JP5315798B2 - Vehicle driving support apparatus and vehicle driving support method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a driving support for a vehicle allowing reduction of a burden in driving operation for obstacle avoidance by a driver. <P>SOLUTION: When deciding that there is a possibility that an obstacle SB present in advance of a travel direction of one's own vehicle SM contacts with the own vehicle SM within a prescribed time, an avoidance direction advantageous to contact avoidance with the obstacle SB is determined. Furthermore, the own vehicle SM is controlled to guide the own vehicle SM to the advantageous avoidance direction in a range of an allowable lateral movement amount set in the own traveling path. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

  The present invention relates to a vehicle driving support for performing driving support for guiding a host vehicle so as to avoid contact with an obstacle.

As a conventional vehicle driving support device for the purpose of avoiding contact with an obstacle, there is a technique described in Patent Document 1, for example.
In the prior art as described in Patent Document 1, a yaw moment for avoiding an obstacle is generated when it is detected that the driver performs a steering operation in order to avoid an obstacle ahead of the host vehicle. To control the host vehicle. Thus, the host vehicle is prevented from colliding with an obstacle.
JP 2007-22232 A.

The vehicle driving assistance for avoiding a collision with an obstacle as described above is so close that the host vehicle may collide with the obstacle, and when the driver performs avoidance steering to avoid the collision. carry out. In other words, driving assistance is provided for the first time when the level of urgency is high.
For this reason, when the driver performs avoidance steering, the behavior of the host vehicle may increase. In addition, since driving assistance is performed in response to an operation for avoiding an obstacle by the driver, there is a possibility that the burden of the steering operation for avoiding the obstacle by the driver may not be reduced.
The present invention focuses on the above points, and an object of the present invention is to provide vehicle driving assistance that can reduce the burden of driving operation for avoiding obstacles by the driver.

  In order to solve the above-mentioned problem, the present invention has a possibility that an obstacle present in the forward direction of the own vehicle may come into contact with the own vehicle within a predetermined time before the driver actually performs the avoidance operation. If it is determined, an avoidance direction advantageous for avoiding contact with the obstacle is determined, and the host vehicle is guided in the advantageous avoidance direction within a range of an allowable lateral movement set in the host vehicle travel path.

  According to the present invention, it is possible to guide the host vehicle in advance in a direction advantageous for obstacle avoidance before the driver actually performs the avoidance operation. As a result, the risk of contact with the obstacle increases and the vehicle is guided to a state where it is easy to avoid when the driver actually performs the steering operation for avoidance. Therefore, the burden of avoidance operation by the driver can be expected to be reduced.

(First embodiment)
Next, a first embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a schematic diagram of a device configuration of a vehicle equipped with the vehicle driving support device of the present embodiment.
(Constitution)
In the example of this vehicle, as shown in FIG. 1, a total of four wheels 1FR to 1RL including left and right front wheels and left and right rear wheels are provided as wheels. This vehicle is an example of a front wheel drive type in which the output of the internal combustion engine is transmitted to the front wheels, and the front wheels are steered wheels. Then, the steering system controls the steering amount of the front wheels. However, other vehicle systems may be used.

The steering system includes a steering actuator 2. The steer actuator 2 includes hydraulic or electric power steering as a drive unit, and controls the steer torque applied to the steered wheels based on an external signal.
The host vehicle SM is equipped with various sensors as follows. The sensors are a stereo camera 3, an acceleration sensor 4, a yaw rate sensor 5, a steering angle sensor 6, and wheel speed sensors 7FR to 7RL. These sensors 3 to 7 output detection information to the avoidance guidance controller 9.

  The stereo camera 3 is provided in front of the vehicle interior. The stereo camera 3 detects information ahead of the host vehicle SM by imaging the obstacle SB in front of the host vehicle SM and the road environment in front of the host vehicle SM. Here, with the stereo configuration, the relative distance to the detected object can be detected. Further, by extracting features related to a road boundary such as a white line by image processing, it is possible to acquire boundary information of the own vehicle traveling road and road boundary information ahead of the own vehicle SM including other than the own vehicle traveling road.

The acceleration sensor 4 detects acceleration in a specific direction generated in the vehicle. The acceleration sensor 4 is constituted by a device using, for example, a piezoelectric element. For example, the acceleration generated in the vertical direction and the horizontal direction of the vehicle body is detected. Then, by integrating the detected acceleration signal at a predetermined time interval, the vehicle body longitudinal and lateral speeds can also be detected.
The yaw rate sensor 5 detects a rate signal in the yaw direction generated at the center of gravity of the vehicle. The yaw rate sensor 5 is constituted by a known device using, for example, a crystal resonator or a semiconductor.
Wheel speed sensors 7FR to 7RL detect the rotation speed of each wheel. The number of rotations is detected by a pulse signal generated according to the wheel rotation of each wheel.

The steering angle sensor 6 detects the rotation angle of steering. For example, an encoder is installed on the pinion side of a front wheel steering mechanism configured by a rack-pinion system, and the rotation angle of steering is detected by the encoder.
A buzzer 8 is provided. The buzzer 8 is a device that generates a warning sound based on a signal from the avoidance guidance controller 9.
The avoidance guidance controller 9 is composed of a microprocessor. The microprocessor is an integrated circuit including an A / D conversion circuit, a D / A conversion circuit, a central processing unit, a memory, and the like. Then, according to the program stored in the memory, the turning amount is calculated based on the information of the signals detected by the various sensors and is output to the steer actuator 2.

Next, the avoidance guidance controller 9 will be further described.
As shown in FIG. 2, the avoidance guidance controller 9 includes a host vehicle motion state acquisition unit 9A, a vehicle forward information acquisition unit 9B, an avoidance determination unit 9C, an avoidance operation estimation unit 9D, a host vehicle guidance amount calculation unit 9E, and a host vehicle guidance. Control command output means 9F, obstacle avoidance control means 9G, and emergency stop control means 9H are provided.
The own vehicle motion state acquisition means 9A receives detection signals from the acceleration sensor 4, the yaw rate sensor 5, the wheel speed sensors 7FR to 7RL, and the steering angle sensor 6. And the own vehicle movement state acquisition means 9A acquires the information showing the movement state of the own vehicle SM by processing those detection signals in an integrated manner. That is, the motion state of the host vehicle SM is detected.

  The vehicle front information acquisition unit 9B performs image processing on the image information captured by the stereo camera 3. That is, the vehicle front information acquisition unit 9B detects road boundaries such as white lines and curbs and obstacles SB based on information extracted by image processing. And the information showing the motion state of the obstacle SB ahead of the own vehicle SM and the information on the road boundary are acquired. Further, the distance to the obstacle SB in front of the host vehicle SM is also detected.

  The avoidance determination unit 9C may come into contact with the detected forward obstacle SB within a predetermined time (for example, within 5 seconds) based on the acquired information of the host vehicle motion state acquisition unit 9A and the vehicle forward information acquisition unit 9B. It is determined whether or not. That is, the movement prediction of the own vehicle SM from the movement state (traveling direction, etc.) of the own vehicle SM acquired by the own vehicle movement state acquisition means 9A and the movement prediction of the obstacle SB acquired by the vehicle forward information acquisition means 9B are performed. Then, it is determined whether or not contact is possible within a predetermined time. Note that it is possible to determine whether or not contact is possible based on one of the motion prediction of the host vehicle SM or the motion prediction of the obstacle SB. That is, for example, the obstacle SB may be determined as being stopped, or may be determined as being moved at a predetermined constant speed. Alternatively, the determination may be made on the assumption that the host vehicle SM moves straight at a constant speed.

The avoidance operation estimation means 9D estimates whether or not a driving operation for avoiding the obstacle SB has been performed. For example, it is estimated that the driving operation for avoiding the obstacle SB is performed when the steering operation more than a predetermined value is detected after the obstacle SB is detected.
The own vehicle guidance amount calculation means 9E operates when the avoidance determination means 9C determines that the obstacle SB has a risk of contact. As shown in FIG. 3, the own vehicle guidance amount calculation means 9E includes an avoidance direction determination means 9Ea, own vehicle control amount calculation means 9Eb, and an alarm signal output means 9Ec.

The avoidance direction determination means 9Ea determines an avoidance direction that is advantageous for avoiding contact with the obstacle SB within the road boundary.
The own vehicle control amount calculation means 9Eb calculates the own vehicle guidance amount for guiding the own vehicle SM in the advantageous avoidance direction within the range of the allowable lateral movement set in the travel path.
Here, the range of the allowable lateral movement amount may be the own vehicle traveling path itself. Alternatively, it may be a range that is inside by a predetermined threshold from the vehicle traveling road. Alternatively, a range that protrudes to the left and right by a predetermined range from the lateral width of the host vehicle SM may be used with the host vehicle SM as a reference.

The alarm signal output means 9Ec transmits a signal to the buzzer 8 to notify the driver of the danger by sound when it is necessary to guide the vehicle outside the traveling road.
In addition, the host vehicle guidance control command output unit 9F calculates a command for the host vehicle braking amount calculation unit 9Eb to realize the host vehicle guidance amount, and outputs the command to the steering actuator 2.
The avoidance guidance controller 9 includes an obstacle SB avoidance control unit and an emergency stop control unit at the time of steering detection as other braking / driving control units.

Next, the process of the part related to the vehicle driving support of the present case in the process of the avoidance guidance controller 9 will be described with reference to FIG.
Here, in order to make the contents of each process easy to understand, a scene as shown in FIG. 5 is assumed as an example. In the scene shown in FIG. 5, the host vehicle SM is traveling on a straight road on one lane on one side, and a stopped vehicle as an obstacle SB is placed on the shoulder in front of the host vehicle SM traveling lane (own vehicle traveling path). It is a scene that exists. That is, the case where the stop vehicle is on the left side with respect to the host vehicle SM is illustrated.

The processing of the avoidance guidance controller 9 operates at a predetermined sampling period.
First, in step S100, a camera image is read from the stereo camera 3 and a detection signal is read by each sensor. These processing results are stored in a memory in the microprocessor.
Then, in order to describe the motion state, a coordinate system is determined from the image image captured by the camera. Furthermore, the information on the host vehicle SM, the front obstacle SB, and the road boundary is associated with the unified coordinate values.

  In this embodiment, as shown in FIG. 5, the coordinate system is set such that the road traveling direction is the X axis and the direction perpendicular to the X axis is the Y axis. It is not necessary to align the X axis with the center line. With respect to the coordinates set in this way, the position information of the center of gravity of the host vehicle SM is associated with (Xv, Yv), and the position information of the rear center of the obstacle SB is associated with (Xb, Yb). Further, the Y coordinate of the road boundary is expressed as YL and YR respectively on the left and right, and the Y coordinate of the center line is expressed as YC.

Further, the moving speeds of the host vehicle SM and the obstacle SB in the coordinate system thus determined are calculated.
In the example of FIG. 5, since the host vehicle SM is traveling in the X-axis direction, the host vehicle speed can be described as (Vvx, Vvy) = (Vv, 0) based on the detection signal of the wheel speed sensor. Further, since the obstacle SB is stopped, the obstacle SB speed can be expressed as (Vbx, Vby) = (0, 0) by the difference from the previously acquired obstacle position information.

Next, in step S200, based on the coordinate information set in step S100, it is determined whether or not there is a possibility that the host vehicle SM will contact the obstacle SB from the current time to a predetermined time in the future. If it is determined that the host vehicle SM may come into contact with the obstacle SB, the process proceeds to step S300. On the other hand, when it is determined that there is no possibility that the host vehicle SM is in contact with the obstacle SB, the vehicle returns.
In step S300, it is determined whether the driver is performing a steering operation for avoidance. Here, the state that proceeds to step S300 is a state in which the host vehicle SM may come into contact with the obstacle SB. Therefore, it is possible to estimate that the driver's steering operation exceeding a predetermined value is a steering operation for avoidance.

The determination of the avoidance steering operation is performed as follows, for example.
That is, the difference Δδstr between the current steering angle signal acquired from the steering angle sensor 6 and the previously acquired steering angle signal is calculated. Then, the absolute value of the difference Δδstr is compared with the threshold value ΔδTH as in the following equation (1).
| Δδstr | <ΔδTH (1)
If the absolute value of the difference Δδstr is less than the threshold ΔδTH, it is determined that the driver is not trying to avoid the front obstacle SB, and the process proceeds to step S400. On the other hand, if the absolute value of the difference Δδstr is equal to or greater than the threshold value ΔδTH, the driver is willing to avoid the forward obstacle SB, and determines that the driver has determined the avoidance direction, and other avoidance control (steering) The process proceeds to the obstacle SB avoidance control means 9G) at the time of detection. The other avoidance control means 9G is, for example, control as in the above-described prior art, steering control for avoidance, or avoidance control for applying automatic braking according to the estimated contact time with the obstacle SB.

Here, in determining whether or not the steering operation is to avoid the obstacle SB, the determination may be made based on the current steering angle δstr instead of using the difference Δδstr. Alternatively, the determination based on the steering angle δstr and the determination based on the difference Δδstr may be combined. Further, the threshold value ΔδTH may be changed according to the speed of the host vehicle SM. Alternatively, the determination may be made based on the steering speed or the like.
Next, in step S400, an avoidance direction advantageous for avoiding the obstacle SB is determined.
That is, considering whether to avoid the obstacle SB in the right direction or in the left direction, it is determined whether the avoidance in the direction is larger in avoidance.

Next, in step S500, a steering guidance amount for guiding the host vehicle SM in the advantageous avoidance direction determined in step S400 is determined.
Next, in step S600, a command value corresponding to the steering guidance amount is output to the steer actuator 2. Then return.
Next, an example of the presence / absence determination process (avoidance determination unit 9C) of the possibility of the host vehicle SM coming into contact with the obstacle SB corresponding to the process of step S200 will be described with reference to FIG.
In the present embodiment, it is determined whether there is a risk of contact between the current time (t = 0) and Tcj seconds ahead. It shows using FIG. 6 and FIG.

  First, in step S210, based on the position (Xv, Yv) of the own vehicle at the current time (t = 0) and the size of the own vehicle SM stored in the memory in advance, the position information (own vehicle SM) Vehicle SM area information) is calculated. Further, the position information of the four corners of the obstacle SB (information on the obstacle SB area) is calculated from the obstacle position (Xb, Yb) and the size of the obstacle SB estimated from the information of the camera image. Then, from these pieces of information, the host vehicle SM area and the obstacle SB area in the coordinate system determined in step S100 are calculated.

Here, the left front of the vehicle body of the host vehicle SM is (X _v1 , Y _v1 ), the front right is (X _v2 , Y _v2 ), the left rear is (X _v3 , Y _v3 ), and the right rear is (X _v4 , Y _v4 ). Is written. In addition, the left front of the obstacle SB is represented as ( _Xb1 , _Yb1 ), the right front is represented as ( _Xb2 , _Yb2 ), the left rear is represented as ( _Xb3 , _Yb3 ), and the right rear is represented as ( _Xb4 , _Yb4 ). .
In the following description, regarding the subscript of the variable representing the coordinates of the host vehicle SM or the obstacle SB, 1 is the left front, 2 is the right front, 3 is the left rear, and 4 is the right rear. To do.
These coordinates can be expressed as the following equations (2) to (5). However, the length from the center of gravity of the own vehicle SM to the front side is dv _xf , the length from the center of gravity of the own vehicle SM to the rear side is dv _xr , the width of the own vehicle SM is dv _y , and the vertical width of the obstacle SB is db _x , the width of the obstacle SB and db _y.

Then, the own vehicle SM area Sv (0) and the obstacle SB area Sb (0) are specified based on the position information at the current time (t = 0) that can be expressed by the equations (2) to (5). To do. In the present embodiment, the region in this case indicates an area in the four corners.
Next, in step S220, as shown in the following equation (6), whether or not there is an area where the own vehicle SM area Sv (t) and the obstacle SB area Sb (t) overlap at the time t seconds from the current time. Determine whether.

If the expression (6) is true, that is, if there are overlapping areas, it is determined that there is a risk of contact between the host vehicle SM and the obstacle SB after t seconds, and the process proceeds to step S300. On the other hand, if the expression (6) is false, it is determined that the risk of contact is low by t seconds and the process proceeds to step S230.
In step S230, first, the time t is advanced by tcj seconds as in the following equation (7).
t ← t + tcj (7)
Subsequently, it is determined whether or not the value at time t exceeds Tcj, as in the following equation (8).
t <Tcj (8)
If the value of time t is less than Tcj, the process proceeds to step S240 in order to continue predicting future danger.

On the other hand, if the value of time t is equal to or greater than Tcj, it is determined that the risk of contact with the obstacle SB is low between the current time (t = 0) and the predetermined time Tcj, and the process returns.
In step S240, the host vehicle SM area and the obstacle SB area at time t seconds are calculated. Then, the process returns to step S220. Here, two methods are shown.
The first process of step S240 calculates the own vehicle SM area and the obstacle SB area at time t seconds from the speed information of the own vehicle SM and the obstacle SB at the current time.
For example, in the scene shown in FIG. 4, since the host vehicle SM is traveling straight and the obstacle SB is stopped, each corner coordinate can be expressed as follows.

And based on the positional information which can be illustrated with (9)-(12) Formula, the own vehicle SM area | region Sv (t) and the obstruction SB area | region Sb (t) in time t second are calculated.
Next, the second processing method of step S240 will be described.
That is, when the movement prediction of the host vehicle SM is performed more precisely than the above-described trajectory prediction method (first processing method) based on the speed information at the current time, the following movement of the host vehicle SM is performed. A method for predicting the trajectory of the host vehicle SM using a vehicle model that expresses
First, an example of a vehicle model is shown below.

Here, as shown in FIG. 7, θ, ν, β, γ, and δw respectively represent the own vehicle SM facing yaw angle, speed, slip angle, yaw rate, and front wheel turning angle. Then, together with the own vehicle position (x _v , y _v ), it is treated as a state variable of the own vehicle SM in the present embodiment.
Further, δs and αlon represent the steering angle and the vehicle longitudinal acceleration / deceleration, respectively. And it handles as a vehicle input in this embodiment.
M, I, Lf, Lr, Ng, and Ts are the own vehicle SM parameters in this embodiment, and are respectively the vehicle mass, the yaw inertia, the distance from the center of gravity to the front wheel axis, the distance from the center of gravity to the rear wheel axis, and the steering gear ratio. , Represents the steer time constant.

(15) Fy _f and Fy _{r in the} equation (16) represent the front and rear wheel tire lateral forces.
Here, the tire lateral force generally has a non-linear characteristic with respect to the tire slip angle. However, when the tire slip angle is small, it can be assumed to be linear.
In the normal running region as in the scene of FIG. 5, the tire slip angle can be regarded as small. Therefore, if the slip angles of the front and rear wheels are α _f and α _r , and the cornering powers of the front and rear wheels are K _f and K _r , respectively, they can be expressed as follows.

  When the vehicle model represented by the above equations (13) to (23) is used, the motion state of the vehicle used in this embodiment is a seven-dimensional vector. An input for operating the vehicle is a two-dimensional vector. These can be expressed as follows:

Next, a method for acquiring the initial state vector in the vehicle model shown in the equation (24) will be described.
x _v and y _v is calculated in the stage of step S100, the Xv and Yv.
γ can be obtained from the output of the yaw rate sensor 5.
θ is detected by integration of the yaw rate sensor 5 signal on the basis of the vehicle posture at a certain time. Or it determines from the angle which the own vehicle SM and the own vehicle travel path make from image information.
δw can be obtained by dividing the steering angle δs by the gear ratio Ng.
β can be expressed by the following equation, where ν _X is the vehicle vertical velocity and ν _Y is the horizontal velocity.

Then, according to the equation (27), β is detected by ν _{X being} a value obtained by integrating the acceleration in the vehicle vertical direction detected by the acceleration sensor 4 attached to the vehicle, and ν _Y being a value obtained by integrating the acceleration in the vehicle lateral direction. What is necessary is just to obtain | require by using. Assuming that β is very small, it can be approximated as ν = ν _X.
Through the above processing, the initial state of the vehicle model can be acquired.
If the input steering angle δs and the longitudinal acceleration αlon shown in the equation (25) are continuously given to the vehicle model in the initial state, the locus of the center of gravity of the vehicle can be obtained.
Using the barycentric coordinates (x _v (t), x _v (t)) and the vehicle posture θ (t) at the future time t, the following is expressed.

Using the equations (28) to (31), the own vehicle SM area Sv (t) is calculated and contact determination is performed. In this case, the reliability can be expected to be higher than that of the own vehicle SM area calculated by the equations (9) and (10).
Next, an example of a determination process (avoidance direction determination means 9Ea) that determines whether the avoidance in which direction is larger in the avoidance direction, which is the process in step S400, will be described next.
The time when the expression (6) in the contact determination in the process of step S200 is false, that is, the time t when the host vehicle SM contacts the obstacle SB is defined as Ttc. Then, the moving position of the obstacle SB after Ttc seconds is (xb ′ (Ttc), yb ′ (Ttc)).
At this time, the distance from the obstacle SB to the left and right road boundaries can be expressed by the following equation (32).
DL = | yb '(t) -YL |
DR = | yb '(t) -YR |
... (32)

In the situation shown in FIG. 5, since the obstacle position does not change, (xb ′ (Ttc), yb ′ (Ttc)) = (Xb, Yb) remains.
Here, in the above equation (32), the distance between the center point of the obstacle SB and the road boundary is used, but the distance between the right and left coordinates of the obstacle SB and each adjacent road boundary. May be used individually.
And about the distance DL and distance DR, the determination shown to following (9) Formula is performed, respectively.
DR <DTH
And DL <DTH
... (33)
Here, DTH is a threshold value for determining whether there is an avoidance margin.

If both the distance DL and the distance DR are less than the threshold value DTH, it is determined that there is no avoidance margin in the left-right direction. When it is determined that there is no avoidance margin in the left-right direction, the function shifts to the emergency stop control means 9H, which is another control means, or an alarm system.
On the other hand, if both the distance DL and the distance DR are not less than the threshold value DTH, there is an avoidance margin in at least one of the left and right directions. In this case, the left and right avoidance margins are compared. The distance DL and the distance DR correspond to the left and right avoidance margins.

Thus, it is determined which direction has a higher avoidance margin and is advantageous for avoidance. The judgment formula is shown in the following formula (34).
DR <DL (34)
When the distance DL is larger, the avoidance in the left direction has a larger margin. On the other hand, when the distance DR is larger, the avoidance to the right avoidance has a larger margin.
Thus, an avoidance direction advantageous for avoiding the obstacle SB is determined.
In the situation shown in FIG. 5, a result that DR is larger than DL can be expected.
Here, the avoidance direction that is the processing purpose of step S400 is determined, and the margin (DR or DL) of the selected direction is replaced with DVBR.
The above is the avoidance direction determination method.

Next, the process of determining the steering guidance amount in the advantageous avoidance direction determined in step S400 (the own vehicle control amount calculation means 9Eb), which is the process in step S500, will be described with reference to FIG.
First, in S510, how much it is necessary to move laterally in the avoidance direction determined in step S400 by this time Ttc is calculated. Here, the time when the contact between the host vehicle SM and the obstacle SB is expected is Ttc.
For example, when the formula (34) is true, it is determined that left avoidance is more advantageous. Then, the lateral movement amount can be calculated from the Y coordinates of the right rear part of the obstacle SB and the left front part of the host vehicle SM at the current time.
This can be expressed as follows when expressed in different cases.

Next, in step S520, the initial state Xv (0) of the vehicle model is acquired. This processing may be performed in the same manner as described in step S220. Therefore, the description is omitted.
In step S240, when the trajectory prediction using the vehicle model is adopted, the initial state Xv (0) of the vehicle model is acquired. Therefore, the process itself of step S520 can be omitted.
Next, in step S530, a predetermined minimum command steering angle δref is set, and trajectory prediction is performed when the vehicle model is constantly input.
It can be expressed as follows when expressed in different cases.

A trajectory is calculated when δref and αlon (0) at the current time are constantly input to the model from t = 0 to t = Ttc, and after Ttc seconds, using equations (28) to (31) The coordinates of the four corners of the host vehicle SM are calculated.
Next, in step S540, it is determined whether the constant input δref is appropriate using the avoidance lateral movement amount calculated in step S510 and the host vehicle SM travel locus calculated in step S530.
It can be expressed as follows when expressed in different cases.

（３７）式が真であれば、一定入力δref（ｔ）を加えることで、Ｔｔｃ秒後に回避が可能と見なすことができる。この場合には、ステップＳ５７０へと移行する。
一方、（３７）式が偽であれば、一定入力δrefでは、接触危険があると判断し、Ｓ５５０へと移行する。
ステップＳ５５０では、指令舵角δrefの再設定を行う。
ＤRとＤLの大きさによって、以下のように場合分けされる。
δref ＝δref ＋Δδｗ （ＤR ＜ ＤLの場合 ）
δref ＝δref −Δδｗ （ＤR ≧ ＤLの場合 ）
・・・（３８）

If equation (37) is true, it can be considered that avoidance is possible after Ttc seconds by adding a constant input δref (t). In this case, the process proceeds to step S570.
On the other hand, if the expression (37) is false, it is determined that there is a contact risk with the constant input δref, and the process proceeds to S550.
In step S550, the command steering angle δref is reset.
Depending on the size of DR and DL, cases are classified as follows.
δref = δref + Δδw (when DR <DL)
δref = δref−Δδw (when DR ≧ DL)
... (38)

Similarly to step SS530, a locus is calculated when δref and αLON (0) at the current time are constantly input to the model from t = 0 to t = Ttc, and (28) to (31) are calculated. ) To calculate the coordinates of the four corners of the host vehicle SM after Ttc seconds.
In step S560, the magnitude of the command steering angle δref is evaluated.
That is, as described below, it is determined whether or not the absolute value of the command steering angle δref is smaller than a certain evaluation reference value δTHref.
| Δref | <δTHref (39)
Here, ΔTHref is a threshold value for determining that the steering amount is large, and is determined based on the current vehicle speed.

If the equation (39) is true, that is, if the absolute value of the command steering angle δref is smaller than the evaluation reference value δTHref, the process returns to the determination in step S540. In this case, it can be determined that there is still a margin in steering.
On the other hand, if the equation (39) is false, when the absolute value of the command steering angle δref is equal to or larger than the evaluation reference value δTHref, it is determined that the steering operation for avoidance needs to be large. In this case, the process shifts to an emergency brake system.
In this way, by repeatedly performing steps S540, S550, and S560, the steering amount for guiding in a direction that is advantageous for avoidance can be determined.
On the other hand, in step S570, evaluation based on the Y coordinate of the host vehicle SM at the present time is performed.

In the scene of FIG. 5, the case determination is performed as follows.
Yv> YL-.DELTA.Y (when DR <DL)
Yv ≦ Yc + ΔY (when DR ≧ DL)
(40)
Here, ΔY in equation (40) represents a margin in the horizontal position. If this ΔY is small, guidance is permitted to the vicinity of the boundary of the own vehicle travel path.
If the expression (40) is true, the process proceeds to step S580.
On the other hand, if the expression (40) is false, the steering command δref determined in step S540 is output to the steering actuator 2. Thereby, only steering control is performed.

When the host vehicle SM travel lane cannot be detected during the process of step S570, the following is performed. That is, the lateral position information at the time when this apparatus starts the guidance control is stored in the memory. And the restriction | limiting with respect to a horizontal movement is provided by determining beforehand the amount of horizontal movement which can move from the guidance control start time.
Next, in step S580, correction is performed by multiplying the steering command by a correction coefficient based on the current Y coordinate of the host vehicle SM.
δref = K × δref

In this embodiment, the correction coefficient K is decreased as the host vehicle position approaches the boundary of the host vehicle travel path, and the steering command is decreased as the vehicle position approaches the boundary of the host vehicle travel path.
For example, the correction coefficient is set as shown in FIG. That is, K is set to “1” when the distance from the boundary position of the host vehicle traveling path is greater than ΔY. When approaching the boundary position of the host vehicle traveling path to less than ΔY, K is decreased as the boundary of the host vehicle traveling path is approached.
Then, this steering command is transmitted to the steer actuator 2 and an alarm signal is transmitted to the alarm device, both steering control and alarm are performed, and the driver is informed of the danger ahead.

  Here, the acceleration sensor 4, the yaw rate sensor 5, the wheel speed sensors 7FR to 7RL, and the steering angle sensor 6 are means for detecting the state of the host vehicle, and the host vehicle motion state acquisition unit 9A constitutes the host vehicle state acquisition unit. To do. The stereo camera constitutes a forward obstacle detection means. The stereo camera also constitutes road information detection means. The avoidance determination unit 9C constitutes an obstacle determination unit. The avoidance operation estimation unit 9D and step S300 constitute an avoidance operation estimation unit. The own vehicle guidance amount calculation means 9E constitutes own vehicle guidance amount calculation means. The steering system (steer actuator 2) constitutes own vehicle guidance control means. The buzzer 8 constitutes alarm means.

(Operation)
The guidance operation for avoiding the obstacle SB in the vehicle driving support of the present embodiment will be described by taking the case shown in FIG. 5 as an example.
As shown in FIG. 5, for example, it is assumed that a front vehicle that is an obstacle SB is stopped on the shoulder.
If the current driving state of the host vehicle SM is maintained, if it is determined that there is a possibility of contacting the preceding vehicle after a predetermined time (for example, after 5 seconds), it is advantageous for avoiding the obstacle SB including the adjacent lane. The correct avoidance direction.
In the scene shown in FIG. 4, the advantageous avoidance direction is determined to be the right side.

Once the advantageous avoidance direction is determined, the vehicle steering amount of the wheel steering for guiding the host vehicle SM in the advantageous avoidance direction within the range of the allowable lateral movement set in the own vehicle travel path is calculated. The traveling direction of the host vehicle SM is changed by the calculated guidance amount.
The allowable lateral movement amount may be the width of the vehicle traveling path itself. It may be between the left and right boundaries of the own vehicle traveling path. Or the left-right width which left the margin width from the own vehicle SM by the predetermined width may be sufficient.

However, when steering by the driver for emergency avoidance is detected during the guidance, the process proceeds to emergency avoidance processing. In this case, since the vehicle state is easily guided in advance, the steering wheel operation for avoiding the obstacle SB by the driver can be reduced. In addition, since the vehicle is guided in an advantageous avoidance direction, the driver can easily determine the avoidance direction.
On the other hand, if it is determined that there is no obstacle SB that may come into contact without detecting the steering by the driver for emergency avoidance, the guidance process in the advantageous avoidance direction is stopped.
In addition, even when guiding in a direction that is advantageous for avoidance, if it cannot be avoided without departing from the vehicle traveling path, a warning is given to the driver to urge attention to take an avoidance action. .

(Effect of 1st Embodiment)
(1) Before the driver actually performs the avoidance operation, the host vehicle SM can be automatically guided in a direction advantageous for avoiding the obstacle SB. As a result, the risk of contact with the obstacle SB increases, and when the driver actually performs the avoidance operation, the vehicle state is guided in advance so that the burden of the avoidance operation by the driver is reduced. Reduction can be expected.

(2) In principle, the host vehicle SM is guided within an allowable range set in the host vehicle travel path. As a result, the host vehicle SM can be guided while minimizing secondary contact damage that may occur during lateral movement.
At this time, by acquiring the road boundary information, an allowable range can be set as the own vehicle travel path.
Here, when the road boundary information is not acquired, an amount obtained by adding the left and right movement widths to the lateral width of the host vehicle SM may be used as the “permissible range set on the host vehicle travel path”. In this case, when the host vehicle SM is traveling near one of the white lines, there is a possibility that the allowable range set for the host vehicle travel path may slightly protrude from the actual host vehicle travel path. There is.

(3) Further, based on the motion prediction of the host vehicle SM and the motion prediction of the obstacle SB, an advantageous avoidance direction and host vehicle guidance amount are calculated. Since the prediction of the future movement of the host vehicle SM and the front obstacle SB is made, the guidance amount of the host vehicle is determined in consideration of the time after the current time, and the future contact damage is more accurately determined. Can be prevented.
(4) The vehicle guidance amount is realized by the steering actuator 2, that is, the steering mechanism. As a result, any vehicle that can control a steering mechanism that has become common now can be applied without using a new actuator.
(5) In the case where it is necessary to deviate from the vehicle traveling path, a warning is given to the driver in addition to the guidance control. As a result of the future prediction, it is possible to notify the driver in advance of the future danger that will be avoided by departing from the traveling lane of the host vehicle SM.

(Modification)
(1) The example using the general buzzer 8 as an alarm means was shown. Instead of this, an alarm means for transmitting an alarm signal through the five senses of the driver, such as visual and physical sensations, such as image display and steering vibration, may be employed. The effect is similar.
It may be a thing.
(2) Moreover, in the said embodiment, the steering mechanism is illustrated as the own vehicle SM guidance control means which implement | achieves the own vehicle guidance amount. As in the second embodiment, a braking / driving mechanism may be employed as the vehicle SM guidance control means.

(3) Moreover, the road boundary information ahead of the own vehicle SM is acquired based on the image acquired from the stereo camera. Road boundary information ahead of the host vehicle SM may be acquired by scanning with an electromagnetic wave such as a laser beam.
(4) Based on the acquired road boundary information, the range of lateral movement allowing the own vehicle traveling path itself is set. Instead of this, an allowable lateral movement range may be obtained by adding a preset lateral movement allowance to the lateral width of the host vehicle SM. In this case, it is not always necessary to acquire road boundary information.

(Second Embodiment)
Next, a second embodiment according to the present invention will be described with reference to the drawings. The same devices as those in the first embodiment will be described with the same reference numerals.
In the first embodiment, the example in which the host vehicle SM is guided in an advantageous avoidance direction by controlling the steering of the wheels has been described. In the second embodiment, an example in which the host vehicle SM is guided in an advantageous avoidance direction by the braking / driving force will be described.
FIG. 11 is a schematic diagram of an apparatus configuration of a vehicle equipped with the vehicle driving support apparatus of the second embodiment.

(Constitution)
As shown in FIG. 11, the vehicle of this embodiment is a front-wheel drive and front-wheel steering vehicle having an internal combustion engine. It also has a brake system that can control the braking amount of each wheel independently. That is, each wheel is equipped with a brake device 21. The brake device 21 applies a braking force to the corresponding wheel in response to a command from the avoidance guidance controller 9. The brake device 21 is for decelerating the traveling vehicle. Moreover, the braking amount of each wheel can be adjusted independently.

However, the vehicle configuration of the present embodiment is not limited to this configuration.
The basic configuration of this embodiment is the same as that of the first embodiment. That is, the stereo camera 3, the acceleration sensor 4, the yaw rate sensor 5, the wheel speed sensors 7FR to 7RL, the steer actuator 2, and the buzzer 8 are equipped in the same manner as in the first embodiment.
Further, the own vehicle guidance amount calculation means 9E and the own vehicle guidance control command output means 9F in the avoidance guidance controller 9 of the present embodiment calculate the own vehicle guidance amount for guidance by braking control instead of steering control. The command is output.

Next, processing of a portion related to vehicle driving assistance in processing of the avoidance guidance controller 9 of the present embodiment will be described with reference to FIG.
Here, in order to make the contents of the processing easy to understand, a scene shown in FIG. 13 is assumed as an example. The scene shown in FIG. 13 is a scene in which the host vehicle SM is traveling on a straight road on one lane on one side, and a vehicle traveling in front of the host vehicle SM traveling lane stops on the left side.
The processing of the avoidance guidance controller 9 operates at a predetermined sampling period.

First, in step S1100, the camera image is taken in from the stereo camera 3 and the detection signal is read by each sensor. These processing results are stored in a memory in the microprocessor.
Then, as in the first embodiment, the position information (Xv, Yv) of the own vehicle SM center point, the position information (Xb, Yb) of the rear center of the obstacle SB, and the Y coordinate of the left and right road boundaries on the unified coordinates. Associate YL, YR, and the center line Y coordinate YC. Then, the moving speed and acceleration / deceleration of the host vehicle SM and the obstacle SB in the determined coordinate system are calculated.

In FIG. 13, since the host vehicle SM is traveling in the X-axis direction, the host vehicle speed is (Vv _x , Vv _y ) = (Vv, 0) based on a signal from the wheel speed sensor. The host vehicle SM acceleration / deceleration is (Gv _x , Gv _y ) = (αlon, αlat) based on a signal from the acceleration sensor 4.
Also, the obstacle SB is moving leftward while decelerating. Therefore, the obstacle SB rate is can be determined from the difference between the previous acquired obstacle position information, the title and (Vb _x, Vb _y). Further, the acceleration of the obstacle SB is can be determined from the difference, referred to as (Gb _x, Gb _y).

  Next, in step S1200, based on the coordinate information set in step S1100, it is determined whether or not the host vehicle SM may contact the obstacle SB from the current time to a predetermined time in the future. If it is determined that the host vehicle SM may come into contact with the obstacle SB, the process proceeds to step S1300. On the other hand, when it is determined that there is no possibility that the host vehicle SM will contact the obstacle SB, the vehicle returns.

  In step S1300, it is determined whether the driver is performing a steering operation for avoidance. Here, the state transitioning to step S1300 is a state in which the host vehicle SM may come into contact with the obstacle SB. Therefore, it is possible to estimate that the driver's steering operation exceeding a predetermined value is a steering operation for avoidance. For the determination, for example, the processing described in the first embodiment may be used.

  And when it determines with the driver not trying to avoid the front obstacle SB, it progresses to step S1400. On the other hand, when it is determined that the driver intends to avoid the front obstacle SB and the driver determines the avoidance direction, other avoidance control (obstacle SB avoidance control means 9G at the time of steering detection) Migrate to The other avoidance control means 9G is, for example, control as in the above-described prior art or avoidance control in which automatic braking is applied according to the estimated contact time with the obstacle SB.

Next, in step S1400, an avoidance direction advantageous for avoiding the obstacle SB is determined.
That is, considering whether to avoid the obstacle SB in the right direction or in the left direction, it is determined whether the avoidance in the direction is larger in avoidance. The determination process described in the first embodiment may be employed.
Next, in step S1500, the braking induction amount of each wheel for guiding the host vehicle SM in the advantageous avoidance direction determined in S1400 is determined.
Next, in step S1600, a command value corresponding to the braking induction amount of each wheel is output to the brake device 21.

Next, an example of the presence / absence determination process (avoidance determination unit 9C) of the possibility of the host vehicle SM coming into contact with the obstacle SB corresponding to the process of step S1200 will be described with reference to FIG.
In the present embodiment, it is determined whether there is a risk of contact between the current time (t = 0) and Tcj seconds ahead.
First, in step S1210, the host vehicle SM area and the obstacle SB area at the current time are expressed on the unified coordinates set in step S1100. The processing in step S1210 is the same as that in the first embodiment.

  Next, in step S1220, it is determined whether or not there is an area where the host vehicle SM area Sv (t) and the obstacle SB area Sb (t) overlap at time t seconds. If it is determined that there is an overlapping area, the process proceeds to step S1300. On the other hand, if there is no overlapping area, it is determined that the risk of contact is low by t seconds and the process proceeds to step S1230. The processing in step S1220 is the same as that in the first embodiment.

In step S1230, time t is advanced by Tcj seconds. Subsequently, it is determined whether or not the value of time t is smaller than a predetermined time Tcj. If the value of time t is less than Tcj, the process proceeds to step S1240 in order to continue predicting future danger. On the other hand, if the value of time t is equal to or greater than Tcj, it is determined that the risk of contact with the obstacle SB is low between the current time and the predetermined time Tcj, and the process returns. The processing in step S1230 is the same as that in the first embodiment.
The basic processing from step S1210 to step S1230 is the same as in the first embodiment. Therefore, detailed description is omitted.

Next, in step S1240, the own vehicle SM area and the obstacle SB area at time t seconds are calculated from the speed information of the own vehicle SM and the obstacle SB at the current time. Then, control goes to a step S1220.
As the processing in step S1240, the same method as the processing in the first embodiment may be employed.
Here, a case where a model expressed by another method is used will be described.
Examples of vehicle models are shown in equations (41) to (46). In this model example, it is assumed that no significant skidding has occurred in the vehicle.

Here, the model (the model exemplified in the first embodiment) that can be expressed by the above-described equations (12) to (22) is a simplified physical motion of the vehicle.
On the other hand, in the example of the vehicle model exemplified in the second embodiment, as shown in FIG. 15, the vehicle is regarded as a mass point, and the degrees of freedom in the vertical (X-axis direction), horizontal (Y-axis direction), and yaw directions. It has something.

Further, x _v , y _v , θ, and γ in the above formula are the same as those in the first embodiment. That is, x _v , y _v , θ, and γ are the vehicle coordinate, the Y coordinate, the yaw angle, and the yaw rate in the unified coordinate system, respectively.
Further, νx is the speed in the vertical direction of the host vehicle SM. νY represents the speed in the lateral direction of the host vehicle SM.
αlon and αlat are vehicle longitudinal acceleration and lateral acceleration, respectively.
Mbr is a moment in the yaw direction generated by braking of each wheel.
When the vehicle model represented by the above equations (41) to (46) is used, the motion state of the vehicle used in this embodiment is a six-dimensional vector. An input for operating the vehicle is a three-dimensional vector. And these can be expressed as follows.

Next, a method for acquiring the initial state vector of the vehicle model shown in the equation (47) will be described.
x _v and y _v becomes Xv and Yv calculated in stage S1100. γ can be obtained from the output of the yaw rate sensor 5. θ can be detected by integrating signals from the yaw rate sensor 5 with reference to the vehicle posture at a certain point in time. Alternatively, θ is determined from the angle formed by the host vehicle SM and the host vehicle travel path from the image information. νx is obtained from the integration of the longitudinal acceleration signal or the wheel speed. νY is calculated from the integral of the lateral acceleration signal.

When u ′ _v at the current time shown in the equation (48) is continuously given to the vehicle model in the initial state, the locus of the vehicle center of gravity of the host vehicle SM can be obtained.
Using the barycentric coordinates (x _v (t), y _v (t)) at this future time t and the vehicle attitude θ (t), the equations (27) to (30) shown in the first embodiment are used. Can be expressed as
Here, in the scene shown in FIG. 13, the obstacle SB decelerates, but in S1100, acceleration / deceleration information of the obstacle SB is acquired. For this reason, the trajectories of the four corners of the obstacle SB can be expressed as follows.

And the own vehicle SM area | region Sv (t) is computed from the locus | trajectory prediction of the own vehicle SM calculated using (41)-(46) Formula. And obstacle SB area | region Sb (t) is calculated from the locus | trajectory prediction of the obstacle SB four corners calculated using (50), (51) Formula.
Thereafter, the process proceeds to S1220, and contact determination is performed based on the overlap of the regions as in the above-described equation (5).

Next, an example of processing for determining the braking guidance amount of each wheel for guiding to the advantageous avoidance direction determined in S1400, which is the processing in step S1500 (own vehicle control amount calculation means 9Eb), will be described. FIG. 16 shows the detailed processing flow.
First, in step S1510, how much it is necessary to move laterally in the avoidance direction determined in S1400 is calculated by Ttc. Ttc is a time at which contact between the host vehicle SM and the obstacle SB is expected.

This process may be divided into cases in the avoidance direction determined in step S1300, as in equation (35) of S510 in the first embodiment.
In step S1520, a predetermined minimum braking moment Mbr is set, and a trajectory prediction when a constant input is made to the vehicle model is performed.
It can be expressed as follows when expressed in different cases.
Mref = Mbr (0) +. DELTA.Mbr (when DR <DL)
Mref = Mbr (0)-. DELTA.Mbr (when DR.gtoreq.DL)
(52)
Further, since Mref is a braking force, the vehicle longitudinal force can be expressed as follows.

With respect to Mref, α′lon, and αlat (0) at the current time, a trajectory when a constant input is made to the model from t = 0 to the predicted contact time t = Ttc is calculated. Then, the coordinates of the four corners of the host vehicle SM after Ttc seconds are calculated using equations (27) to (30).
In step S1530, it is determined whether or not the constant input Mref is appropriate using the avoidance lateral movement amount calculated in step S1510 and the host vehicle SM travel locus calculated in step S1520.

In this process, the determination is made for each case in the avoidance direction, as in equation (37) of S540 in the first embodiment.
If it is determined that it is difficult to avoid using Mref, the process proceeds to step S1540. On the other hand, if it is determined that it can be avoided, the process proceeds to step S1560.
In step S1540, the command braking moment is reset.
Cases are divided as follows according to the sizes of DR and DL.
Mref = Mref + ΔMbr (when DR <DL)
Mref = Mref-.DELTA.Mbr (when DR.gtoreq.DL)

Similarly to S1520, when this Mbr, the vertical acceleration α′lon when the Mbr is input, and αlat (0) at the current time are input to the model from t = 0 to t = Ttc. Is calculated. Then, the coordinates of the four corners of the host vehicle SM after Ttc seconds are calculated using equations (27) to (30).
In step S1550, the magnitude of the command moment Mref is evaluated.
The magnitude relationship between a certain evaluation reference value MTHref and Mref is determined as follows.
| Mref | <MTHref (54)
Here, MTHref is a threshold value for determining that the induced moment is large. This threshold is determined based on the current host vehicle speed.
MTHref (Vv), which is a function of the vehicle speed, is generally set smaller as the vehicle speed increases to prevent spin. However, it shall have the following restrictions.

This restriction represents that each wheel cannot be braked more than the wheel load.
Here, by having means for detecting or estimating the road surface friction coefficient of the traveling road surface of the host vehicle SM, it is possible to make a restriction according to the road surface condition.
If the formula (54) is true, it is determined that there is still a margin in the braking moment, and the process proceeds to step S1530 again. On the other hand, if the formula (54) is false, it is determined that the moment command is excessive, and the process proceeds to a known emergency brake system.

By repeating steps S1530 to S1550, the braking moment amount for guiding in a direction that is advantageous for avoidance is determined.
On the other hand, in step S1560, evaluation based on the Y coordinate of the host vehicle SM at the present time is performed.
Even in the scene of FIG. 13, it is possible to make a case-by-case determination like the equation (40) shown in the first embodiment.
If it is determined in step S1560 that there is little risk of departure from the vehicle traveling path, the process proceeds to step S1570. On the other hand, if it is determined that there is a risk of departure, the process proceeds to step S1580.

Here, as in the first embodiment, when the host vehicle SM travel lane cannot be detected, the amount of lateral movement from the control start time is constrained based on the processing history of the microprocessor. Performs processing that replaces the determination of.
In step S1570, each wheel command braking force is determined so as to be the braking moment Mref calculated in step S1530.
That is, if the braking force of each wheel is Fx ₁ (front left), Fx ₂ (front right), Fx ₃ (back left), and Fx ₄ (back right), it can be divided into the following cases.

The braking force command value distributed in this way is output to the braking system (braking device 21), and braking control is performed independently for each wheel.
In step S1580, the moment command is corrected based on the current Y coordinate of the host vehicle SM.
The correction method is performed by multiplying the correction coefficient as in the following equation, for example.
Mref = Mref × K
As shown in FIG. 10 of the first embodiment, the correction coefficient K is a correction coefficient K determined to decrease as it approaches the travel path boundary.
Using this corrected Mref, the distribution as shown in equation (56) is performed, and the braking force command value for each wheel is calculated. Each wheel independent braking control is performed based on this command value. In addition, by transmitting an alarm signal to the alarm device, the driver is informed of the danger ahead.
Other configurations are the same as those in the first embodiment.
(Operation / Action)
The basic operation is the same as that of the first embodiment.
However, in this embodiment, the braking moment is generated by adjusting the braking amount of each wheel. This braking moment guides the host vehicle SM in an advantageous avoidance direction.

(Effect of 2nd Embodiment)
(1) Basic effects are the same as those in the first embodiment. That is, it is possible to automatically guide the host vehicle SM in a direction advantageous for avoiding the obstacle SB before the driver actually performs the avoidance operation. As a result, the risk of contact with the obstacle SB increases, and when the driver actually performs the avoidance operation, the vehicle state is guided in advance so that the burden of the avoidance operation by the driver is reduced. Reduction can be expected.
(2) The host vehicle SM is guided in an advantageous avoidance direction by braking. That is, the steering is not controlled. For this reason, the influence on the steering for avoiding the obstacle SB by the driver is reduced.

(Modification)
(1) In the present embodiment, the host vehicle SM is guided in an advantageous avoidance direction by the braking moment. In accordance with this, guidance by steering may be used together. In this case, the degree of freedom in controlling the guidance amount is improved.
(2) In the present embodiment, a yaw moment for guiding the host vehicle is generated by braking each wheel. A configuration may be adopted in which each wheel can be individually driven and a yaw moment for guiding the host vehicle is generated by controlling the driving of each wheel.

(Third embodiment)
Next, a third embodiment will be described with reference to the drawings.
FIG. 17 is a schematic diagram of a device configuration of a vehicle equipped with the vehicle driving support device of the present embodiment.
The vehicle of this embodiment is a front-wheel drive and front-wheel steering vehicle having an internal combustion engine. A steering system capable of controlling the steering amount of the front wheels is mounted. However, the vehicle configuration of the present embodiment is not limited to this.
Since the basic configuration of the vehicle is the same as the configuration in the first embodiment, detailed description thereof is omitted.

Next, processing of a portion related to vehicle driving assistance in the processing of the avoidance guidance controller 9 in the present embodiment will be described with reference to FIG.
Here, in order to make the contents of the processing easy to understand, a scene shown in FIG. 19 is assumed. The scene shown in FIG. 19 is a scene in which the host vehicle SM is traveling on a straight road on one lane on one side, and a pedestrian crosses from the right side to the left side in front of the host vehicle SM traveling lane. Even if the pedestrian is in the oncoming lane, it can be an obstacle SB.

The processing of the avoidance guidance controller 9 operates at a predetermined sampling period.
First, in step S2100, a camera image is captured from the stereo camera 3 and a detection signal is read by each sensor. These processing results are stored in a memory in the microprocessor.
Then, as in each of the above-described embodiments, the position information (Xv, Yv) of the own vehicle SM center point, the position information (Xb, Yb) of the rear center of the obstacle SB, and the Y of the left and right road boundaries on the unified coordinates. Coordinates YL and YR are associated with the center line Y coordinate YC. Then, the moving speed of the host vehicle SM and the obstacle SB in the determined coordinate system is calculated.

In the scene as shown in FIG. 19, the host vehicle SM is traveling in the X-axis direction. Therefore, the vehicle speed is (Vv _x , Vv _y ) = (Vv, 0) from the wheel speed sensor.
Also, the obstacle SB moves at a constant speed in the left direction. Therefore, the obstacle SB rate, the difference between the previously acquired obstacle position information, can be expressed as _{(Vb x, Vb y) =} (Vb, 0).
Next, in step S2200, based on the coordinate information set in step S2100, it is determined whether or not there is a possibility that the host vehicle SM is in contact with the obstacle SB from the present to a predetermined time in the future.

The process of step S2200 may be the same process as step S200 of the first embodiment. That is, the coordinates of the four corners are obtained from the trajectory prediction based on the vehicle model and the motion prediction of the obstacle SB. And it judges based on the area | region in these four corner coordinates.
In step S2300, considering the case where the obstacle SB is avoided in the right direction and the case where the obstacle SB is avoided in the left direction, it is determined in which direction the avoidance is greater when avoiding. Here, processing by a method different from the first and second embodiments will be described later.
In step S2400, the steering sliding amount for making the advantageous avoidance direction determined in step S2300 is determined, and in step S2500, the command is output.

Next, a description will be given of a method for determining in which direction in step S2300 the avoidance is larger when avoiding.
In the third embodiment, a case where a determination is made based on optimization calculation will be exemplified with reference to FIG.
First, in step S2310, a vehicle model, an obstacle SB model, and an evaluation function are set as settings necessary for the optimization calculation.
As the vehicle model, the vehicle model used in step S2200 is used.
That is, it can be expressed as follows.

In the present embodiment, Arufalon of u _v shown in (58) equation is assumed to be constant for the current longitudinal acceleration.
The details of the model can be expressed by equations (12) to (22).
Next, the obstacle SB model can be described by a differential equation as shown below. Here, it is considered that constant velocity linear motion is performed based on the current moving speed information.

  Here, if the movement of the obstacle SB can be detected in more detail, the obstacle SB model expressed by the equations (60) and (61) can be complicated. However, in order to suppress the calculation load on the microprocessor, the simplest model is used in this embodiment.

  Next, an evaluation function for evaluating the avoidance path is set. The evaluation function can be expressed as the following expression based on the predicted value of the vehicle state vector x for the input u applied to the vehicle from the current time t0 to the predetermined estimated time t0 + Tf.

The first term on the right side of the equation (64) is an equation (terminating evaluation equation) for evaluating the vehicle motion state at time t0 + Tf. The second term is an expression (section evaluation expression) for evaluating the vehicle motion state in the section from time t0 to t0 + Tf.
Here, the longer Tf is, the longer the future is predicted, but generally the calculation load increases.
Then, the end evaluation item and the section evaluation item are set as follows.
Terminal evaluation item 1: After Tf seconds, the vehicle posture is directed toward the road. Section evaluation item 2: Keep away from the front obstacle SB for Tf seconds.
Section evaluation item 3: Prevents the own vehicle SM from deviating from the road boundary for Tf seconds. Section evaluation item 4: Prevents the own vehicle SM from deviating from the traveling road of the section. Section evaluation item 5: Tf seconds. The steering control input of the host vehicle SM is made as small as possible. The evaluation formulas corresponding to these evaluation items are as follows.
・ Terminal evaluation item 1

・ Section evaluation item 2 (σx and σY are parameters indicating the shape of the function)

・ Section evaluation item 3 (Δ is the margin parameter for approaching the road boundary)

・ Section evaluation item 4

・ Section evaluation item 5

When each of the above evaluation items is weighted using the parameters w ₁ , w ₂ , w ₃ , w ₄ , and w ₅ , the first and second terms in the equation (54) can be expressed as follows. it can.

Here, the weighting parameters are set so that the balance between the evaluation items is improved. However, since the evaluation item 2 is the most important item, it is usually set larger.
Moreover, in the evaluation items 3 and 4, it is better that the evaluation item 3 is clearly stronger. Therefore, w3> w4.
Thus, the evaluation function setting is completed, and the process of step S2310 is completed.

Using the initial state of the host vehicle SM model, the initial state of the obstacle SB model, and the evaluation function determined in step S2310, optimization calculation is performed in the subsequent processing. In other words, a numerically optimal avoidance route is calculated from the viewpoint of the evaluation function set this time.
In step S2320, a time series avoidance operation amount map that is stored in advance in a memory in the microprocessor and is created based on various scenes is closest to the current avoidance scene and avoids in the right direction. Select. The selected one is used as the initial solution when performing the optimization calculation.

A method for creating a time-series avoidance manipulated variable map assumed in the present embodiment will be described.
Considering the scene as shown in FIG. 21 using the vehicle model described in step S2310, as shown in FIG. 22, taking the own vehicle speed and the relative distance between the own vehicle SM and the obstacle SB as the evaluation axes, A simulation when the vehicle avoids the obstacle SB in the right direction under each condition is performed offline, and the time-series steering amount at that time is acquired.

This is simulated for all the set conditions to create an initial solution map index, which is stored in the microprocessor memory together with each time-series manipulated variable.
Similarly, when avoiding to the left, the result of the offline simulation is stored in the memory.
Next, in step S2320, when this map is referenced from the memory, the closest condition is selected based on the current vehicle speed and the distance to the obstacle SB, and stored in the memory address indicated by the selected index. The series operation amount is read (see FIG. 22).

This is the initial solution for the optimization calculation. The time-series avoidance manipulated variable map in this embodiment uses the vehicle speed and the relative longitudinal distance as the evaluation axes. In addition, by adding the relative vertical distance, the moving speed of the obstacle SB, and the like to the evaluation axis, it is possible to set the time series operation amount corresponding to the actual scene as the initial solution. However, since the memory capacity in the microprocessor is limited, it is necessary to create a map in consideration of the trade-off between memory capacity and versatility.
The above is the processing content in step S2320.

Next, in step S2330, optimization calculation is performed based on the vehicle model, obstacle SB model, evaluation function, and initial solution set in steps S2310 and S2320.
The problem of obtaining an operation amount that minimizes the evaluation function that can be expressed by equation (64) is generally called an optimal control problem. Various algorithms are considered as known techniques for obtaining the numerical solution.
Here, as an example of a known technique, reference is made to T. Ohtsuka, “A continuation / GMRES method for fast computing of nonlinear receeding horizon control”, Automatica, vol, 40, 563/574, 2004.

The optimal operation amount is calculated using such an algorithm. In the case of this embodiment, (58) from equation, the input u _v is Arufalon is put constant. Therefore, δs becomes an object of optimization, and each operation amount from time t0 to t0 + Tf can be calculated in time series.
In actual operation amount calculation, the evaluation interval is divided by an appropriate number of steps N and discretized. Then, the value of the operation amount at each step time is calculated. That is, N time-series inputs as shown below can be obtained.

Here, the larger the N, the shorter the sample time interval, and the more accurate predicted operation amount can be calculated. However, since the number of data to be calculated increases correspondingly, the calculation load of the microprocessor increases. Therefore, it is desirable to determine N so that the sample time interval is 10 to 100 msec.
The above is the processing content in step S2330.
Next, in step S2340, an avoidance margin for right avoidance is calculated. Here, a method based on the trajectory of the vehicle model when the optimum time series operation is performed is used.
First, the vehicle trajectory (x _v (τ), y _v (τ)) when the optimal time-series operation amount calculated in step S2330 is input to the vehicle model set in step S2310 and the movement trajectory of the obstacle SB ( xb (τ), yb (τ)). That is, the time series distance DVB_R (t) from the center of gravity of the host vehicle SM to the obstacle SB is calculated as follows. The DVB_R (t) having the smallest distance is defined as DVEmin_R.

Also, the time series distance DVR_R (t) is calculated for the vehicle trajectory (x _v (τ), y _v (τ)) and the right road boundary as follows. The DVR_R (t) having the smallest distance is defined as DVRmin_R.
Also, the time series distance DVR_R (t) is calculated for the vehicle trajectory (x _v (τ), y _v (τ)) and the right road boundary as follows. The DVR_R (t) having the smallest distance is defined as DVRmin_R.

Using this DVBmin_R and DVRmin_R, a margin DR for avoiding the right side described below is calculated.
_{DR = w VB × DVBmin_R + w} VR × DVRmin_R ··· (77)
In equation (77), w _Vb represents a weighting factor for the degree of proximity to the obstacle SB. _wVR represents a weighting factor for the degree of approach to the road boundary. In order to give priority to avoiding the obstacle SB, the weight is normally set so that w _Vb > w _VR .
The above is the processing content in step S2340.
Here, the processing from steps S2320 to S2340 is processing for calculating a margin when avoiding to the right side. The subsequent steps S2350 to S2370 are processes for calculating the margin when avoiding to the left.

Since the basic processing contents in steps S2350 to S2370 are the same as the processing contents in steps S2320 to S2340, the details are omitted.
In step S2350, from the time-series avoidance operation amount map created based on various scenes stored in advance in the memory in the microprocessor, the closest to the current avoidance scene and the avoidance to the left is performed. And use this as the initial solution when performing optimization calculations.
In step S2360, optimization calculation is performed based on the vehicle model, obstacle SB model, evaluation function, and initial solution set in steps S2310 and S2350. Again, N time-series inputs can be obtained as shown below.

  In step S2370, the time series distance DVB_L (t) from the center of gravity of the host vehicle SM to the obstacle SB is calculated using the optimization calculation result of step S2360 and the vehicle model. The DVB_L (t) having the smallest distance is defined as DVEmin_L.

Also, the time series distance DVR_L (t) is calculated for the vehicle trajectory (x _v (τ), y _v (τ)) and the left road boundary as follows. The DVR_L (t) having the smallest distance is defined as DVRmin_L.

Using this DVBmin_L and DVRmin_L, a margin DL for avoiding the right side as described below is calculated.
_{DL = w VB × DVBmin_R + w} VR × DVRmin_R ··· (83)
The above is the processing content in steps S2350 to S2370.
In step S2380, it is determined whether there is an avoidance margin on either the left or right side.
That is, the same determination as that of the expression (33) in the first embodiment is performed for DR and DL calculated in steps S2340 and S2370.
If both equations (33) are true, it is determined that there is no avoidance margin in the left-right direction. In this case, the function shifts from the system of the present embodiment to an emergency brake system as found in the known art.

On the other hand, if any of the equations (33) is false, it is determined that there is an avoidance margin in the left-right direction. In this case, the process proceeds to step S2390.
In step S2390, the left and right avoidance allowances are compared to determine which direction has a higher avoidance allowance and is advantageous for avoidance. The determination formula is the same as the formula (34).
If the equation (34) is true, the leftward avoidance has a larger margin, and if it is false, the avoidance has a larger margin. Thereby, the avoidance direction is determined.

Next, processing for determining the steering guidance amount in order to make the avoidance in the direction determined in step S2300 advantageous, which is processing in step S2400, will be described.
First, in step S2410, the time when the contact between the host vehicle SM and the obstacle SB is expected is Ttc. Then, by this Ttc, the amount of lateral movement in the direction determined in step S2300 is calculated.
In this process, cases are classified according to the direction determined in step S2300, as in equation (35) of step S510 in the first embodiment.

Next, in step S2420, the optimized steering pattern expressed by equation (72) or (78) is input to the vehicle model expressed by equation (59).
Αlon at this time is a constant input based on the current vertical acceleration / deceleration.
In step S2430, it is determined whether or not there is a risk of contact with the moving forward obstacle SB after Ttc seconds when the operation is performed with the steering pattern optimized in step S2300.

In this process, as in the expression (37) of S540 in the first embodiment, the determination is made according to the avoidance direction.
If it is determined by this determination that it is difficult to avoid the steering pattern optimized in step S2300, the process proceeds to a known emergency brake system. On the other hand, if it is determined that avoidance is possible, the process proceeds to step S2440.
In step S2440, it is determined whether or not steering for avoidance is performed based on the driver's steering information.

As an example, the following processing is performed. That is, the difference Δδs between the current steering angle signal acquired from the steering angle sensor 6 and the previously acquired steering angle signal is calculated. Then, this value is compared with the threshold value ΔδTH as follows.
| Δδs |> ΔδTH (84)
If the formula (84) is true, the driver intends to avoid the forward obstacle SB, and the process proceeds to step S2450. On the other hand, if the formula (84) is false, the driver determines that he is not trying to avoid the forward obstacle SB, and proceeds to step S2460.
In step S2450, a steering command torque that assists the driver's avoidance operation is calculated based on the optimized steering amount in step S2300.
That is, Δδs is calculated as follows based on the avoidance direction advantageous for avoidance determined in step S2300.

Here, δs (0) indicates the steering amount by the driver. Then, the difference from the actually required steering amount is obtained as Δδs by the equation (85).
From this Δδs, a torque command is calculated using a map of Δδs and the torque command value δtq as shown in FIG.
Here, the reason why the dead zone δdz is used in FIG. 24 is to give priority to the driver operation amount.
In step S2460, it is determined whether or not the vehicle deviates from the SM traveling lane.
Here, a method different from the determination based only on the current information as in the first and second embodiments will be described. That is, using the time-series steering pattern calculated in step S2300, a determination is made after predicting the future.
First, the time series Y coordinate when the time series steering pattern is input to the vehicle model is expressed as follows.

Using this, the following determination is performed for all elements of y ^* _v .

If even one of the N determinations expressed by the equation (87) is false, it is determined that the vehicle can travel on the SM traveling lane for Tf seconds, and the process proceeds to step S2470.
Here, as in the first and second embodiments, if the host vehicle SM travel lane cannot be detected, it can be replaced by limiting the amount of lateral movement from the control start point based on the processing history of the microprocessor. Processing can be performed.

On the other hand, if all the N determinations expressed by the equation (87) are true, it is determined that the vehicle can be avoided while remaining in the own vehicle SM travel lane. Then, a steering torque command value is calculated based on the command steering angle δ ^* _SR (t0) obtained by the optimization. The calculated steering torque command value is transmitted as a signal to the steering system.
In step S2470, the steering command is corrected based on the current Y coordinate of the host vehicle SM.
The correction method is the same as the method of S580 in the first embodiment.
Then, the corrected steering command is transmitted to the steering actuator 2. In addition, an alarm signal is transmitted to the alarm device. As a result, both steering control and warning are performed to inform the driver of the danger ahead.

(Operation)
The basic operation is the same as that in the first embodiment.
However, when a steering operation for avoidance is detected, a steering command torque that assists the driver's avoidance operation is calculated and output in step S2450 without shifting to another avoidance control. In other words, the range of allowable lateral movement is not limited to the own vehicle traveling path, but the range of allowable lateral movement is substantially expanded to the outside of the own vehicle traveling path to assist the driver's avoidance operation. To do.

(Effect of this embodiment)
(1) The effect of this embodiment is the same as that of the first embodiment.
(2) The permissible guidance range is expanded according to the driver's steering amount. This allows the guidance system to switch from automatic guidance control to assist control without the hindrance to the driver's operation when the driver tries to deviate from the vehicle traveling path in order to achieve the avoidance purpose. .
(Modification)
(1) Here, in the present embodiment, a vehicle capable of steering control has been described. Instead of this, even in a vehicle capable of independent wheel braking control as in the second embodiment, it is possible to realize a guidance system based on optimization calculation by setting a vehicle model.

BRIEF DESCRIPTION OF THE DRAWINGS It is a schematic diagram of the apparatus structure which concerns on 1st Embodiment based on this invention. It is a figure which shows the structure of an avoidance guidance controller. It is a figure which shows the structure of the own vehicle guidance amount calculation means. It is a figure which shows the process of the avoidance guidance controller which concerns on 1st Embodiment based on this invention. It is a figure which shows an example of a scene. It is a figure which shows the determination process of the contact risk degree which concerns on 1st Embodiment based on this invention. It is a figure explaining the feature point showing the own vehicle area | region and obstacle area | region which concern on 1st Embodiment based on this invention. It is a figure for demonstrating the state vector of the vehicle model which concerns on 1st Embodiment based on this invention. It is a figure for demonstrating the process for determining the guidance steering amount which concerns on 1st Embodiment based on this invention. It is a figure explaining the map of the guidance steering amount correction | amendment which concerns on 1st Embodiment based on this invention. It is a schematic diagram of the apparatus structure concerning 2nd Embodiment based on this invention. It is a figure which shows the process of the avoidance guidance controller which concerns on 2nd Embodiment based on this invention. It is a figure explaining an example of a scene. It is a figure which shows the determination process of the contact risk degree which concerns on 2nd Embodiment based on this invention. It is a figure for demonstrating the state vector of the vehicle model which concerns on 2nd Embodiment based on this invention. It is a figure explaining the determination process of the induced braking moment amount which concerns on 2nd Embodiment based on this invention. It is a schematic diagram of the apparatus structure which concerns on 3rd Embodiment based on this invention. It is a figure which shows the process of the avoidance guidance controller which concerns on 3rd Embodiment based on this invention. It is a figure explaining an example of a scene. It is a figure explaining the process for determining the direction advantageous for the avoidance which concerns on 3rd Embodiment based on this invention. It is an assumption figure at the time of calculating | requiring the initial solution map structure for the optimization calculation which concerns on 3rd Embodiment based on this invention. It is a figure explaining the initial solution map structure for the optimization calculation which concerns on 3rd Embodiment based on this invention. It is a figure explaining the determination process of the guidance steering amount which concerns on 3rd Embodiment based on this invention. It is a figure for demonstrating the map for the assist steering torque calculation which concerns on 3rd Embodiment based on this invention.

Explanation of symbols

2 Steer actuator (own vehicle guidance control means)
3 Stereo camera 4 Acceleration sensor 5 Yaw rate sensor 6 Steering angle sensor 7FR to 7RL Wheel speed sensor 8 Buzzer (alarm means)
9 avoidance guidance controller 9A own vehicle motion state acquisition means 9B vehicle forward information acquisition means 9C avoidance determination means 9D avoidance operation estimation means 9E own vehicle guidance amount calculation means 9Ea avoidance direction determination means 9Eb own vehicle control amount calculation means 9Ec alarm signal output means 9F Own vehicle guidance control command output means 9G Obstacle avoidance control means 9H Emergency stop control means 21 Brake device (own vehicle guidance control means)
SB obstacle SM own vehicle

Claims (7)

Own vehicle state acquisition means for acquiring the movement state of the own vehicle;
Obstacle detection means for detecting an obstacle present in the forward direction of the host vehicle;
An obstacle determination means for determining whether the obstacle detected by the obstacle detection means is an obstacle that may contact the host vehicle within a predetermined time;
Avoidance operation estimation means for estimating whether or not the driver has performed a steering operation for obstacle avoidance;
When the obstacle determining means determines that the obstacle has a risk of contact, if the avoidance operation estimating means estimates that the steering operation for avoiding the obstacle has not been performed, contact with the obstacle An own vehicle that determines an avoidance direction advantageous for avoidance and further calculates an own vehicle guidance amount for guiding the own vehicle in an advantageous avoidance direction within a range of allowable lateral movement set on the own vehicle traveling path A guidance amount calculating means;
Own vehicle guidance control means for realizing the own vehicle guidance amount calculated by the own vehicle guidance amount calculation means;
I have a,
When the driver's guidance control means detects a steering operation for avoidance by the driver and determines that it is necessary to avoid the vehicle's travel path, the vehicle's guidance control means A vehicular driving support device that calculates a guidance amount that widens a range of movement and assists a driver's steering operation . Provided with road information detection means for acquiring road boundary information ahead of the vehicle,
The own vehicle guidance amount calculating means is
2. The vehicle driving support device according to claim 1, wherein a range of the allowable lateral movement amount is set based on road boundary information ahead of the host vehicle acquired by the road information detecting means. The own vehicle guidance amount calculating means is
The advantageous avoidance direction and the own vehicle guidance amount are calculated based on at least one of the own vehicle motion prediction obtained from the own vehicle state acquisition means and the obstacle motion prediction obtained from the forward obstacle detection means. The vehicle driving support apparatus according to claim 1 or 2, wherein The vehicle guidance control means is
4. The self-vehicle guidance amount is realized by using at least one of a steering mechanism capable of steering control and a braking / driving mechanism capable of braking / driving control. The vehicle driving support device described in 1. Equipped with warning means to inform the driver of the danger,
The own vehicle guidance amount calculating means is
In order to avoid obstacles, when it is determined that it is necessary to perform avoidance guidance that deviates from the vehicle traveling path, a warning is given to the driver by issuing an alarm through the alarm means. The vehicle driving support device according to any one of claims 1 to 4. A steering operation so as to deviate from the vehicle running path by the OPERATION person, a vehicle according to any one of claims 1 to 5, characterized in that the steering operation of the avoidance object of the driver Driving assistance device. If an obstacle present in the forward direction of the host vehicle is detected and the obstacle is determined to be an obstacle that may come into contact with the host vehicle within a predetermined time, it is determined whether or not there is a driving operation for avoiding the obstacle. Regardless of the wheel direction, an avoidance direction that is advantageous for avoiding contact with the obstacle is determined, and the vehicle is guided in the advantageous avoidance direction within a range of allowable lateral movement set in the own vehicle travel path. To control the steering or braking / driving force ,
If it is determined that it is necessary to avoid the vehicle driving path when the driver's steering operation is detected, the range of lateral movement allowed outside the vehicle driving path is widened and driving A driving support method for a vehicle, comprising calculating a guidance amount that assists a person's steering operation .

Images