant term of the centerline equation, and C1 to C5 are the coefficients of the first to fifth terms of the centerline equation, respectively.

[0023] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention discloses a target screening optimization algorithm when a self-vehicle changes lanes. Firstly, the ACC longitudinal control function of the self-vehicle is activated, the self-vehicle parameters, the parameters of the front target vehicle, the self-lane parameters, the left lane parameters and the right lane parameters are obtained, the center line equations of the three lanes, namely the self-lane, the left lane and the right lane, are calculated, and it is judged whether to change lanes and to which lane to change lanes; when the lane to be switched is determined and entered, the current lane center line of the self-vehicle is calculated; when the self-vehicle just enters the lane to be switched, no lane change judgment is performed within 3 seconds; after the self-vehicle enters the lane to be switched, no lane change judgment is performed within 3 seconds, and the lane change work of the self-vehicle is completed.

[0024] The present invention enables the driver to select a target more reliably and effectively during the lane changing process, makes the ACC control during the lane changing process more intelligent and smooth, and makes the ACC control more in line with the driver's actual intention and more safe and reliable.

[0025] The target screening optimization algorithm of the self-vehicle when changing lanes of the present invention has the advantages of more reliable and effective selection of the following vehicle target, smoother ACC control, and improved safety and reliability of lane changing operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 The present invention is a flow chart of the target screening optimization algorithm of the vehicle when changing lanes.

[0027] Figure 2It is a schematic diagram of the preview point of the ego vehicle from the current starting point of the ego lane to the target lane in the target screening optimization algorithm of the ego vehicle when changing lanes of the present invention.

[0028] Figure 3 This is a schematic diagram of the lane changing at the first time point T1 (the front of the vehicle just enters the lane to be switched) during the lane changing process of the target screening optimization algorithm of the present invention.

[0029] Figure 4 This is a schematic diagram of the lane changing at the second time point T2 (after the vehicle enters the target lane to be switched) in the lane changing process of the target screening optimization algorithm when the vehicle changes lanes according to the present invention.

[0030] Figure 5 It is a structural block diagram of the state machine of ACC longitudinal control in the prior art.

[0031] The present invention will be further described below through specific implementation modes in conjunction with the accompanying drawings. DETAILED DESCRIPTION

[0032] See also Figure 1 to Figure 4 The present invention provides a target screening optimization algorithm for a self-vehicle when changing lanes, comprising the following steps: Step 1: Activate the ACC longitudinal control function of the ego vehicle, obtain the ego vehicle parameters through the ego vehicle control system, obtain the parameters of the target vehicle in front through the ego vehicle camera, and obtain the ego lane parameters, left lane parameters, and right lane parameters through the ego vehicle camera; The ACC longitudinal control function is activated to obtain relevant input parameters of the vehicle, the target vehicle in front and the camera; Step 2: Calculate the centerline equation of the self-lane, the centerline equation of the left lane and the centerline equation of the right lane according to the parameters of the self-lane, the parameters of the left lane and the parameters of the right lane; Based on the various parameters and information collected in step 1 and the image obtained by the camera, the ego vehicle control system calculates the center line equations of the ego lane and the left and right lanes of the ego lane through internal built-in software and hardware; Step 3: According to the vehicle parameters, the centerline equation of the lane, the centerline equation of the left lane and the centerline equation of the right lane, determine whether the vehicle needs to change lanes, change lanes to the left lane or change lanes to the right lane; Based on the information in step 2 and the vehicle information parameters in step 1, calculate and determine the motion state of the vehicle: left lane change or right lane change, and determine whether to change lanes and the direction of lane change; Step 4: When determining and entering the lane to be switched, calculate the current lane centerline of the vehicle; According to the information in step 3, when entering the left lane change or the right lane change, a quintic polynomial is fitted using the current motion parameters of the vehicle and the parameters of the center line of the left or right lane. The quintic polynomial is used as the moving curve of the switching process from the current lane center line of the vehicle to the lane center line of the lane to be switched; in fact, the moving curve is the moving curve of the projection point of the rear axle center point of the vehicle on the ground during the lane change process; the width of the lane line of the current lane of the vehicle adopts the width of the vehicle body. At the same time, the following target of the lane where the vehicle was originally located is released, and the updated lane line (the lane line determined according to the current lane center line of the vehicle; during the lane change, the lane line of the vehicle is no longer the lane line of the actual road before, but a virtual lane line that changes with the position of the vehicle; is used to perform target screening and select the following target in front; the following target represents the target vehicle in front to be followed after ACC is activated; When calculating the current lane center line of the vehicle, in addition to using a quintic polynomial, a cubic polynomial fitting method can also be used to fit the lane line using the motion parameters of the vehicle's starting point and a preview point (in the present invention, the preview point refers to the projection point of the rear axle center point on the ground after the vehicle finally completes the lane change process; the preview lateral position and preview longitudinal position hereinafter are the position coordinates of the preview point in the longitudinal and lateral directions.)

[0033] Step 5: When the vehicle just enters the lane to be switched, no lane change judgment will be performed within 3 seconds; According to the information in step 4, when the ego vehicle just enters the lane to be changed, no lane change judgment is performed within 3 seconds to prevent the ego vehicle from being judged to continue changing lanes because the ego vehicle has not yet adjusted its state, resulting in misselection of the target; When the vehicle just enters the lane to be changed, in order to prevent the misjudgment of vehicles continuously entering the lane change, the time delay method in this algorithm can be used. In addition, when the vehicle just enters the adjacent lane, a certain compensation can be made for the threshold of re-entering the lane change, making the conditions for the vehicle to enter the lane change more stringent, and preventing the misjudgment of vehicles continuously entering the lane change state.

[0034] Step 6: After the vehicle enters the lane to be switched, no lane change judgment will be made within 3 seconds, and the lane change of the vehicle is completed.

[0035] Based on the information in step 5, the vehicle will no longer make lane change judgments within 3 seconds after entering the adjacent lane to prevent the vehicle from being misjudged as changing lanes when the steering wheel is straightened in the centerline area of ​​the lane, resulting in misselection of the target.

[0036] In specific implementation, when returning to the corresponding adjacent lane, in order to prevent the vehicle from misselecting lane changes, the time delay method in this algorithm can be used. In addition, when the vehicle just enters the adjacent lane, a certain compensation can be made for the threshold for re-entering the lane change, making the conditions for the vehicle to enter the lane change more stringent, and preventing the misjudgment of the vehicle continuously entering the lane change state.

[0037] The target screening optimization algorithm of the self-vehicle when changing lanes of the present invention, in order to make the self-vehicle screening more accurate when changing lanes after the ACC function is turned on, firstly judges that the self-vehicle wants to change lanes according to the movement of the self-vehicle, then virtualizes the lane line of the lane change according to the information such as the movement parameters of the self-vehicle, and then uses the lane line of the lane change to screen the target of the following vehicle. And in the process of changing lanes, the movement state of the self-vehicle is tracked, and the lane change is judged after the self-vehicle has just changed lanes and after a certain time (for example, 3 seconds) is met, so as to prevent the vehicle from mistakenly selecting to continue changing lanes or changing back to the original lane during the process of changing lanes because the angle of the steering wheel is too large, resulting in the vehicle mistakenly selecting a vehicle in a non-target lane during the process of changing lanes, thereby making the lane changing operation more stable and reliable. The present invention makes the driver's choice of target more reliable and effective during the process of changing lanes, makes the ACC control during the process of changing lanes more intelligent and smooth, makes the control of ACC more in line with the driver's actual intention, and is more safe and reliable.

[0038] During specific implementation, in step 1, the vehicle parameters include the lateral speed, longitudinal speed, lateral acceleration, longitudinal acceleration and steering wheel angle of the vehicle.

[0039] During specific implementation, in step 1, the parameters of the front target vehicle include the lateral speed, longitudinal speed, lateral acceleration, longitudinal acceleration, lateral position and longitudinal position of the front vehicle.

[0040] In specific implementation, in step 1, the self-lane parameters include the cubic polynomial coefficient of the left lane line of the self-lane, the cubic polynomial coefficient of the right lane line, the lane line credibility and the lane line length.

[0041] During specific implementation, in step 1, the left lane parameters include the cubic polynomial coefficient of the left lane line of the left lane, the cubic polynomial coefficient of the right lane line, the lane line credibility and the lane line length.

[0042] During specific implementation, in step 1, the right lane parameters include the cubic polynomial coefficients of the left lane line of the right lane, the cubic polynomial coefficients of the right lane line, the lane line credibility and the lane line length.

[0043] The lane line credibility is used to represent whether the lane line is in an available state. Whether the lane line can be used as a real lane line is related to whether the lane marking line is blurred, whether there are other similar lane marking lines near the lane line, etc.

[0044] In the specific implementation, in step 2, the centerline equation and lane width of the self-lane are obtained through the self-vehicle camera. ; The lane centerline equation is shown in the following formula (1); (1); In formula (1), y0 is the dependent variable of the lane centerline equation, x is the independent variable, a0 is the constant term of the lane centerline equation, and a1 to a3 are the coefficients of the linear to cubic terms of the lane centerline equation, respectively.

[0045] In the specific implementation, in the step 3, when judging whether the vehicle needs to change lanes, change lanes to the left lane or change lanes to the right lane, when the vehicle is in the initial state, the distance △ between the center position of the vehicle at the starting position and the lane line y 0 <0, the vehicle at the preview point is within a certain range to the left of the lane line, and the distance between the vehicle position and the lane line is △ y preview <-1.0, and the absolute value of the steering wheel angle A ego If the angle is less than 30° and the time from exiting the last lane change to the left or right exceeds 3 seconds, the vehicle is considered to be changing lanes on the left.

[0046] In the specific implementation, in the step 3, when judging whether the vehicle needs to change lanes, change lanes to the left lane or change lanes to the right lane, when the vehicle is in the initial state, the distance △ between the center position of the vehicle at the starting position and the lane line y 0 >0, and at the preview point, the vehicle is within a certain range to the right of the lane line, and the distance between the vehicle position and the lane line is △ y preview >1.0, and the steering wheel angle is the absolute value A ego If the angle is less than 30° and the time from exiting the last lane change to the left or right exceeds 3 seconds, the vehicle is considered to be changing lanes to the right.

[0047] In the specific implementation, in step 4, the current lane centerline of the vehicle is expressed by a quintic polynomial using the following formula (9); (9) In formula (9), C0 is the constant term of the centerline equation, and C1 to C5 are the coefficients of the first to fifth terms of the centerline equation, respectively.

[0048] In order to make the selection of the following vehicle more accurate when the ego vehicle is changing lanes, in the present invention, it is necessary to first determine whether the vehicle is changing lanes based on the parameters of the lane ahead detected by the sensor and the relevant motion parameters of the ego vehicle. When the ego vehicle is changing lanes, the lane line needs to be optimized. According to the current position parameters of the ego vehicle and the parameters of the lane centerline of the lane after the lane change, the lane centerline of the lane change process is fitted with a quintic polynomial, and the following vehicle target is selected using the fitted lane centerline.

[0049] The left and right lane lines of the ego lane are obtained based on the video information captured by the camera, and then the lane center line of the own lane is obtained by calculation based on the left and right lane lines of the ego lane. The lane center line of the ego lane is expressed by a cubic polynomial of formula (1). In specific implementation, the ego vehicle control system can obtain the lane center line equation of the own lane in real time through the video of the camera.

[0050] According to the motion parameters of the vehicle, the longitudinal velocity of the vehicle is obtained v x_ego Longitudinal acceleration a x_ego , lateral speed v y_ego , lateral acceleration a y_ego , vehicle width W ego , steering wheel angle of the vehicle A ego , and then use the constant acceleration model to estimate the next moving position of the vehicle based on the current motion parameters of the vehicle.

[0051] Calculate the current coordinate position of the vehicle at time 0. In the initial state, the vehicle position ( x 0_ego , y 0_ego )for x 0_ego =0, y 0_ego = 0, the position of the lane line is y 0_lane =a0; Based on this, the distance △y0 between the center position of the vehicle at the starting position (in specific calculation, the center position uses the projection point of the rear axle center point of the vehicle on the ground) and the lane line is shown in the following formula (2).

[0052] (2); According to the speed of the ego vehicle, the ego vehicle's motion position and the position of the lane line 1 second ahead are previewed, and the constant acceleration model is used to estimate the ego vehicle state. In order to avoid the ego vehicle previewing distance 1 second ahead being too close at low speed, which will lead to deviations in the calculation of the vehicle's motion trend, the minimum value of the vehicle preview distance is set to 10m.

[0053] The longitudinal position of the vehicle after 1 second is the preview longitudinal position. x ego_preview , the expression is shown in the following formula (3).

[0054] (3); Substituting t=1s into formula (3), we obtain: .

[0055] and will x ego_preview The minimum value of is restricted to obtain: ; Then calculate the preview lateral position of the vehicle after 1 second y ego_preview , the expression is shown in the following formula (4).

[0056] (4); Substituting t=1s into formula (4), we obtain formula (5).

[0057] (5); At the vehicle preview point, the lateral position of the lane line at the preview point y lane_preview See the following formula (6).

[0058] (6); Formula (6), x ego_preview_final Indicates the longitudinal position of the vehicle in the last second of the lane changing process.

[0059] Then at the preview point, the distance between the vehicle's position and the lane line is △ y preview See the following formula (7).

[0060] (7); pass y lane_preview and y preview And other parameters to determine whether the vehicle needs to change lanes.

[0061] Judging whether the vehicle needs to change lanes includes the following two processes.

[0062] 1. The process of changing lanes to the left lane: When the vehicle is in the initial state △ y 0 <0, the vehicle at the preview point is within a certain range to the left of the lane line, △ y preview <-1.0, and the absolute value of the steering wheel angle A ego If the angle is less than 30° and the time from exiting the last left or right lane change is more than 3 seconds, the vehicle is considered to be changing lanes on the left; 2. The process of changing lanes to the right lane: When the vehicle is in the initial state △ y 0 >0, and at the preview point, the vehicle is within a certain range to the right of the lane line, △ y preview >1.0, and the steering wheel angle is the absolute value A ego If the angle is less than 30° and the time from exiting the last left or right lane change is more than 3 seconds, the vehicle is considered to be changing lanes on the right; When it is determined that the vehicle is changing lanes to the left or right, ACC releases the following target in the lane and uses the following target selected by the fitted lane line to perform following control.

[0063] In the present invention, the lane line fitting algorithm is the same when changing lanes on the left and on the right. The only difference is the parameters corresponding to the center line of the left or right lane when changing lanes. Therefore, the algorithm is illustrated by taking the lane changing on one side as an example. The present invention takes the lane changing on the right as an example to illustrate.

[0064] When the ego vehicle changes lanes, the starting point and end point of the quintic polynomial are set as follows: the starting point is the coordinate origin of the ego vehicle, and its related parameters can be obtained according to the sensors of the ego vehicle; the end point is the position parameter of the lane centerline preview point of the destination lane.

[0065] The centerline equation of the right lane is shown in the following formula (8).

[0066] (8); When changing lanes to the right, the fifth-order polynomial of the vehicle centerline equation during the lane changing process is shown in the following formula (9).

[0067] (9) In formula (9), C0 is the constant term of the centerline equation, and C1 to C5 are the coefficients of the first to fifth terms of the centerline equation, respectively.

[0068] The first-order derivative of the function f(x) in formula (9) with respect to x yields the following formula (10).

[0069] (10); The function f(x) in formula (9) can be differentiated by the second order with respect to x to obtain the following formula (11).

[0070] (11); When x = 0, substitute the motion parameters at the origin of the vehicle into formula (9), formula (10), and formula (11) to obtain formula (12), formula (13), and formula (14), respectively.

[0071] (12); (13); (14); In Formula 12 to Formula 14, y 0 is the lateral position error of the ego vehicle. The starting point of the motion planning of the ego vehicle is the current point of the ego vehicle, so it is equal to 0; v y0 is the lateral velocity of the vehicle, which can be measured by the vehicle's sensor; a y0 is the lateral acceleration of the vehicle, which can be measured by the vehicle's sensor.

[0072] When the vehicle enters the right lane change state, at the preview point x p = x ego_preview_final Substituting into the right lane centerline equation (9) yields equation (15).

[0073] (15) The function f(x) in formula (15) is p The first-order derivative gives the following formula (16).

[0074] (16); The function f(x) in formula (15) is p The second-order derivative gives the following formula (17).

[0075] (17); The parameters y p is the lateral position of the center line of the right lane at the preview point, see the following formula (18).

[0076] (18); parameter v ypis the first-order derivative of the center line of the right lane at the preview point, see the following formula (19).

[0077] (19); Parameters for a yp is the second-order derivative of the center line of the right lane at the preview point, see the following formula (20).

[0078] (20); By combining equations (12), (13), (14), (15), (16), (17), (18), (19) and (20), we can obtain three parameters, namely C3, C4 and C5, as shown in equation (21).

[0079] (twenty one); Then, the obtained parameters C0, C1, C2, C3, C4 and C5 are substituted into formula (9) as the lane centerline of the ego vehicle when the ego vehicle changes lanes.

[0080] The width of the lane line also needs to be optimized. Since the vehicle's trajectory changes rapidly during lane change or turning, in order to prevent the misselection of the target next to it, the width of the lane of the vehicle directly uses the width of the vehicle when the vehicle changes lanes, that is, W lane = W ego .

[0081] The vehicle uses a fifth-order polynomial to fit the lane line during the lane change process from the current starting point of the lane to the preview point of the target lane. A third-order polynomial is used to fit the effective distance from the starting point after the lane change to the lane line. The distance exceeding the lane line is estimated by expanding the lane line at the vanishing point. The specific schematic diagram after implementation is shown below. Figure 2 shown.

[0082] During the vehicle lane changing process, the following two time points need to be considered.

[0083] 1. The first time point T1 is when the front of the vehicle just crosses the lane line; that is, when the vehicle just enters the target lane, the left and right lane lines will jump; at this time, the absolute value of the difference between the lane centerline a0 value of the vehicle's lane and the a0 value of the previous moment is greater than 0.5 times the lane width to determine whether the vehicle has changed lanes and entered the right lane (when the absolute value is not greater than 0.5 times the lane width, it is determined that it has not entered the right lane). At this time, exit the logic of changing lanes on the right, but record the lane line fitted by the lane change for target screening. This is mainly to avoid using the parameters at this time to calculate the lane change, which is easy to calculate the vehicle as continuing to change lanes after entering the right lane, resulting in misselection of the target. The vehicle will make the next lane change judgment 3s after exiting the lane change state. The schematic diagram of lane change at the first time point T1 is as follows Figure 3 shown.

[0084] 2. At the second time point T2, when the vehicle enters the right lane, the steering wheel angle is sometimes too large when changing lanes, and there is often a large reverse angle to return to the right lane. The lane change logic cannot be calculated at this time, because after the driver successfully changes lanes, it is only to return the vehicle to the right lane, not to return to the original lane. The vehicle will make the next lane change judgment 3 seconds after exiting the lane change state. The schematic diagram of the lane change at the second time point T1 is as follows Figure 4 shown.

[0085] The target screening optimization algorithm of the self-vehicle when changing lanes of the present invention has been proved through experiments that after the ACC function is turned on, when the driver changes lanes, whether changing lanes to the left or to the right, the driver's lane-changing intention can be accurately identified, and the following vehicle target in the adjacent lane can be screened for following. When it is judged that the self-vehicle enters the lane change, the following vehicle target in the self-vehicle lane is released, and the fitted lane line is used to perform target screening, and the following vehicle target corresponding to the adjacent lane is screened.

[0086] The present invention has the following technical features.

[0087] 1. The intelligence of ACC longitudinal control function has been improved, which can identify the driver's lane-changing intention and make the system more in line with the driver's operating intention; 2. Improved the comfort of ACC longitudinal control function. When the vehicle changes lanes, the ACC controller releases the following vehicle target in the lane and selects the target in the adjacent lane for tracking control based on the fitted lane line, making the entire lane change process more comfortable. 3. Improve the accuracy of ACC longitudinal control target recognition. When the vehicle just enters the target lane for lane change, and when returning to the center of the steering wheel after entering the target lane, there will be no misjudgment of continuing to change lanes, thus improving the accuracy of target recognition during lane change; 4. Improved the efficiency of the longitudinal function of ACC. By considering the movement trend of the vehicle, the lane line is optimized according to the vehicle, making the target recognition in the lane change process more accurate, avoiding vehicle deceleration caused by misidentification, making the entire longitudinal control more comfortable, and increasing the number of times the driver turns on the ACC function and the efficiency of use; 5. Improve the safety of the ACC system by identifying target vehicles in adjacent lanes in advance, allowing the ACC system to have enough time to perform longitudinal control during lane changes, thereby improving the safety of the system.

[0088] The target screening optimization algorithm of the self-vehicle when changing lanes of the present invention uses the movement trend of the self-vehicle to judge that the vehicle enters the lane changing state, and uses the lane line fitted by the quintic polynomial to select the following target, and considers the process of the vehicle just entering the adjacent lane and adjusting the vehicle in the adjacent lane when changing lanes, so as to prevent the misjudgment of the vehicle continuing to change lanes.

[0089] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0090] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A target selection optimization algorithm for a vehicle when changing lanes, characterized in that: The steps include: Step 1: Activate the ACC longitudinal control function of the ego vehicle, obtain the ego vehicle parameters through the ego vehicle control system, obtain the parameters of the target vehicle in front through the ego vehicle camera, and obtain the ego lane parameters, left lane parameters, and right lane parameters through the ego vehicle camera; Step 2: Calculate the centerline equation of the self-lane, the centerline equation of the left lane and the centerline equation of the right lane according to the parameters of the self-lane, the parameters of the left lane and the parameters of the right lane; Step 3: According to the vehicle parameters, the centerline equation of the lane, the centerline equation of the left lane and the centerline equation of the right lane, determine whether the vehicle needs to change lanes, change lanes to the left lane or change lanes to the right lane; Step 4: When determining and entering the lane to be switched, calculate the current lane centerline of the vehicle; Step 5: When the vehicle just enters the lane to be switched, no lane change judgment will be performed within 3 seconds; Step 6: After the vehicle enters the lane to be switched, no lane change judgment will be made within 3 seconds, and the lane change of the vehicle is completed.

2. The target selection optimization algorithm for a vehicle when changing lanes according to claim 1 is characterized in that: In step 1, the vehicle parameters include the lateral speed, longitudinal speed, lateral acceleration, longitudinal acceleration and steering wheel angle of the vehicle.

3. The target selection optimization algorithm for a vehicle when changing lanes according to claim 1 is characterized in that: In the step 1, the parameters of the front target vehicle include the lateral speed, longitudinal speed, lateral acceleration, longitudinal acceleration, lateral position and longitudinal position of the front vehicle.

4. The target selection optimization algorithm for a vehicle when changing lanes according to claim 1 is characterized in that: In the step 1, the lane parameters include the cubic polynomial coefficient of the left lane line of the lane, the cubic polynomial coefficient of the right lane line, the lane line credibility and the lane line length.

5. The target selection optimization algorithm for a vehicle when changing lanes according to claim 1 is characterized in that: In step 1, the left lane parameters include the cubic polynomial coefficient of the left lane line of the left lane, the cubic polynomial coefficient of the right lane line, the lane line credibility and the lane line length.

6. The target selection optimization algorithm for a vehicle when changing lanes according to claim 1 is characterized in that: In step 1, the right lane parameters include the cubic polynomial coefficients of the left lane line of the right lane, the cubic polynomial coefficients of the right lane line, the lane line credibility and the lane line length.

7. The target selection optimization algorithm for a vehicle when changing lanes according to claim 1 is characterized in that: In step 2, the centerline equation and lane width of the self-lane are obtained through the self-vehicle camera. ; The lane centerline equation is shown in the following formula (1); （1）； In formula (1), y0 is the dependent variable of the lane centerline equation, x is the independent variable, a0 is the constant term of the lane centerline equation, and a1 to a3 are the coefficients of the linear to cubic terms of the lane centerline equation, respectively.

8. The target selection optimization algorithm for a vehicle when changing lanes according to claim 1 is characterized in that: In step 3, when judging whether the vehicle needs to change lanes, change lanes to the left lane or change lanes to the right lane, when the vehicle is in the initial state, the distance △ between the center position of the vehicle at the starting position and the lane line y 0 <0, the vehicle at the preview point is within a certain range to the left of the lane line, and the distance between the vehicle position and the lane line is △ y preview <-1.0, and the absolute value of the steering wheel angle A ego If the angle is less than 30° and the time from exiting the last lane change to the left or right exceeds 3 seconds, the vehicle is considered to have changed lanes on the left.

9. The target selection optimization algorithm for a vehicle when changing lanes according to claim 1 is characterized in that: In step 3, when judging whether the vehicle needs to change lanes, change lanes to the left lane or change lanes to the right lane, when the vehicle is in the initial state, the distance △ between the center position of the vehicle at the starting position and the lane line y 0 >0, and at the preview point, the vehicle is within a certain range to the right of the lane line, and the distance between the vehicle position and the lane line is △ y preview >1.0, and the steering wheel angle is the absolute value A ego If the angle is less than 30° and the time from exiting the last lane change to the left or right exceeds 3 seconds, the vehicle is considered to be changing lanes to the right.

10. The target selection optimization algorithm for a vehicle when changing lanes according to claim 1, characterized in that: In step 4, the current lane centerline of the vehicle is expressed by a quintic polynomial using the following formula (9): （9） In formula (9), C0 is the constant term of the centerline equation, and C1 to C5 are the coefficients of the first to fifth terms of the centerline equation, respectively.

Images