JP2010030399A - Vehicle operation support device and vehicle operation support method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a vehicle driving support, suitably performing support control for a side obstacle even when a lane dividing line is not detected. <P>SOLUTION: When an obstacle at the side of a driver's own vehicle is detected, a future position of the own vehicle in the lapse of a predetermined time is estimated based on steering input of the driver. When the estimated future position of the own vehicle is determined to have reached a determination threshold value with reference to a lane dividing line at the side closer to the obstacle or to be closer to the obstacle than the determination threshold value, the start of the support control for the side obstacle is determined. However, when the lane dividing line at the side closer to the obstacle is not detected, the lane dividing line taken to be reference for determining the above control start is displaced to the outside in the lane width direction. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a vehicle driving support device and a vehicle driving support method for supporting driving of a driver so as to prevent an approach to the obstacle when an obstacle on the side of the host vehicle is detected.

As a device that supports the driving of the driver, for example, there is a technique described in Patent Document 1. In this technique, the presence or absence of steering is determined when the vehicle speed exceeds the set vehicle speed. And the distance to the obstacle which exists in the steering direction side of the own vehicle is detected. When the distance to the obstacle is within the set distance, the steering to the obstacle side is suppressed and controlled. This makes it possible to warn the driver of approaching an obstacle.
Further, when the vehicle speed is equal to or lower than the set vehicle speed, the process does not shift to the control, and the suppression control is canceled while the steering suppression control is in operation. As a result, when the traveling road is estimated to be a curved road, the steering is prevented from being suppressed and the vehicle is prevented from deviating from the traveling road.
JP-A-8-253160

  In the above prior art, even when an obstacle exists on the side of the vehicle, the control is prevented from intervening when estimating a curved road or a lane change. At this time, it is estimated whether the road is a curved road or the like based on the vehicle speed of the host vehicle. That is, since it is not determined whether or not the control is performed by acquiring an actual lane marking, there is a situation in which the control is temporarily terminated even on a straight road. That is, there is a possibility that control intervention and cancellation may occur repeatedly more than necessary.

That is, in such a technique, when the lane marking is not acquired, there is a possibility that the assistance control for the side obstacles may be intervened and canceled more than necessary. Such a thing leads to a driver's uncomfortable feeling.
The present invention has been made paying attention to the above points, and provides vehicle driving support capable of appropriately performing support control for side obstacles even when a lane marking may not be detected. The task is to do.

  In order to solve the above problems, the present invention predicts the future position of the host vehicle after a predetermined time based on the driver's steering input when detecting an obstacle on the side of the host vehicle, and predicts the future of the host vehicle. When the position reaches a predetermined lane width direction position with reference to the lane marking on the side closer to the obstacle, or is determined to be on the obstacle side with respect to the predetermined lane width direction position, support control for the side obstacle is performed. It is determined to start. However, when the lane marking near the obstacle cannot be detected, the lane marking used as a reference when determining the start of the control is displaced outward in the lane width direction.

According to the present invention, since the future position of the host vehicle with respect to the obstacle is determined by detecting the lane division line of the traveling road, even when steering on a curved road, the vehicle is appropriately directed toward the obstacle. It can be determined whether or not. This makes it possible to appropriately perform support control for avoiding a side obstacle even on a curved road.
Further, when it becomes impossible to detect the lane line during traveling, the lane line used as a reference when determining the relative start of control is shifted to the outside in the vehicle width direction. This makes it difficult to determine that control has started. That is, the control start timing is delayed. As a result, the start / end of control is prevented more than necessary, and support control for necessary side obstacles is performed. That is, it is possible to start control when the driver steers intentionally in the direction of the obstacle while suppressing the start of control with a sense of incongruity.
As described above, even when the lane markings cannot be detected, it is possible to appropriately perform the support control for the side obstacle.

Next, embodiments of the present invention will be described with reference to the drawings.
In the present embodiment, a case where a side obstacle support control device is mounted on a rear wheel drive vehicle will be exemplified. The target vehicle may be front-wheel drive or four-wheel drive.
FIG. 1 is a schematic configuration diagram of an apparatus according to the present embodiment.
(Constitution)
This vehicle is equipped with an automatic transmission and a differential gear. Both the front and rear wheels are equipped with a braking device that can independently control the braking force of the left and right wheels.

Reference numeral 1 denotes a brake pedal. The brake pedal 1 is connected to the master cylinder 3 via the booster 2. Reference numeral 4 denotes a reservoir. The master cylinder 3 is connected to each wheel cylinder 6FL to 6RR of each wheel via a fluid pressure circuit 30. As a result, when the brake control is not activated, the brake fluid pressure is increased by the master cylinder 3 in accordance with the depression amount of the brake pedal 1 by the driver. The increased braking fluid pressure is supplied to the wheel cylinders 6FL to 6RR of the wheels 5FL to 5RR through the fluid pressure circuit 30.
The braking fluid pressure control unit 7 controls the actuator in the fluid pressure circuit 30 to individually control the braking fluid pressure to each wheel. Then, the brake fluid pressure to each wheel is controlled to a value corresponding to the command value from the braking / driving force control unit 8. An example of the actuator is a proportional solenoid valve capable of controlling each wheel cylinder hydraulic pressure to an arbitrary braking hydraulic pressure.

  Here, the brake fluid pressure control unit 7 and the fluid pressure circuit 30 may use, for example, a brake fluid pressure control unit used in anti-skid control (ABS), traction control (TCS), or vehicle dynamics control device (VDC). good. The brake fluid pressure control unit 7 may be configured to control the brake fluid pressure of each of the wheel cylinders 6FL to 6RR independently. When a braking fluid pressure command value is input from a braking / driving force control unit 8 to be described later, each braking fluid pressure is controlled according to the braking fluid pressure command value.

The vehicle is provided with a drive torque control unit 12.
The drive torque control unit 12 controls the drive torque to the rear wheels 5RL and 5RR that are drive wheels. This control is realized by controlling the operating state of the engine 9, the selected transmission ratio of the automatic transmission 10, and the throttle opening of the throttle valve 11. That is, the drive torque control unit 12 controls the fuel injection amount and the ignition timing. At the same time, the throttle opening is controlled. Thereby, the operating state of the engine 9 is controlled.

Further, the drive torque control unit 12 outputs the value of the drive torque Tw, which is information at the time of control, to the braking / driving force control unit 8.
The drive torque control unit 12 can also control the drive torque of the rear wheels 5RL and 5RR independently. However, when a driving torque command value is input from the braking / driving force control unit 8, the driving wheel torque is controlled according to the driving torque command value.

Moreover, the imaging part 13 with an image processing function is provided in this vehicle front part. The imaging unit 13 is used to detect the position of the host vehicle in the traveling lane. The imaging unit 13 is configured by a monocular camera including, for example, a CCD (Charge Coupled Device) camera.
And the imaging part 13 images the front of the own vehicle. Then, the imaging unit 13 performs image processing on the captured image in front of the host vehicle, detects a lane marking such as a white line (lane marker), and detects a traveling lane based on the detected lane marking.

Further, the imaging unit 13 determines, based on the detected travel lane, an angle (yaw angle) φfront between the travel lane of the host vehicle and the longitudinal axis of the host vehicle, a lateral displacement Xfront with respect to the travel lane, and a travel lane curvature β. Etc. are calculated. The imaging unit 13 outputs the calculated yaw angle φfront, lateral displacement Xfront, travel lane curvature β, and the like to the braking / driving force control unit 8.
At this time, when the lane marking is not detected, the imaging unit 13 outputs a signal indicating that the lane marking has not been detected to the braking / driving force control unit 8. Further, information on the recognition accuracy of detection is also output to the braking / driving force control unit 8 according to the accuracy of the captured image.

Here, even if the image quality (recognition accuracy) of the captured image is greater than or equal to a predetermined value, there is no lane marking in the vicinity of the toll booth or the like, so it is impossible to detect the lane marking.
Here, the imaging unit 13 detects a lane marking that forms a traveling lane, and calculates the yaw φfront based on the detected lane marking. For this reason, the yaw angle φfront greatly affects the detection accuracy of the lane markings of the imaging unit 13.

  Further, the vehicle is provided with a radar device 22L / R. The radar devices 22L / R are sensors for detecting obstacles traveling in the left and right side directions. The radar device 22L / R is set so as to be able to detect whether or not an obstacle exists in at least a predetermined blind spot area on the side surface. Desirably, detection of the relative lateral position POSXobst, the relative longitudinal position DISTobst, and the relative longitudinal velocity dDISTobst with respect to the obstacle can be detected respectively.

In addition, a master cylinder pressure sensor 17, an accelerator opening sensor 18, a steering angle sensor 19, a direction indicating switch 20, and wheel speed sensors 22FL to 22RR are provided.
The master cylinder pressure sensor 17 detects the output pressure of the master cylinder 3, that is, the master cylinder hydraulic pressure Pm. The accelerator opening sensor 18 detects the amount of depression of the accelerator pedal, that is, the accelerator opening θt. The steering angle sensor 19 detects the steering angle (steering angle) δ of the steering wheel 21. The direction indication switch 20 detects a direction indication operation by the direction indicator. The wheel speed sensors 22FL to 22RR detect the rotational speeds of the wheels 5FL to 5RR, so-called wheel speeds Vwi (i = fl, fr, rl, rr). These sensors and the like output the detected detection signal to the braking / driving force control unit 8.

As shown in FIG. 2, the braking / driving force control unit 8 includes a future position predicting means 8A, an avoidance control start detecting means 8B, and an obstacle avoidance control means 8C. The avoidance control start detection unit 8B includes a lane marking adjustment unit 8Ba.
The future position prediction means 8A predicts the future position of the host vehicle MM after the forward gaze time Tt, which is a predetermined time, based on the driver's steering input.
When the avoidance control start detection means 8B determines that an obstacle on the side of the host vehicle is detected, the future position of the host vehicle is an obstacle based on the lane marking near the obstacle. The start of control is detected when the object distance X2obst is reached.

When the avoidance control start detection means 8B detects the start of control, the obstacle avoidance control means 8C calculates a yaw moment Ms for controlling the host vehicle so as to prevent the vehicle from approaching the obstacle.
If the lane line adjusting unit 8Ba determines that the detected lane line cannot be detected, the lane line adjusting unit 8Ba sets a virtual lane line outside the lane line detected by the lane line detecting unit. While the lane marking is not detected, the lane marking is changed to a virtual lane marking based on the lane marking when the lane marking is detected.

Next, processing of the braking / driving force control unit 8 will be described with reference to FIG.
The processing of the braking / driving force control unit 8 is executed by timer interruption every predetermined sampling time ΔT every 10 msec, for example. Although no communication process is provided in the process shown in FIG. 3, the information acquired by the arithmetic process is updated and stored in the storage device as needed, and necessary information is read out from the storage device as needed.
First, in step S10, various data are read from each sensor, controller and control unit. Specifically, each wheel speed Vwi, steering angle δ, master cylinder hydraulic pressure Pm, and direction switch signal detected by each sensor are acquired.

Next, the vehicle speed V is calculated in step S20. That is, the vehicle speed V is calculated based on the wheel speed Vwi as in the following formula (1).
V = (Vwr1 + Vwrr) / 2 (: front wheel drive)
V = (Vwfl + Vwfr) / 2 (: In the case of rear wheel drive)
... (1)
Here, Vwfl and Vwfr are the wheel speeds of the left and right front wheels, respectively. Vwrl and Vwrr are the wheel speeds of the left and right rear wheels, respectively. That is, in the above equation (1), the vehicle speed V is calculated as an average value of the wheel speeds of the driven wheels. In this embodiment, since the vehicle is a rear-wheel drive vehicle, the vehicle speed V is calculated from the latter equation, that is, the wheel speed of the front wheels.

Further, when another automatic braking control device such as ABS (Anti-lock Break System) control is operating, an estimated vehicle speed estimated by the other braking control device is acquired, and the vehicle speed V Used as
Next, in step S30, based on the signals from the left and right radar devices 22L / R, the presence / absence of the presence SMobs / Robsts of the obstacle SM is acquired for the left and right sides of the host vehicle MM. In addition, when using a sensor with higher detection accuracy, the relative position and relative speed of the side obstacle SM with respect to the host vehicle MM are also acquired. Here, as shown in FIG. 4, the side of the host vehicle MM includes an obliquely rearward position with respect to the host vehicle MM.

Next, in step S35, the detection status of the lane marking is determined.
That is, based on the recognition information from the imaging unit 13, it is determined whether left and right lane markings are detected, and lane marking flags CAMLASTright and CAMLASTleft are set. The lane marking flags CALMOSTright and CAMLASTleft are information on the right and left lane markings, respectively. The lane marking flags CAMLASTright and CAMLASTleft are set to “0” when a lane marking is detected, and set to “1” when a lane marking cannot be detected. If the recognition accuracy is worse than normal, “2” is set.

Further, the continuation time after the detection of the lane markings becomes impossible is counted separately for each of the left and right lane markings. This duration is referred to as lost time LTr, LTl. When the corresponding lane marking is detected, “0” is cleared.
Next, in step S40, the lateral displacement Xfront of the host vehicle MM and the curvature βfront of the traveling lane on the currently traveling road are read from the imaging unit 13.
Further, the yaw angle φfront of the host vehicle MM with respect to the currently traveling road is calculated. This yaw angle φfront is used to detect the running situation in the lane.
In the present embodiment, the yaw angle φfront uses a measured value obtained by the imaging unit 13.

However, when the lane line on the steering side has not been detected based on the lane line flags CAMLOSTright and CAMLASTleft, the yaw angle φfront is calculated based on the lane line that has been detected so far. For example, the yaw angle φfront is calculated based on the nearby lane markings imaged by the imaging unit 13. In this case, for example, the yaw angle φfront is calculated by the following equation (2) using the lateral displacement Xfront of the host vehicle MM. Note that the same calculation may be performed when a lane marking is detected.
φfront = tan ⁻¹ (dX ′ / V (= dX / dY)) (2)
here,
dX: A change amount of the lateral displacement X per unit time dY: A change amount in the traveling direction per unit time dX ′: A differential value of the change amount dX.

  Further, when the yaw angle φfront is calculated based on a nearby lane marking, it is not limited to calculating the yaw angle φfront using the lateral displacement X as in the above equation (2). For example, a lane marking detected in the vicinity may be extended far and the yaw angle φfront may be calculated based on the extended lane marking. If the lane marking cannot be detected, the yaw angle φfront may be calculated by this process.

Next, in step S50, a neutral yaw rate φ′path is calculated.
If the lane line on the steering side can be detected based on the lane line flags CAMLOSTright and CAMLASTleft, the neutral yaw rate φ′path is calculated by the following equation (3).
φ'path = βfront × V (3)
The neutral yaw rate φ′path is a yaw rate necessary for the host vehicle MM to maintain traveling along the traveling path. The neutral yaw rate φ'path is zero when traveling on a straight road. However, on the curved road, the neutral yaw rate φ′path changes depending on the curvature βfront. Therefore, when calculating the neutral yaw rate φ′path, the curvature βfront of the travel lane is used.

As the neutral yaw rate φ'path for maintaining this travel route, the average value φ'ave of the yaw rate φ 'during a predetermined time is used, or a filter with a large time constant is applied to the yaw rate φ'. The value may be simply calculated.
On the other hand, if the lane line on the steering side is not detected based on the lane line flags CALMOSTright and CAMLASTleft, “0” is set to the neutral yaw rate φ′path.
This is because the lane marking is not detected and the estimation accuracy of the neutral yaw rate φ′path is not good. The neutral yaw rate φ′path = “0” means that a straight road is estimated.

  However, if the steering-side lost time is within a predetermined time, the neutral yaw rate φ'path immediately before the lane marking can no longer be detected is not "0" and the steering direction is on the curve-out side. May use the previous neutral yaw rate φ′path as the current neutral yaw rate φ′path. Alternatively, the average value φ′ave of the yaw rate φ ′ during a predetermined time may be used, or a value obtained by multiplying the yaw rate φ ′ by a filter having a large time constant may be used.

Next, in step S60, a forward gaze time Tt (= vehicle head distance) is set.
The forward gaze time Tt is a time for determining a threshold value for predicting the contact situation of the driver with the future obstacle SM. For example, the forward gaze time Tt is set to 1 second.
Further, target yaw rates ψdriver and ψdriverhosei are calculated.
The target yaw rate ψdriver is calculated from the steering angle δ and the vehicle speed V as in the following equation. This target yaw rate ψdriver is a target yaw rate generated in response to steering. Kv is a gain.
Ψdriver = Kv ・ δ ・ V

Further, the target yaw rate ψdriverhosei is calculated by the following equation (4). This target yaw rate Ψ driverhosei is a value obtained by removing the yaw rate φ′path necessary for traveling on the traveling path from the target yaw rate Ψ driver. This eliminates the influence of steering for traveling on a curved road.
Ψdriverhosei = Ψdriver − φ'path (4)

Next, in step S70, the predicted vehicle position ΔXb in the lateral direction with respect to the current travel path position is calculated based on the following equation (5). This own vehicle predicted position ΔXb is used to determine whether or not to change the lane after leaving the travel path. That is, the host vehicle predicted position ΔXb is used for starting avoidance support control for the obstacle SM. Actually, the own vehicle predicted position ΔXb is also obtained separately on the left and right.
ΔXb = (K1φ + K2φm + K3φm ′) (5)
here,
φ: Yaw angle φm: Target yaw angular velocity φm ': Target yaw angular acceleration.

The target yaw angular velocity φm is expressed by the following equation.
φm = Ψdriverhosei × Tt
The target yaw angular acceleration φm ′ is expressed by the following formula.
φm ′ = φm × Tt2
Here, when the forward gaze distance L is used in order to set the vehicle predicted position ΔXb to the dimension of the yaw angle, it can be expressed by the following equation.
ΔXb = L · (k1φ + k2φm × T + k3φm ′ × Tt2)
Here, the front gaze distance L and the front gaze time Tt are in the relationship of the following formula.
Forward gaze distance L = front gaze time Tt × vehicle speed V

Considering these characteristics, the set gain K1 is a value that is a function of the vehicle speed. The set gain K2 is a value that is a function of the vehicle speed and the forward gaze time. The set gain K3 is a value that is a function of the vehicle speed and the square of the forward gaze time.
Note that the predicted position of the host vehicle MM may be calculated by select high by separately obtaining the steering angle component and the steering speed component as in the following equation.
ΔXb = max (K2φm, K3∫φm ′)
Next, in step S80, a determination threshold for starting control is set. This determination threshold value is a determination threshold value for determining whether to start avoidance control for the side obstacle SM.

  In the present embodiment, when it is detected that an obstacle SM exists within a predetermined obstacle detection range set in advance in the side direction of the host vehicle MM, the vehicle is displaced outward from the lane line based on the lane line position. The predetermined lane width direction position is set as an obstacle distance X2obst, and the obstacle distance X2obst is set as a determination threshold (see FIG. 4). The obstacle distance X2obst is a value that is set assuming that the obstacle SM exists virtually. That is, processing is performed assuming that the obstacle SM exists at the obstacle distance X2obst displaced outward from the lane line in the lane width direction. The displacement amount from the lane marking that sets the obstacle distance X2obst may be set to “0”. In this case, the lane line and the obstacle distance X2obst are at the same position.

  Further, in the detection of the radar device 22L / R, when the distance from the obstacle SM can be detected with a predetermined accuracy, the lateral relative distance ΔO (= X0 + X2obst) between the host vehicle MM and the obstacle SM is detected. Then, using the lateral relative distance ΔO, (ΔXO−X0) is set to X2obst which is a predetermined lane width direction position. Then, the X2obst is used as the determination threshold as shown in FIG. When the lane marking is not taken into account (for example, when the target yaw rate is not corrected with the neutral yaw rate φ′path), the following may occur. That is, when the lateral relative distance ΔO becomes small, it becomes impossible to distinguish whether the relative driving between the host vehicle and the obstacle is for maintaining the route of the travel path or for changing the lane.

  In this case, the lateral distance (ΔXO−X0) between the lane marking and the obstacle is the obstacle distance X2obst. However, even if the lateral relative distance ΔO can be detected, the lateral distance (ΔXO−X0) itself does not need to be the obstacle distance X2obst. For example, (ΔXO−X0) −α may be used as the obstacle distance X2obst with some margin, or a value obtained by multiplying (ΔXO−X0) by a gain less than 1 may be used as the obstacle distance X2obst.

Here, an XY coordinate system is used in which the Y axis is taken in the direction along the road and the X axis is taken in the direction perpendicular to the road, that is, in the lane width direction. Then, the lateral position of the obstacle SM is detected on the X-axis coordinates. Based on the lateral position, the lateral relative distance ΔO is obtained.
Here, the obstacle detection range set as whether or not the obstacle SM is detected is set to be a predetermined vertical / horizontal position on the side of the host vehicle MM. The vertical position may be set such that the greater the relative speed at which the obstacle SM approaches the host vehicle MM, the wider the obstacle detection range.

Next, in step S83, the obstacle distance X2obst is adjusted.
That is, when the lane marking is not detected, the obstacle distance X2obst is adjusted.
When the lane marking flags CAMLOSTright and CAMLASTleft are “0”, that is, when the lane marking is detected, the obstacle distance X2obst itself is set for the determination obstacle distance X2obstvir as follows.
X2obstvir = X2obst

On the other hand, when the lane marking flags CAMLOSTright and CAMLASTleft are “1” or “2”, the determination obstacle distance X2 obstvir is obtained as follows. The case of “2” may be treated the same as “0”.
Here, when the lane marking flags CALMOSTright and CAMLASTleft are “0”, that is, every time a lane marking is detected in each control cycle, the width H of the traveling lane is updated and stored. As a result, the width H of the traveling lane immediately before the lane marking can no longer be detected can be acquired.

First, the minimum obstacle margin distance ΔX2 obstmin is obtained using a map as shown in FIG. 5 based on the width H of the traveling lane when the lane marking has been detected. That is, the minimum obstacle margin distance ΔX2 obstmin is calculated as the lane width H immediately before the lane marking becomes undetectable increases.
Next, as shown in the following formula, a select row of the obstacle distance X2obst and the minimum obstacle margin distance ΔX2obstmin is performed, and the smaller one is set as the determination obstacle distance X2obstvir.
X2obstvir = min (X2obst, ΔX2obstmin)

Next, in step S86, the position of the lane marking is adjusted.
When the lane marking flags CAMLOSTright and CAMLASTleft are “0”, that is, when the lane marking is detected, the obstacle distance XO itself is set for the determination lateral position XOvir as follows.
XOvir = XO
On the other hand, when the lane line flag flags CAMLASTright and CAMLASTleft are “1” or “2”, the lane line is changed to a virtual lane line as follows to obtain the determination lateral position XOvir. The case of “2” may be treated the same as “0”.
The lateral position X0 of the lane marking line immediately before the lane marking can no longer be detected is set as the lateral position X0 of the lane marking line used as the current reference.

First, a position correction amount X0add that is an offset amount outward from the lane marking is obtained. The position correction amount X0add is set to a larger value as the lost time is longer as shown in FIG.
Then, the determination lateral position XOvir is obtained by the following equation.
XOvir = X0 + X0add
The determination lateral position XOvir represented by this equation is a value that defines the position of the virtual lane marking, and is a value obtained by offsetting the reference lane marking to the outside by the position correction amount X0add. That is, the determination lateral position XOvir is a lateral distance between the host vehicle and the lane line or the virtual lane line.

The upper limit value XOvir_max of the determination lateral position XOvir is calculated based on the lane width H immediately before the upper limit value lane marking cannot be detected. That is, the upper limit value XOvir_max is calculated based on a map as shown in FIG. That is, the larger the lane width H, the larger the value. For example, the value is half of the lane width H.
If XOvir exceeds the upper limit value XOvir_max as in the following equation, the upper limit value XOvir_max is set as the determination lateral position XOvir.
XOvir = min (XOvir, XOvir_max)

Next, in step S88, the forward gazing point threshold value is corrected.
ΔOm expressed by the following formula is a forward gazing point threshold value. The forward gazing point threshold value ΔOm is a total value of the determination obstacle distance X2obstvir and the determination lateral position XOvir.
ΔOm = ΔX2obstvir + ΔX0vir
Here, the obstacle distance for determination X2 obstvir is a determination threshold value corresponding to a predetermined position in the lane width direction determined based on the lane line (virtual lane line). Further, the determination lateral position XOvir is a lateral distance from the own vehicle to the lane line or the virtual lane line.

When the lane marking is not detectable and the relative position of the obstacle with respect to the host vehicle can be detected, the forward gazing point threshold value ΔOm is corrected according to the position of the obstacle as follows. .
First, the relative vertical position information DIST_tale is acquired. The relative vertical position information DIST_tale is information on the vertical position of the obstacle with respect to the host vehicle. The vertical position is a relative position in the direction along the lane (Y-axis direction).
The relative vertical position information DIST_tale is set as follows. As shown in FIG. 8, the traveling direction of the host vehicle and the traveling direction of the host vehicle are negative. The distance between the front end position of the obstacle and the rear end position of the host vehicle. That is, the rear end position of the host vehicle is calculated as the origin.

Then, based on the relative vertical position information DIST_tale, when the rear end position of the host vehicle and the front end position of the obstacle are less than a predetermined value, and when the host vehicle and the obstacle overlap when viewed from the lane width direction, Corrects the forward gazing point threshold value ΔOm when the front part of the obstacle is located in front of the front part of the host vehicle.
That is, when the relative vertical position information DIST_tale is less than the predetermined value tale1, the forward gazing point threshold value ΔOm is corrected. That is, the forward gazing point threshold value ΔOm is corrected to decrease toward the lateral relative distance ΔO.

For example, the decrease correction is performed as follows.
That is, the lateral relative distance ΔO between the host vehicle and the obstacle is acquired.
Next, as shown in FIG. 9, with DIST_tale = predetermined value tale1, the current forward gaze point threshold value ΔOm is set, and the relative vertical position information DIST_tale = “− L” is set to be the horizontal relative distance ΔO. Sets a straight line LINE of inclination. Note that L is a distance in the Y-axis direction of the front gazing point (front gazing distance).
Then, using the function defined by the straight line LINE, a value corresponding to the current value of the relative vertical position information DIST_tale is set as the forward gaze point threshold value ΔOm ^* .
Then, the forward gazing point threshold value ΔOm is corrected by the following formula.
ΔOm = ΔOm ^*

Next, in step S90, control start determination is performed.
First, it is determined whether or not the obstacle SM exists. If there are no obstacles on the left and right, the obstacle avoidance control determination flag Fout_obst is set to OFF. Then, the process proceeds to step S100.
On the other hand, when the obstacle SM exists on at least one of the left and right sides, it is determined that the control is started when the following equation is satisfied for the lane marking line side where the obstacle exists. That is, the determination is made by comparing the forward gazing point threshold value ΔOm with the predicted future position ΔXb of the host vehicle MM.
ΔOm ≧ ΔXb

Then, when the above conditions are satisfied, it is determined that the control of the obstacle SM is started, assuming that there is a lane change operation to the obstacle SM side. When it is determined that the control for the obstacle SM is started, the obstacle avoidance control Fout_obst is set to ON. When the above condition is not satisfied, that is, when the future predicted position ΔXb is less than the determination threshold, the obstacle avoidance control determination flag Fout_obst is set to OFF.
Note that the future predicted position ΔXb is determined as ΔXbL / ΔXbR for each of the left and right sides of the vehicle and is individually determined.

Further, the target obstacle SM is not only set for the vehicle in the rear side direction of the host vehicle MM, but may also be a control target for the oncoming vehicle in front of the adjacent lane.
Here, when determining whether or not the future predicted position ΔXb is less than the determination threshold value, a hysteresis of F may be provided as ΔOm <ΔXb−F. That is, a dead zone may be set. That is, a dead zone may be provided between the control intervention threshold and the control end threshold.

  Also, Fout_obst can be set to ON when Fout_obst is OFF. Further, as a condition for enabling Fout_obst to be set to ON, a temporal condition may be added, for example, after a predetermined time has elapsed after Fout_obst is set to OFF. Further, when a predetermined time Tcontrol elapses after it is determined that Fout_obst is ON, the control may be terminated by setting Fout_obst = OFF.

Further, during the execution of the obstacle avoidance control, the control execution direction Dout_obst is determined based on the determination direction of the future predicted position. If the future predicted position is on the left, Dout_obst = LEFT is set, and if it is on the right, Dout_obst = RIGHT is set.
Here, when the anti-skid control (ABS), the traction control (TCS), or the vehicle dynamics control device (VDC) is operating, the obstacle avoidance control determination flag Fout_obst is set to OFF. This is to prevent the obstacle avoidance control from being activated while the automatic braking control is in operation.

In this determination method, threshold values are set for each of the yaw angle φ, the steering angle δ, and the steering speed δ ′ in the direction of the obstacle SM, and those threshold values approach the obstacle SM (or the obstacle SM). This is synonymous with setting so that the control start timing is more difficult to determine. This is because the target yaw rate φm ′ is determined by the relationship between the steering angle (and the vehicle speed) as is widely used in formulas.
Next, in step S100, an alarm generation process is performed.
Here, if it is determined in step S90 that the control start position has been reached, an alarm is generated.

  The alarm may be generated before the front gaze point based on the above-mentioned front gaze time reaches the control start position. For example, a predetermined gain Kbuzz (> 1) is applied so as to be longer than the forward gaze time Tt used for the detection in step S90. Then, an alarm is generated when it is determined that the forward gazing point calculated based on the equation (5) using (Tt × Kbuzz) has reached the control start position in step S90. Further, in step S90, it may be determined that the operation of the obstacle avoidance system is started, an alarm is generated, and then control is started after a predetermined time has elapsed.

Next, in step S110, a target yaw moment Ms is set.
When the obstacle avoidance control determination flag Fout_obst is ON, the target yaw moment Ms is calculated by the above formula as follows. When the obstacle avoidance control determination flag Fout_obst is OFF, the target yaw moment Ms is set to 0, and the process proceeds to the next step S120.
That is, when the obstacle avoidance control determination flag Fout_obst is ON, the target yaw moment Ms is obtained by the following equation.
Ms = K1recv × K2recv × ΔXs (6)
ΔXs = (K1mon · φ + K2mom · φm)

  Here, K1recv is a proportional gain (yaw inertia moment) determined from vehicle specifications. K2recv is a gain that varies according to the vehicle speed V. An example of the gain K2recv is shown in FIG. As shown in FIG. 10, for example, the gain K2recv has a large value in the low speed range, and when the vehicle speed V reaches a certain value, it has an inversely proportional relationship with the vehicle speed V. Thereafter, when the vehicle speed V reaches a certain value, it becomes a constant value with a small value. . The set gain K1mom is a value that is a function of the vehicle speed. The set gain K2mom is a value that is a function of the vehicle speed and the forward gaze time.

According to the equation (6), the target yaw moment Ms increases as the yaw angle φ with respect to the lane marking line or the yaw rate that is constantly generated by the steering increased by the driver increases.
Alternatively, the target yaw moment Ms may be calculated from the following equation (7). This equation (7) is synonymous with multiplying equation (6) by a gain K3 (= 1 / Tt ² ). The gain K3 is a gain that decreases as the forward gaze time Tt increases.
Ms = K1recv × ΔXb / (L × Tt ² ) (7)

  Using the above equation (7) indicating how much time T should be taken to control the yaw angle, the following is obtained. That is, by setting the control time T to coincide with the forward gaze time Tt, when the front gaze point at the control start timing is shortened, the time T for returning the vehicle is shortened. As a result, the control amount increases. That is, even when the control start timing is delayed, the control amount at the start of the control increases. Further, when the control start timing is made earlier, the control amount becomes smaller. As a result, it is possible to perform control with less discomfort according to the situation, regardless of the setting of the forward gazing point for the driver.

Note that the determination of Fout_obst predicts a future course change based on the steering information.
Here, in the case where the lane departure prevention control is provided in addition to the main control, one of the first and the second starts when the main control starts and when the lane departure prevention control starts (Fout_LDP = 1). Depending on whether the control is started first, the control that has been started first may be given priority, and the other control may not be performed until the control ends.

Next, in step S120, the process returns after calculating and outputting a command for generating the target yaw moment Ms for obstacle avoidance.
Here, in the present embodiment, an example in which a yaw moment is generated using a braking / driving force as means for generating a yaw rate Ms for avoiding an obstacle will be described below.
When the steering reaction force control device is used as a means for generating the yaw rate, the steering reaction force Frstr may be generated as Frstr = K × Ms.

When a steering control device is used as a means for generating the yaw rate, the steering angle STRθ may be given to the steering as a result obtained as STRθ = K × Ms ′.
Alternatively, the steering control device may be used as a means for generating the yaw rate, and the steering force (steering torque) may be calculated as STRtrg = K × Ms.

When the target yaw moment Ms is 0, that is, when the determination result that the yaw moment control is not performed is obtained, as shown in the following equations (8) and (9), the target brake hydraulic pressure of each wheel Psi (i = fl, fr, rl, rr) is set to the brake fluid pressures Pmf, Pmr.
Psfl = Psfr = Pmf (8)
Psrl = Psrr = Pmr (9)
Here, Pmf is the brake fluid pressure for the front wheels. Further, Pmr is the braking fluid pressure for the rear wheels, and is a value calculated based on the braking fluid pressure Pmf for the front wheels in consideration of the front-rear distribution. For example, if the driver is operating a brake, the brake fluid pressures Pmf and Pmr are values corresponding to the operation amount of the brake operation (master cylinder fluid pressure Pm).

On the other hand, when the absolute value of the target yaw moment Ms is larger than 0, that is, when the determination result that the obstacle avoidance control is started is obtained, the following processing is performed.
That is, the front wheel target braking hydraulic pressure difference ΔPsf and the rear wheel target braking hydraulic pressure difference ΔPsr are calculated based on the target yaw moment Ms. Specifically, the target braking hydraulic pressure differences ΔPsf and ΔPsr are calculated by the following equations (10) and (11).
ΔPsf = 2 · Kbf · (Ms × FRratio) / T (10)
ΔPsr = 2 · Kbr · (Ms × (1−FR ratio)) / T (11)
here,
FRratio: threshold value for setting T: tread Kbf, Kbr: conversion factors for the front and rear wheels when the braking force is converted to the braking hydraulic pressure.

In addition, the said tread T is handled as the same value before and behind here for convenience. Kbf and Kbr are determined by the brake specifications.
Thus, the braking force generated by the wheels is distributed according to the magnitude of the target yaw moment Ms. That is, a predetermined value is given to each target braking hydraulic pressure difference ΔPsf, ΔPsr, and a braking force difference is generated between the left and right wheels respectively. Then, the final target brake fluid pressure Psi (i = fl, fr, rl, rr) of each wheel is calculated using the calculated target brake fluid pressure differences ΔPsf, ΔPsr.

Specifically, when the execution direction Dout_str is LEFT, that is, when the obstacle avoidance control for the left obstacle SM is performed, the target braking hydraulic pressure Psi (i = fl, fr, rl, rr) is calculated.
Psfl = Pmf
Psfr = Pmf + ΔPsf
Psrl = Pmr
Psrr = Pmr + ΔPsr
(12)

Further, when the execution direction Dout is RIGHT, that is, when there is a tendency to deviate from the lane marking on the right side, the target braking hydraulic pressure Psi (i = fl, fr, rl, rr) of each wheel is calculated according to the following equation (13). ) Is calculated.
Psfl = Pmf + ΔPsf
Psfr = Pmf
Psrl = Pmr + ΔPsr
Psrr = Pmr
... (13)

According to the equations (12) and (13), the braking / driving force difference between the left and right wheels is generated so that the braking force of the wheels on the lane departure avoidance side is increased.
Further, here, as shown in the equations (12) and (13), the brake operation by the driver, that is, the target brake fluid pressure Psi (i = fl, fr) of each wheel in consideration of the brake fluid pressures Pmf, Pmr. , Rl, rr).
Then, the braking / driving force control unit 8 sets the target braking fluid pressure Psi (i = fl, fr, rl, rr) of each wheel calculated in this way as a braking fluid pressure command value to the braking fluid pressure control unit 7. Output.

(Operation / Action)
Based on the yaw angle φ, yaw angular velocity φm, and the like, which are the traveling state of the host vehicle, the host vehicle predicted position ΔXb is obtained as the future position of the host vehicle after the forward gaze time T.
When the predicted vehicle position ΔXb on the side where the obstacle SM is detected is equal to or greater than the obstacle distance X2obst, which is a determination threshold outside the lane line, with reference to the lane line, assistance for obstacle avoidance Control is started (see FIG. 4). When it is determined that the assist control is started, a target yaw moment Ms is calculated as a control amount based on the host vehicle predicted position ΔXb, and the braking / driving force is controlled so as to generate the target yaw moment Ms. As a result, the host vehicle is controlled in a direction that prevents access to an obstacle.

Here, as the time during which the lane marking cannot be detected continues, the probability of the travel route decreases. In response to this, when the lane line cannot be detected, a virtual lane line is set in which the position of the lane line is displaced outward from the lane line immediately before the lane line can no longer be detected. The control start is determined based on the virtual lane marking.
Here, when the lane marking is not detected in the state of FIG. 4, it can be assumed that the forward route is a curved road (indicated by a one-dot chain line in FIG. 8) as shown in FIG. In other words, even if the relative lateral distance between the host vehicle and the obstacle is shortened, if the lane marking cannot be detected, it cannot be accurately recognized whether the lane is changed or the front is a curved road.

At this time, in this embodiment, the position of the determination threshold is offset to the outside of the lane by changing the lane marking used as a reference for determining the start of control to the virtual lane marking. As a result, the adjustment is made in a direction that makes it difficult to start the control.
As a result, even if the lane marking cannot be detected, the start / end of control is prevented more than necessary. That is, by reducing the frequency of control intervention, it is possible to suppress changes in vehicle behavior due to side obstacle avoidance control, and to reduce discomfort to the driver.

In addition, since only the control start timing is delayed, support control for necessary side obstacles is performed.
In other words, it is possible to start control when the driver intentionally steers in the direction of the obstacle SM while suppressing the start of control with an uncomfortable feeling.
Here, the radar device 22L / R constitutes an obstacle detection means. The imaging unit 13 constitutes a lane detection unit. The steering angle sensor 19 constitutes a steering input detection means.
Step S70 constitutes future position prediction means 8A. Steps S80 to S90 constitute avoidance control start detection means 8B. Steps S100 to S210 constitute the obstacle avoidance control means 8C. Step S86 constitutes the lane marking adjustment means 8Ba. The determination threshold corresponds to a predetermined lane width direction position.

(Effect of this embodiment)
(1) When the avoidance control start detecting means 8B detects the obstacle, the judgment threshold value (predetermined based on the future position of the host vehicle and the lane marking nearer the obstacle) The start of control is detected based on the relationship with the position in the lane width direction. At this time, if it is determined that the lane line can no longer be detected while the lane line is being detected, the lane line adjustment means sets the virtual lane line outside the detected lane line position in the lane width direction position. Then, the reference lane marking is changed to the virtual lane marking. That is, when it becomes impossible to detect a lane line during traveling, the lane line used as a reference when determining the start of control is shifted outward in the vehicle width direction.

  This makes it difficult to determine that the control is started when it becomes impossible to detect a lane line during traveling. That is, the control start timing is delayed. As a result, the start / end of control is prevented more than necessary, and support control for necessary side obstacles is performed. That is, it is possible to start control when the driver steers intentionally in the direction of the obstacle while suppressing the start of control with a sense of incongruity.

As described above, even when the lane markings cannot be detected, it is possible to appropriately perform the support control for the side obstacle.
Here, in the state in which the lane marking is detected, the lane marking on the road is detected to determine the future position of the vehicle with respect to the obstacle. Thus, by using the lane marking as a reference, it is possible to determine whether or not the vehicle is properly heading toward the obstacle even when steering on a curved road. This makes it possible to appropriately perform support control for avoiding a side obstacle even on a curved road.

(2) The lane marking adjustment means displaces the virtual lane marking in the vehicle width direction outward direction as the duration of the state in which the lane marking cannot be detected becomes longer.
Here, as the recognition status of the lane markings continues to deteriorate, the probability of the route information decreases. In addition, the probability that the traveling direction of the host vehicle changes is also increased. In view of this, as the recognition status of the lane markings continues to deteriorate, the control start determination is more difficult to be performed. As a result, it is possible to further reduce that the control is performed at a timing at which the driver feels uncomfortable.
For example, when steering is performed to maintain a traveling path such as a curve, unnecessary control is less likely to be entered, and the uncomfortable feeling can be reduced.

(3) The lane marking adjustment means sets the upper limit of the offset amount for displacing the virtual lane marking outward in the vehicle width direction based on the lane width of the lane marking detected by the lane marking detection means. The larger the lane width, the larger the value. That is, the maximum change width of the virtual lane marking is limited based on the width of the travel path estimated by the lane width before the lane marking cannot be detected.
As a result, it is possible to perform control at a timing that is close to the driver according to the lane width. In other words, when driving with a wide distance between lane markings on a road with a large distance between obstacles, or when driving with a narrow distance between lane markings, the timing according to the lane width Control can be started with.

(4) The lane marking adjustment means displaces the virtual lane marking to the outside in the vehicle width direction as the vehicle speed when the detected lane marking cannot be detected is decreased.
The lower the vehicle speed, the greater the vibration of the vehicle, that is, the yaw rate tends to increase, and the control intervention and termination may be repeated unnecessarily. On the other hand, when the control intervention element is increased, unnecessary control intervention can be suppressed.
That is, even in a situation where the vehicle speed is low and a smaller turn may occur, it is possible to reduce the control being performed at a timing that makes the driver feel uncomfortable.

(5) The avoidance control start detection means sets the determination threshold value on the outer side in the lane width direction from the lane line according to the distance in the lane width direction from the lane line to the obstacle detected by the lane line detection means. Set to position.
Thereby, the control start timing can be set depending on whether or not the host vehicle interferes with the obstacle after the forward gaze time Tt.
Here, the actual lateral distance from the lane marking to the obstacle varies depending on the traveling scene.

(6) If it is determined that the lane marking cannot be detected while the lane marking is detected by the lane detecting means, the lane is determined based on the lane width when the lane marking is detected by the lane marking detecting means. The determination threshold is set to the outer side in the lane width direction as the width increases.
Based on the current lane line width information estimated based on the lane line width before the lane line becomes invisible when the lane line becomes invisible, the standard lane line is determined based on the distance between the lane line and the obstacle. Set the distance between the line and the obstacle.
As a result, even when an obstacle happens to be away from the lane line, it can be assumed that the vehicle is traveling at the standard traveling position, and control can be started at standard timing. .

(7) When the vehicle lane marking adjustment means approaches the vehicle and the obstacle so as to overlap at least when viewed from the vehicle lane width direction, the lane width direction distance between the vehicle and the determination threshold value Correct to approach the distance in the lane width direction.
When the vertical approach situation with the obstacle is close to the own vehicle, the situation is such that the linkage between the own vehicle and the side obstacle becomes strong. In view of this, the lane width direction between the host vehicle and the determination threshold is set to the distance in the lane width direction between the actual host vehicle and the obstacle according to the situation where the obstacle approaches the host vehicle. Correct so that the distance gets closer.
As a result, the control start timing approaches a value corresponding to the actual obstacle position. That is, it becomes easier to start the control at a timing that matches the driver's sense, and it is possible to reduce the sense of discomfort.

It is a schematic block diagram of the apparatus which concerns on embodiment based on this invention. It is a figure explaining the structure of the control unit which concerns on embodiment based on this invention. It is a figure explaining the process of the control unit which concerns on 1st Embodiment based on this invention. It is a conceptual diagram which shows the relationship between the own vehicle and an obstruction. It is a figure which shows the example of the relationship between the lane width H and the minimum obstacle margin distance (DELTA) X2 obstmin. It is a figure which shows the example of the relationship between elapsed time (lost time) and position correction amount XOadd. It is a figure which shows the example of the relationship between the lane width H and XOvir_max. It is a figure which shows the example in case the front is a curve road when a lane marking cannot be detected. It is a figure explaining the method of correction | amendment of a front gaze point threshold value. It is a figure explaining gain K2.

Explanation of symbols

8 Braking / driving force control unit 8A Future position prediction means 8B Avoidance control start detection means 8Ba Lane lane marking adjustment means 8C Obstacle avoidance control means 13 Imaging unit 19 Steering angle sensor 22L / R Radar device CAMLASTright, CAMLASTleft Lane lane marking flag dDISTobst Relative Vertical speed DIST Relative vertical position information DISTobst Relative vertical position Fout Obstacle avoidance control determination flag H Lane width L Forward gaze distance LINE Straight line LTr, LTl Lost time Ms Target yaw moment Tt Forward gaze time X0add Position correction amount X2obst Obstacle distance (determination Threshold, predetermined lane width direction position)
Obstacle distance for X2 obstvir judgment (judgment threshold, predetermined lane width direction position)
Xfront Lateral displacement XO Obstacle distance XOadd Position correction amount (offset amount)
Upper limit value XOvir_max Upper limit value XOvir Judgment lateral position βfront Curvature δ Steering angle ΔO Lateral relative distance ΔOm Forward gaze threshold ΔX2 obstmin Minimum obstacle margin distance ΔXb Future predicted position

Claims (8)

Obstacle detection means for detecting obstacles present on the side of the vehicle;
Lane detection means for acquiring information around the vehicle and detecting a lane division line of a running lane;
Steering input detection means for detecting the steering input of the driver;
Future position prediction means for predicting the future position of the host vehicle after a predetermined time based on the steering amount detected by the steering input means;
When the obstacle detection means detects an obstacle, the relationship between the future position of the host vehicle predicted by the future position prediction means and a predetermined lane width direction position determined in advance with reference to the lane marking Based on the avoidance control start detection means for detecting the control start,
When the avoidance control start detection means detects the control start, the avoidance control start detection means includes an obstacle avoidance control means for controlling the host vehicle so as to prevent an approach to the obstacle,
The avoidance control start determination means includes
If it is determined that the lane marking cannot be detected while the lane marking is being detected by the lane detection means, the virtual lane marking is positioned outside the lane marking position detected by the lane marking detection means at a position outside the lane width direction. And a lane marking adjustment means for changing the lane marking used as a reference for the predetermined lane width direction position to the virtual lane marking.   2. The vehicle according to claim 1, wherein the lane marking adjustment means displaces the virtual lane marking outward in the vehicle width direction as the duration of the state in which the lane marking cannot be detected becomes longer. Driving assistance device.   The lane marking adjustment means is configured to set an upper limit of an offset amount for displacing the virtual lane marking outward in the vehicle width direction based on the lane width of the lane marking detected by the lane marking detection unit. The vehicle driving support device according to claim 1 or 2, wherein the vehicle driving support device is limited to a larger value as the lane width is larger.   The lane marking adjustment means displaces the virtual lane marking outward in the vehicle width direction as the vehicle speed when the detected lane marking cannot be detected is lower. Item 4. The vehicle driving support device according to any one of Item 3.   The determination threshold value is set at a position outside the lane line in the lane width direction according to the distance in the lane width direction from the lane line to the obstacle detected by the lane line detection unit. The vehicle driving support device according to any one of claims 1 to 4.   If it is determined that the lane marking cannot be detected while the lane marking is detected by the lane detecting means, the lane width is calculated based on the lane width when the lane marking is detected by the lane marking detecting means. 6. The vehicle driving support device according to claim 5, wherein the predetermined position in the lane width direction is set to the outside in the lane width direction as the value increases.   The lane division line adjusting means is based on the approach situation between the host vehicle and the obstacle in the direction along the lane, and when approaching the host vehicle and the obstacle at least as viewed from the lane width direction, The vehicle according to claim 5 or 6, wherein the distance in the lane width direction between the position in the lane width direction is corrected so as to approach the distance in the lane width direction between the host vehicle and the obstacle. Driving assistance device.   When an obstacle on the side of the host vehicle is detected, the future position of the host vehicle after a predetermined time is predicted based on the driver's steering input, and the predicted future position of the host vehicle is determined based on the lane marking. When the vehicle is controlled so as to prevent the vehicle from approaching the obstacle, the obstacle starts when the vehicle reaches the position in the lane width direction or is determined to be on the obstacle side of the predetermined lane width direction position. A vehicle driving support method, wherein when the lane marking on the side close to the vehicle cannot be detected or the recognition accuracy is poor, the lane marking used as a reference when deciding the start of the control is displaced outward in the lane width direction. .

Images