id="p-0071" num="0070">Accordingly, if the collision risk with the peripheral is relatively small (if the collision risk is equal to or less than the first threshold value), it is possible to improve the driving responsiveness and the operability by rapidly entering the neighboring lane by increasing the lateral speed at the start of the lane change.

In addition, when the collision risk with the peripheral is relatively high (when the collision risk is greater than the first threshold value), the lateral speed at the start of lane change may be reduced and the entry into the neighboring lane may be performed relatively late. Accordingly, it is possible to safely change the lane by exposing the lane changing intention to the peripheral for a long time (increasing the time for the peripheral to recognize the illuminated turn signal lamp of the driving vehicle).

Meanwhile, if the peripheral of the rear side (rear left of rear right side) increases the speed while the vehicle is driving to change the lane according to the lane changing trajectory generated in the lane changing trajectory generation step S130, the collision risk may increase. In this case, since continuing to change the lane may result in a collision with the peripheral, a situation may arise where the vehicle needs to stop changing the lane and return to the existing lane.

FIG. 4 is a control flowchart of a method for generating a lane change trajectory of a vehicle according to an embodiment of the present disclosure, showing a case of returning to the existing lane according to a driving risk during the lane change.

Specifically, after the vehicle control step S140 for controlling the vehicle according to the lane changing trajectory generated in the lane changing trajectory generating step S130 is performed, determining whether the collision risk with the peripheral exceeds a predetermined second threshold value S150 may be performed.

Here, the second threshold value may be a value greater than the first threshold value. That is, if the collision risk exceeds the first threshold value, a lane changing trajectory is generated such that the lateral speed is small, and if the collision risk exceeds the second threshold value which is greater than the first threshold value, the vehicle may be controlled to return to the existing lane since it is determined that there is a high risk to collide with the peripheral if the vehicle continues to drive according to the generated lane changing trajectory. Meanwhile, the second threshold value may be an arbitrary set value like the first threshold value. For example, in case that the collision risk is calculated using a probabilistic collision risk model, a collision risk of 30% may be set as the second threshold value.

If the collision risk with the peripheral exceeds the second threshold value (‘Yes’ in S150), the lane returning trajectory generation step S160 for generating a lane returning trajectory may be performed, and a step of controlling the vehicle S170 for controlling the vehicle according to the generated returning trajectory may be performed.

Meanwhile, if the collision risk with the peripheral does not exceed the second threshold value (‘No’ in S150), the vehicle may be controlled to change the lane according to the previously generated lane changing trajectory (S140).

In this regard, according to the prior art, since the lane changing trajectory in the form of a symmetrical graph is generated regardless of the collision risk with the peripheral, in cases that it is necessary to return to the existing lane due to the collision risk with the peripheral, the driver may feel uncomfortable by generating an unstable lane returning trajectory due to a sudden direction change. In addition, there may be a problem in that the vehicle may have already entered the neighboring lane at the timing of determining whether to return to the existing lane, which may further increase the collision risk with peripheral.

On the other hand, according to method for generating the lane changing trajectory S100′ of an embodiment of the present disclosure, if there is a collision risk with the peripheral (i.e., if the collision risk exceeds the first threshold value), by making the lateral speed at the start of the lane change smaller than the predetermined lateral speed (the lateral speed of the vehicle according to the symmetrical graph), it is possible to generate the returning trajectory as a safer trajectory in case that the vehicle is required to return to the existing lane when the collision risk is increased during the lane change (the collision risk exceeds the second threshold value). In addition, since the initial lateral speed is reduced and the entry into the neighboring lane is performed relatively late, the vehicle can be controlled to return before entering the neighboring lane, thereby reducing the collision risk with the peripheral.

FIG. 5 illustrates diagrams showing trajectories in which the vehicle changes the lane according to the driving risk in the method for generating the lane changing trajectory of the vehicle according to the embodiment of the present disclosure.

In FIG. 5, if the driver executes the lane changing function to change the lane while the vehicle 1 is driving, it is determined whether the collision risk with the peripheral exceeds the first predetermined threshold value.

As shown in FIG. 5(a), if the collision risk with the peripheral does not exceed the predetermined first threshold value, such as in case that there is no peripheral, the vehicle 1 may generate a lane changing trajectory A having an initial lateral speed greater than the lateral speed at the start of the lane change in the lane changing trajectory P (the predetermined lateral speed) in which the graph of data is to be symmetric.

On the other hand, as shown in FIG. 5(b), if the peripheral 2 driving in the neighboring lane where the vehicle 1 is about to cut in exists within a predetermined distance to the rear side (rear left or rear right side) of the vehicle 1, the collision risk with the peripheral 2 may exceed the first threshold value. In this case, the vehicle 1 may generate a lane changing trajectory B having an initial lateral speed smaller than the lateral speed at the start of the lane change in the lane change trajectory P (predetermined lateral speed) in which the graph of data regarding the lane changing trajectory is to be symmetric.

Meanwhile, in FIG. 5(b), while the vehicle 1 is driving along the lane changing trajectory B, if the peripheral 2 accelerates so that the collision risk exceeds the second threshold value, the vehicle 1 may return to the existing lane by generating the lane returning trajectory R. In this case, according to the trajectory B, since the initial entry speed (e.g., initial lateral speed) is smaller than that of the lane changing trajectory P so that it is possible to generate a gradual returning trajectory, the driver's safety and driving convenience can be improved.

FIG. 6 illustrates graphs showing changes in a lateral movement distance, a lateral speed, and a lateral acceleration of the vehicle over time when the vehicle changes the lane according to the method for generating the lane changing trajectory of the present disclosure.

In the embodiment of FIG. 6, in the case of following the lane changing trajectory generated by the lane changing trajectory generation method according to the embodiment of the present disclosure, it is exemplified so that the lane change is completed when the vehicle moves 4 m in the lateral distance, and the lane change is completed within 8 seconds (the first predetermined time) from the start of the lane change. However, the present disclosure is not limited thereto, and if the regulations on the lane changing function (ACSF) are complied, the total lateral movement distance and total lane changing time from the start to the completion of the lane change may be set differently.

Meanwhile, in FIG. 6, the graph shown with a dotted line represents a conventional lane changing trajectory (trajectory P) that is set so that the graph of the lane changing data is symmetric, and the graph shown with a solid line represents a lane changing trajectory (trajectory A) with a large initial lateral speed when the collision risk is equal to or less than the first threshold value in the embodiment of the present disclosure. In addition, the graph shown with a dashed-dotted line represents a lane changing trajectory (trajectory B) with the initial lateral speed set to be small when the collision risk is greater than the first threshold value in the embodiment of the present disclosure.

Looking at the graph regarding the lateral movement distance y over time with reference to FIG. 6(a), in case that the collision risk is equal to or less than the first threshold value in the lane changing trajectory generation method according to the embodiment of the present disclosure, the vehicle may pass the midpoint of the lane change (the lateral movement distance of 2 m) at the timing that approximately 3 seconds (the second predetermined time) has elapsed. In other words, when approximately 3 seconds have elapsed, the entry into the neighboring lane (a part of the driving vehicle overlaps with one lane line of the neighboring lane) may occur.

On the other hand, in case that the collision risk exceeds the first threshold value, the second predetermined time may be approximately 5 seconds. In other words, the vehicle may pass the midpoint of the lane change (the lateral movement distance of 2 m) after approximately 5 seconds. In this way, if the collision risk is relatively high, the entry into the neighboring lane (a part of the driving vehicle overlapping with one lane line of the neighboring lane) may occur at relatively late timing from the start of the lane change.

FIG. 6(b) is a graph showing data on the lateral speed v_yfrom the start to the end of the lane change. As shown in FIG. 6(b), in the case of a graph of lateral speed data where the collision risk is equal to or less than the first threshold value (trajectory A), a lane changing trajectory may be generated with a relatively large lateral speed at the start of the lane change. On the other hand, a graph may be formed such that the lateral speed gradually decreases after reaching the maximum lateral speed.

In the case of a graph where the collision risk is greater than the first threshold value (trajectory B), since a lane changing trajectory is generated with a relatively small lateral speed at the start of the lane change, it may take longer to reach the maximum lateral speed. On the other hand, after reaching the maximum lateral speed, a graph may be formed such that the lateral speed decreases relatively rapidly.

FIG. 6(c) is a graph showing data on the lateral acceleration ay of the vehicle from the start to the end of the lane change. As shown in FIG. 6(c), according to the lane changing trajectory according to the prior art, it has a point symmetrical form about the midpoint (4 seconds point), while the graph where the collision risk is equal to or less than the first threshold value (trajectory A), the absolute value of the lateral acceleration at the initial stage of the lane change increases, and in the graph where the collision risk is greater than the first threshold value (trajectory B), the absolute value of the lateral acceleration at the final stage of the lane change increases.

As shown in the embodiments of FIG. 6(a) to FIG. 6(c), the lane changing trajectory generated according to the embodiment of the present disclosure may have a shape that, based on the midpoint of the lane change (the 4 seconds point in the graph) from the start point (starting time point) to the end point (the ending time point) of the lane change, each of the graphs of the vehicle's lateral movement distance, the lateral speed, and the lateral acceleration over time before the midpoint and each of the graphs of the vehicle's lateral movement distance, the lateral speed, and the lateral acceleration after the midpoint is asymmetric.

Meanwhile, in the embodiment of FIG. 6, the second predetermined time (from the start of lane change to the start of the entry into the neighboring lane) is exemplified as 3 seconds and 5 seconds depending on the collision risk, but the present disclosure is not limited thereto. The second predetermined time in the trajectory A and in the trajectory B may be set differently, as long as the conditions that the entry time point into the neighboring lane in the trajectory A is earlier than that in the trajectory P and the entry time point into the neighboring lane in the trajectory B is later than that in the trajectory P are satisfied.

Further, the time from the entry time point into the neighboring lane to the time point that the lane change is completed may be set as a third predetermined time. That is, the first predetermined time (the time from the start of lane change to the end of the lane change) may be the sum of the second predetermined time and the third predetermined time.

In the embodiment of FIG. 6, since the first predetermined time is illustrated as 8 seconds, in case of the trajectory A (if the second predetermined time is 3 seconds), the third predetermined time may be 5 seconds, and in case of the trajectory B (if the second predetermined time is 5 seconds), the third predetermined time may be 3 seconds.

Meanwhile, the embodiment of FIG. 6 illustrates that the first predetermined time is 8 seconds, but as long as the regulations regarding the lane changing function are to be complied (for example, the lane change must be completed within 10 seconds), the first predetermined time may be set differently. Meanwhile, according to the standards of the regulations regarding the lane changing function, in the embodiments of the present disclosure, the second predetermined time may be in the range of approximately 3 to 5 seconds, and the third predetermined time may be approximately within 5 seconds.

FIG. 7 is a graph for explaining a method for generating a trajectory using a spline in the method for generating the lane changing trajectory according to the embodiment of the present disclosure.

Referring to FIG. 7, the lane changing trajectory according to the present disclosure may be a spline-curve based trajectory. For example, the lane changing trajectory of the present disclosure may be based on a spline curve calculated by two 5th degree polynomials.

Looking more specifically with reference to FIG. 7, the time point at which the midpoint (mid distance point) of the lateral movement distance (the entry point into the neighboring lane) is reached may be set as t_mid, and the time point at which the lane change is completed may be set as t_end. Further, the lateral speed of the vehicle at the time point of reaching the midpoint of the lateral movement distance may be represented as v_ytmid. Here, t_mid, t_end, and v_ytmidmay be tuning parameters. For example, in the embodiment of FIG. 6, it is illustrated that t_endis 8 seconds, and t_midis 3 seconds (in case that the initial lateral speed is large according to the trajectory A) or 5 seconds (in case that the initial lateral speed is small according to the trajectory B), and v_ytmidis 0.937 m/s.

Here, if the trajectory from the start to t_midwhich is the time point of reaching the midpoint (the time point of entry into the neighboring lane) is P₁(t), and the trajectory from t_midto t_end(the ending time point of lane change) is P₂(t), then each trajectory may be expressed in below two equations.

$P_{1} (t) = c_{0, 1} + c_{1, 1} t + c_{2, 1} t^{2} + c_{3, 1} t^{3} + c_{4, 1} t^{4} + c_{5, 1} t^{5} = \sum_{i = 0}^{5} c_{i, 1} t^{i}$

$P_{2} (t) = c_{0, 2} + c_{1, 2} t + c_{2, 2} t^{2} + c_{3, 2} t^{3} + c_{4, 2} t^{4} + c_{5, 2} t^{5} = \sum_{i = 0}^{5} c_{i, 2} t^{i}$

Here, the spline boundary conditions may include the following eight boundary conditions.

$\begin{matrix} P_{1} (0) = c_{0, 1} = y_{0} (Lateral position at the starting time point) & (1) \end{matrix}$

$\begin{matrix} \frac{{dP}_{1}}{dt} (0) = c_{1, 1} = v_{y_{0}} (Lateral speed at the starting time point) & (2) \end{matrix}$

$\begin{matrix} \frac{d^{2} P_{1}}{{dt}^{2}} (0) = 2 c_{2, 1} = a_{y_{0}} (Lateral acceleration at the starting time point) & (3) \end{matrix}$

$\begin{matrix} P_{2} (t_{end}) = \sum_{i = 0}^{5} c_{i, 2} t_{end}^{i} = y_{t_{end}} (Lateral position at the ending time point) & (4) \end{matrix}$

$\begin{matrix} \frac{{dP}_{2}}{dt} (t_{end}) = v_{y_{t_{end}}} = 0 (Lateral speed at the ending time point) & (5) \end{matrix}$

$\begin{matrix} \frac{d^{2} P_{2}}{{dt}^{2}} (t_{end}) = a_{y_{t_{end}}} = 0 (Lateral acceleration at the ending time point) & (6) \end{matrix}$

$\begin{matrix} P_{1} (t_{mid}) = \sum_{i = 0}^{5} c_{i, 1} t_{mid}^{i} = y_{t_{mid}} (Lateral position at the midpoint reaching time point) & (7) \end{matrix}$

$\begin{matrix} \frac{{dP}_{1}}{dt} (t_{mid}) = v_{y_{t_{mid}}} (Lateral speed at the midpoint reaching time point) & (8) \end{matrix}$

Further, the spline continuity conditions at the midpoint reaching time point (t_mid) may include the following four continuity conditions.

$\begin{matrix} P_{1} (t_{mid}) = P_{2} (t_{mid}) (Lateral position continuity equation at the midpoint reaching time point) & (i) \end{matrix}$

$\begin{matrix} \frac{{dP}_{1}}{dt} (t_{mid}) = \frac{{dP}_{2}}{dt} (t_{mid}) (Lateral speed continuity equation at the midpoint reaching time point) & (ii) \end{matrix}$

$\begin{matrix} \frac{d^{2} P_{1}}{{dt}^{2}} (t_{mid}) = \frac{d^{2} P_{2}}{{dt}^{2}} (t_{mid}) (Lateral acceleration continuity equation at the midpoint reaching time point) & (iii) \end{matrix}$

d
3

⁢

P
1

dt
3

⁢

(

t
mid

)

=

d
3

⁢

P
2

dt
3

⁢

(

t
mid

)

⁢

(

Lateral
⁢

jerk
⁢

continuity