JP2014019262A - Vehicular drive support device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To appropriately integrate a steering intention and the like of a driver and a steering support matching a target path of a travel lane even under a complex traffic environment such as an obstacle existing, preventing the driver from feeling discomfort or fear.SOLUTION: A control gain is variably set in accordance with recognized obstacle data; using the variably set control gain, an output value in accordance with a recognition form of left/light while line recognized by a forward recognition device 31 is calculated as a feed forward output value Iff; an output value in accordance with the recognition form of the left/right white line recognized by the forward recognition device 31 and with a traveling state of a vehicle itself is calculated as a feedback output value Ifb; and a value produced by adding the two output values is converted into a torque value, to output to a motor drive part 21 as a steering torque (a control amount) Ti.

Description

  The present invention relates to a vehicle driving support apparatus that sets a steering angle so as to travel appropriately based on a recognized white line.

  In recent years, various driving support devices have been developed and put to practical use in vehicles with the development of electronics technology. As such a driving support device, a white line is recognized by a camera mounted on the vehicle and a traveling path of the own vehicle is set, and a steering force is added based on this, or a driver is driven by automatic steering. Some perform load reduction. As such a driving support device, for example, in Japanese Patent Application Laid-Open No. 2007-8281 (hereinafter, Patent Document 1), in driving support control that performs lane keeping control based on a detected white line and a set virtual white line, A technique is disclosed in which, when an obstacle is detected, a lane keeping control is performed by setting a virtual white line inside the travel lane from the obstacle based on the distance from the host vehicle to the obstacle.

JP 2007-8281 A

  By the way, in the driving assistance control, in the steering assistance system in which the driver and the system interfere with each other via the steering wheel, there are cases where it is not always desirable to perform the steering automatically along the target route set by the system. For example, even if the obstacle is the same stop vehicle, there is a possibility that the door will open when there is an occupant in the vehicle, but in the absence, the possibility is low. In addition, the possibility of a pedestrian jumping out from behind an obstacle greatly varies depending on whether the environment in which the vehicle is traveling is an urban area such as a shopping street, a general road or a farm road with little traffic. Although the driver can empirically detect these complex factors and set the target route, it is extremely difficult to set the target route by grasping all these factors with the steering assist system. Therefore, in the system that performs the driving support control as disclosed in Patent Document 1 as described above, the target route set by the system and the target route intended by the driver are different, giving the driver a sense of incongruity or fear. There is a risk of it.

  The present invention has been made in view of the above circumstances, and even in a complicated traffic environment where obstacles exist, steering intention such as driver avoidance steering and steering assistance along the target route of the driving lane It is an object of the present invention to provide a vehicle driving support device capable of performing natural and stable steering support control without causing a sense of incongruity and fear to the driver by appropriately combining the above.

  One aspect of the vehicle driving support apparatus of the present invention is a vehicle equipped with a front environment recognition means for recognizing a front left and right white line and an obstacle, and a steering control means capable of setting a steering angle independently of a driver input. In the driving support apparatus, the steering control unit recognizes the feedforward control unit that calculates an output value corresponding to the recognized shape of the left and right white lines recognized by the front environment recognition unit as a feedforward output value, and is recognized by the front environment recognition unit. At least one of feedback control means for calculating an output value corresponding to the recognized shape of the left and right white lines and the traveling state of the host vehicle as a feedback output value, and weighting of the calculation term in the calculation of the feedforward output value and the feedback In the calculation of the output value, the weight of the calculation term is variably set according to the obstacle recognized by the front environment recognition means. With the control gain setting means.

  According to the vehicle driving support device of the present invention, even in a complicated traffic environment in which an obstacle exists, steering intention such as driver avoidance steering and steering assistance along the target route of the driving lane are performed. It is possible to perform natural and stable steering support control without causing a sense of incongruity or fear to the driver by appropriately merging.

1 is a configuration explanatory diagram of a vehicle steering system according to an embodiment of the present invention. FIG. It is a functional block diagram of the steering control part which concerns on one Embodiment of this invention. It is a flowchart of the steering control program which concerns on one Embodiment of this invention. It is a flowchart of the control gain setting control which concerns on one Embodiment of this invention. FIG. 5 is a flowchart continuing from FIG. 4. 6 is a flowchart continuing from FIG. 5. It is a flowchart continuing from FIG. It is a flowchart of the feedforward control which concerns on one Embodiment of this invention. It is a flowchart of the feedback control which concerns on one Embodiment of this invention. It is explanatory drawing of the curvature of the white line used by the feedforward control which concerns on one Embodiment of this invention. It is explanatory drawing of the deviation | shift amount (1st feedback control amount) from the center of the white line on either side in the driving | running | working position where the own vehicle used by feedback control which concerns on one Embodiment of this invention is estimated. It is explanatory drawing of the calculation based on the driving | running | working attitude | position of the own vehicle used by the feedback control which concerns on one Embodiment of this invention. It is explanatory drawing of the steering intention of the driver obtained from the steering torque by the driver which concerns on one Embodiment of this invention. It is explanatory drawing of the obstruction ahead of the own vehicle which concerns on one Embodiment of this invention. It is an example of the driving | running route | route when an obstruction exists in front of the own vehicle which concerns on one Embodiment of this invention. It is a time chart which shows an example of the control gain set when drive | working like FIG.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
In FIG. 1, reference numeral 1 denotes an electric power steering device in which a steering angle can be set independently of a driver input. It is rotatably supported, and one end thereof extends to the driver's seat side and the other end extends to the engine room side. A steering wheel 4 is fixed to an end portion of the steering shaft 2 on the driver's seat side, and a pinion shaft 5 is connected to an end portion extending to the engine room side.

  A steering gear box 6 extending in the vehicle width direction is disposed in the engine room, and a rack shaft 7 is inserted into and supported by the steering gear box 6 so as to be reciprocally movable. A rack (not shown) formed on the rack shaft 7 is engaged with a pinion formed on the pinion shaft 5 to form a rack and pinion type steering gear mechanism.

  The left and right ends of the rack shaft 7 protrude from the end of the steering gear box 6, and a front knuckle 9 is connected to the end via a tie rod 8. The front knuckle 9 rotatably supports left and right wheels 10L and 10R as steering wheels, and is supported by a vehicle body frame via a king pin (not shown) so as to be steerable. Accordingly, when the steering wheel 4 is operated and the steering shaft 2 and the pinion shaft 5 are rotated, the rack shaft 7 is moved in the left-right direction by the rotation of the pinion shaft 5, and the front knuckle 9 is moved to the king pin (not shown). ) And the left and right wheels 10L, 10R are steered in the left-right direction.

  Further, an electric motor 12 is connected to the pinion shaft 5 via an assist transmission mechanism 11 so that the steering torque applied to the steering wheel 4 by the electric motor 12 and a set steering angle are obtained. The appropriate steering torque is added. The electric motor 12 is driven via a motor drive unit 21 so as to have a steering torque (control amount) Ti set by a steering control unit 20 described later. The steering control unit 20 also has a steering torque assist function as described above, but the description of the steering torque assist function is omitted in this embodiment.

  The steering control unit 20 includes a front recognition device 31 as a front environment recognition means for recognizing a front left and right white line and a preceding vehicle, a vehicle speed sensor 32 for detecting a vehicle speed V, a handle angle sensor 33 for detecting a steering angle θH, a steering torque. A steering torque sensor 34 for detecting Ts, a yaw rate sensor 35 for detecting yaw rate (dθ / dt), and a lateral acceleration sensor 36 for detecting lateral acceleration Gy are connected.

  The front recognition device 31 is, for example, a set of CCD cameras that are mounted at a predetermined interval in front of a ceiling in a vehicle interior and that captures an object outside the vehicle from different viewpoints, and a stereo image that processes image data from the CCD camera. And a processing device.

  The processing of image data from the CCD camera in the stereo image processing device of the forward recognition device 31 is performed as follows, for example. First, distance information is obtained from a set of stereo image pairs taken in the traveling direction of the host vehicle captured by the CCD camera from the corresponding positional deviation amount, and a distance image is generated.

  In recognition of white line data, based on the knowledge that the white line is brighter than the road surface, the brightness change in the width direction of the road is evaluated, and the positions of the left and right white lines on the image plane are specified on the image plane. To do. The position (x, y, z) of the white line in the real space is known based on the position (i, j) on the image plane and the parallax calculated with respect to this position, that is, based on the distance information. Calculated from the coordinate conversion formula. In this embodiment, for example, as shown in FIG. 10, the coordinate system of the real space set based on the position of the host vehicle is the road surface directly below the center of the stereo camera, and the vehicle width direction is the x axis. The vehicle height direction is the y-axis, and the vehicle length direction (distance direction) is the z-axis. At this time, the xz plane (y = 0) coincides with the road surface when the road is flat. The road model is expressed by dividing the traveling lane of the host vehicle on the road into a plurality of sections in the distance direction, and connecting the left and right white lines in each section with a predetermined approximation.

  In recognition of side wall and three-dimensional object data, side wall data such as guardrails and curbs that exist along the road are compared with prestored frames (windows) such as three-dimensional side wall data and three-dimensional object data. The three-dimensional object is classified into other three-dimensional objects such as an automobile, a two-wheeled vehicle, a pedestrian, and a utility pole, and extracted. In the three-dimensional object (hereinafter also referred to as an obstacle) data, as shown in FIG. 14, the position of the three-dimensional object and the position of the left and right white lines are compared, and the three-dimensional object exists on the left side or the right side of the traveling lane. Determine whether it is a three-dimensional object. As a result, in the case of a three-dimensional object existing on the left side of the traveling lane, the distance between the three-dimensional object and the host vehicle is risk_d_l, and the distance from the preset reference position to the outermost end protruding inside the traveling lane Information such as risk_xl, speed risk_vl with respect to the width direction of the traveling lane, and the type of the three-dimensional object is detected as the three-dimensional object information. In the case of a three-dimensional object existing on the right side of the travel lane, the distance between the three-dimensional object and the host vehicle is risk_d_r, and the distance risk_xr from the preset reference position to the extreme end protruding inside the travel lane. Information such as the speed risk_vr with respect to the width direction of the traveling lane and the type of the three-dimensional object is detected as the three-dimensional object information. Information thus obtained, that is, white line data, side data such as guardrails and curbs along the road, and three-dimensional object (obstacle) data, are input to the steering control unit 20. Further, in the embodiment of the present invention, since the independent control is performed on the obstacle present on the left side and the obstacle present on the right side of the driving lane, the last subscript “l” of the suffixed code is the left side, The subscript “r” is shown on the right side.

  Then, the steering control unit 20 variably sets the control gain according to the recognized obstacle data based on each of the input signals described above, and is recognized by the front recognition device 31 using the variably set control gain. The output value corresponding to the recognized shape of the left and right white lines is calculated as a feedforward output value Iff, and the output value corresponding to the recognized shape of the left and right white lines recognized by the front recognition device 31 and the traveling state of the host vehicle is a feedback output value Ifb. The value obtained by adding these values is converted into a torque value and output to the motor drive unit 21 as a steering torque (control amount) Ti.

  Therefore, as shown in FIG. 2, the steering control unit 20 mainly includes a control gain setting control unit 20a, a feedforward control unit 20b, a feedback control unit 20c, and a steering torque calculation unit 20d as steering control means. Yes.

  The control gain setting control unit 20 a receives the image information recognized from the front recognition device 31, receives the vehicle speed V from the vehicle speed sensor 32, receives the handle angle θH from the handle angle sensor 33, and steers from the steering torque sensor 34. Torque Ts is input, and yaw rate (dθ / dt) is input from the yaw rate sensor 35. Each control gain Gff, Gfb1, Gfb2 is set according to flowcharts shown in FIGS. 4 to 7 described later, each control gain Gff is output to the feedforward control unit 20b, and the control gains Gfb1, Gfb2 are fed back to the feedback control unit 20c. It is configured to output to.

Hereinafter, the setting of each control gain Gff, Gfb1, and Gfb2 will be described with reference to the flowcharts shown in FIGS.
First, in step (hereinafter abbreviated as “S”) 201, the distance risk_d_l between the three-dimensional object existing on the left side of the travel lane and the host vehicle is less than a preset first threshold Cd1 (risk_d_l <Cd1). In addition, when the distance risk_d_r between the three-dimensional object existing on the right side of the travel lane and the host vehicle is less than a preset first threshold Cd1 (risk_d_r <Cd1), the obstacle present on both sides of the travel lane and the vehicle It is determined whether the distance to the vehicle is short.

  As a result of this determination, if risk_d_l <Cd1 and risk_d_r <Cd1 are satisfied, the process proceeds to S202, and whether or not the steering torque Ts is equal to or less than a preset torque value (−Tc) (Ts ≦ −Tc). judge.

  Here, in the embodiment of the present application, as shown in FIG. 13, when the steering torque Ts is a sign of (+), the left steering of the driver is shown, and when the sign of (−) is shown, the right steering of the driver is shown. Shall be shown. Further, the region where the steering torque Ts is −Tc <Ts <Tc is set as a dead zone where the determination of the steering direction of the driver is unstable.

  As a result of the determination in S202, if Ts ≦ −Tc and it is determined that the driver intends to avoid right in the right steering state, the process proceeds to S203, the left avoidance flag F_start_l is set to 0, and the right avoidance flag F_start_r is set to 1. And proceed to S209.

  If Ts> −Tc as a result of the determination in S202, the process proceeds to S204, in which it is determined whether the steering torque Ts is equal to or greater than a preset torque value Tc (Ts ≧ Tc).

  As a result of the determination in S204, if Ts ≧ Tc and it is determined that the driver intends to avoid left in the left steering state, the process proceeds to S205, the left avoidance flag F_start_l is set to 1, and the right avoidance flag F_start_r is set to 0. And proceed to S209.

  Further, as a result of the determination in S204, if Ts <Tc, that is, if -Tc <Ts <Tc, the process proceeds to S208, where it is determined that the driver is not in the steering state and there is no intention of avoiding both the left and right sides. Thus, both the left avoidance flag F_start_l and the right avoidance flag F_start_r are set to 0, and the process proceeds to S209.

  On the other hand, if the conditions of risk_d_l <Cd1 and risk_d_r <Cd1 are not satisfied in S201 described above, the process proceeds to S206, and it is determined whether risk_d_r <Cd1 is satisfied. When risk_d_r <Cd1 is satisfied (when the distance risk_d_r between the three-dimensional object present on the right side of the travel lane and the host vehicle is short), the processing from S202 described above is repeated.

  Conversely, if risk_d_r <Cd1 is not satisfied in S206, the process proceeds to S207, and it is determined whether risk_d_l <Cd1 is satisfied. When risk_d_l <Cd1 is satisfied (when the distance risk_d_l between the three-dimensional object existing on the left side of the travel lane and the host vehicle is short), the processing from S204 described above is repeated.

  If risk_d_l <Cd1 is not satisfied, that is, if the distance between the obstacles on both sides of the driving lane and the host vehicle is long (including the case where there is no vehicle), the process proceeds to S208, and the left avoidance flag F_start_l and the right avoidance flag Both F_start_r are set to 0 and the process proceeds to S209.

After the left avoidance flag F_start_l and the right avoidance flag F_start_r are set in any of S203, S205, and S208, the process proceeds to S209, and as an initial setting, for example, a control gain Gff related to feedforward control, feedback control as follows First and second control gains Gfb1 and Gfb2 are determined.
Gff = Gff_base−dGff_l−dGff_r (1)
Gfb1 = Gfb1_base-dGfb1_1-dGfb1_r (2)
Gfb2 = Gfb2_base-dGfb2_l-dGfb2_r (3)
Here, Gff_base, Gfb1_base, and Gfb2_base are control gain values in the case of performing normal lane keep control that is set in advance by experiment, calculation, or the like. Further, dGff_l and dGff_r are gain subtraction amounts generated in association with the risk in the left direction and the right direction in the feedforward control, respectively. This is the amount of gain subtraction that occurs with risk, and is set to 0 for each gain by default.

  After setting the control gains Gff, Gfb1, and Gfb2 in S209, the process proceeds to S210, where the left avoidance flag F_start_l = 1 is in the left steering (avoidance) state, or the right avoidance flag F_start_r = 1 and the driver is right It is determined whether it is in a steering (avoidance) state.

  In S210, if F_start_l = F_start_r = 0 and the driver has no intention of avoiding either the left or right side of the travel lane, or if the distance to the obstacle is long (S208), the program exits as it is. In this case, the subtraction amounts dGff_l, dGff_r, dGfb1_l, dGfb1_r, dGfb2_l, and dGfb2_r for each control gain to be set are all 0, and normal lane keep control is performed.

In S210 described above, if F_start_l = 1 and the driver is in the left steering (avoidance) state, or if the right avoidance flag F_start_r = 1 and the driver is in the right steering (avoidance) state, the process proceeds to S211 and the following (4 ) To (9), the control gain subtraction amounts dGff_l, dGff_r, dGfb1_l, dGfb1_r, dGfb2_l, and dGfb2_r are calculated.
First, the control gain subtraction amount dGff_l related to the feedforward control for the left obstacle is calculated by, for example, the following equation (4).
dGff_l = G1, risk_xl + G2, risk_vl + risk_kind_l (4)
Further, the subtraction amount dGff_r of the control gain related to the feedforward control with respect to the obstacle on the right side is calculated by, for example, the following equation (5).
dGff_r = G1, risk_xr + G2, risk_vr + risk_kind_r (5)
The control gain subtraction amount dGfb1_l related to the feedback control for the left obstacle is calculated by, for example, the following equation (6).
Gfb1_l = G1, risk_xl + G2, risk_vl + risk_kind_l (6)
The subtraction amount dGfb1_r of the control gain related to the feedback control for the right obstacle is calculated by the following equation (7), for example.
dGfb1_r = G1.risk_xr + G2.risk_vr + risk_kind_r (7)
The subtraction amount dGfb2_l of the control gain related to the feedback control for the left obstacle is calculated by, for example, the following equation (8).
dGfb2_l = G1 / risk_xl + G2 / risk_vl + risk_kind_l (8)
The subtraction amount dGfb2_r of the control gain related to the feedback control for the right obstacle is calculated by the following equation (9), for example.
dGfb2_r = G1.risk_xr + G2.risk_vr + risk_kind_r (9)
In the above formulas (4) to (9), G1 and G2 are values set in advance by experiments, calculations, etc., and risk_kind_r is a risk set in advance for each type of obstacle, For example, it is set with reference to a map or the like set in advance as (pedestrian risk)> (motorcycle risk)> (automobile risk).

  The control gain subtraction amounts dGff_l, dGff_r, dGfb1_l, dGfb1_r, dGfb2_l, dGfb2_r calculated above are input to the equations (1), (2), and (3). Therefore, the amount of control gain subtraction for feedforward control and feedback control varies depending on the presence or absence of obstacles, their position, speed, and type.

  Next, the process proceeds to S212, and the control gain Gff related to the feedforward control and the first and second control gains Gfb1 and Gfb2 related to the feedback control are calculated and output, for example, by the above-described equations (7) to (9). .

  That is, as described above, the control gain Gff is a control gain for calculating the output value corresponding to the recognized shape of the left and right white lines as the feedforward output value Iff by the feedforward control, and the control gain Gff is the normal control gain Gff_base. On the other hand, it is determined by reducing the value according to the equation (1), whereby the control amount by the steering assist control is reduced, and the interference with the steering of the driver is reduced. Details of the feedforward control will be described later.

  Similarly, the control gains Gfb1 and Gfb2 are control gains for calculating an output value according to the recognized shape of the left and right white lines and the traveling state of the host vehicle as the feedback output value Ifb by feedback control. This is determined by reducing the gains Gfb1_base and Gfb2_base according to the expressions (2) and (3), whereby the control amount by the steering assist control is reduced and the interference with the steering of the driver is reduced. Details of the feedback control will be described later.

  Next, the process proceeds to S213, where it is determined whether F_start_l = 1 and the driver is in the left steering (avoidance) state. If F_start_l = 1 (the driver is in the left steering (avoidance) state), the process proceeds to S214 and the right side of the driving lane The distance risk_d_r between the three-dimensional object existing in the vehicle and the host vehicle is less than a preset second threshold Cd2 (risk_d_r <Cd2), and the obstacle and the vehicle present on the right side of the travel lane are close to each other Alternatively, it is determined whether the yaw rate (dθ / dt) or the handle angle θH is in a constant state.

  As a result of this determination, if risk_d_r <Cd2, or the yaw rate (dθ / dt) or the handle angle θH becomes constant, the process proceeds to S216, and if this condition is not satisfied, the process returns to S212. Repeat the process.

  On the other hand, if the condition of F_start_l = 1 is not satisfied in S213 (that is, if the driver is in the right steering (avoidance) state with F_start_r = 1), the process proceeds to S215, and the three-dimensional object and the vehicle existing on the left side of the driving lane The distance risk_d_l is less than a preset second threshold Cd2 (risk_d_l <Cd2), and the distance between the obstacle present on the left side of the driving lane and the vehicle is close, or the yaw rate (dθ / dt Or, it is determined whether or not the steering wheel angle θH is in a constant state.

  As a result of this determination, if risk_d_l <Cd2, or the yaw rate (dθ / dt) or the steering wheel angle θH becomes constant, the process proceeds to S216. If this condition is not satisfied, the process from S212 is resumed. Repeat the process.

  When the process proceeds from S214 or S215 to S216, the subtraction amount dGfb2_l of the second control gain Gfb2 related to the feedback control for the left obstacle and the subtraction amount of the second control gain Gfb2 related to the feedback control for the right obstacle dGfb2_r is reset to 0. Note that dGfb1_l and dGfb2_r are left as they are. That is, as will be described later, the second control gain Gfb2 related to the feedback control is a control gain set according to the traveling state of the host vehicle (particularly, the traveling inclination of the host vehicle with respect to the left and right white lines), and the above-described S214. Alternatively, when the vehicle posture is suitable for passing an obstacle by satisfying the condition of S215, only the feedback control with respect to the yaw angle θ of the vehicle is stopped and returned to avoid the driver. It is possible to control against disturbances such as cant and crosswind that are unexpectedly applied to the vehicle without disturbing the operation.

Proceeding to S216, Gff, Gfb1, and Gfb2 are calculated in the same manner as in equations (1) to (3) described above.
Next, the process proceeds to S217, and, for example, the control gain Gff related to the feedforward control and the first and second control gains Gfb1 and Gfb2 related to the feedback control are calculated and output by the above-described equations (7) to (9). .

  Next, the process proceeds to S218, where it is determined whether F_start_l = 1 and the driver is in the left steering (avoidance) state. If F_start_l = 1 (the driver is in the left steering (avoidance) state), the process proceeds to S219 and the right side of the driving lane The distance risk_d_r between the three-dimensional object existing in the vehicle and the host vehicle is less than a preset third threshold Cd3 (risk_d_r <Cd3), and the obstacle present on the right side of the travel lane is close to the host vehicle It is determined whether or not. As a result of this determination, if risk_d_r <Cd3, the process proceeds to S221, and if this condition is not satisfied, the processing from S216 is repeated again.

  On the other hand, if the condition of F_start_l = 1 is not satisfied in S218 (that is, if the driver is in the right steering (avoidance) state with F_start_r = 1), the process proceeds to S220, and the three-dimensional object and the vehicle existing on the left side of the driving lane Is less than a preset third threshold Cd3 (risk_d_l <Cd3), it is determined whether the distance between the obstacle present on the left side of the travel lane and the host vehicle is close. As a result of this determination, if risk_d_l <Cd3, the process proceeds to S221. If this condition is not satisfied, the process from S216 is repeated again.

  When the process proceeds from S219 or S220 to S221, the subtraction amount dGff for the obstacle on the left side of the control gain Gff related to the feedforward control, the subtraction amount dGff_r for the obstacle on the right side, and the first control gain Gfb1 related to the feedback control The subtraction amount dGfb1_l for the left obstacle and the subtraction amount dGfb1_r for the right obstacle are reset to zero. Also, a preset subtraction amount d_Gyaew is set as the subtraction amount dGfb2_l for the left obstacle and the subtraction amount dGfb2_r for the right obstacle of the second control gain Gfb2 related to feedback control.

  Next, the process proceeds to S223, and, for example, the control gain Gff related to the feedforward control and the first and second control gains Gfb1 and Gfb2 related to the feedback control are calculated and output by the above-described equations (1) to (3). .

  That is, as will be described later, when the process proceeds to S221 while satisfying the above-described conditions of S219 or S220, the obstacle is avoided and the vehicle returns to the center of the traveling lane. In such a situation, since the difference between the vehicle and the white line is generated again, the output value by the second control gain Gfb2 is a component for steering the vehicle in the direction opposite to the direction of returning the vehicle to the center of the lane. For this reason, d_Gyaew is set with respect to dGfb2_l and dGfb2_r, and the second control gain Gfb2 related to feedback control is decreased to facilitate the return of the vehicle to the center of the lane. Further, it is desirable to change the value of d_Gyaew depending on the positional relationship between the vehicle and the center of the white line. In other words, when the vehicle is sufficiently close to the center of the white line, the value of d_Gyaew is brought close to 0 and the second control gain Gfb2 is increased to promote the direction of the lane and the vehicle to be parallel. .

  Then, the process proceeds to S224, where it is determined whether or not the host vehicle is traveling substantially in the center of the travel lane. If the vehicle is traveling approximately in the center, the process proceeds to S225. If the vehicle is not traveling approximately in the center, S221 is performed. Repeat the process from.

  In S225, the subtraction amounts dGfb2_l and dGfb2_r of the second control gain Gfb2 related to the feedback control set in S221 described above are also set to 0, and in S226, Gfb2 is again calculated and output by the above-described equation (3). Then, the routine exits so as to return to the normal steering assist control. Thus, the control gain setting control unit 20a is provided as a control gain setting means.

The feedforward control unit 20b receives image information from the front recognition device 31, and receives a control gain Gff from the control gain setting control unit 20a. Then, for example, according to the following equation (22), an output value corresponding to the recognized shape of the left and right white lines is calculated as a feedforward output value Iff and output to the steering torque calculator 20d.
Iff = Gff. (Rr + Rl) / 2 (22)
Here, Rr is a curvature component due to the right white line, and Rl is a curvature component due to the left white line. Specifically, the curvature components Rr and Rl of the left and right white lines are the coefficients of the quadratic terms calculated by the second least square method for the candidate points constituting each of the left and right white lines as shown in FIG. It is determined by using. For example, when the white line is approximated by a quadratic expression of x = A · z ² + B · z + C, the value of A is used as the curvature component. The white line curvature components Rr and Rl may be the respective white line curvatures themselves.

  Next, the feedforward control executed by the feedforward control unit 20b will be described with reference to the flowchart of FIG.

  First, in step S301, the control gain Gff is read from the control gain setting control unit 20a.

  In step S302, the curvature component Rl due to the left white line and the curvature component Rr due to the right white line are calculated.

  Then, the process proceeds to S303, where the feedforward output value Iff is calculated according to the above equation (22) and output to the steering torque calculation unit 20d. Thus, the feedforward control part 20b is provided as a feedforward control means.

The feedback control unit 20 c receives image information from the front recognition device 31, receives a vehicle speed V from the vehicle speed sensor 32, receives a yaw rate (dθ / dt) from the yaw rate sensor 35, and receives a lateral acceleration Gy from the lateral acceleration sensor 36. The control gains Gfb1 and Gfb2 are input from the control gain setting control unit 20a. And, for example, according to the following equation (23), the shape of the left and right white lines based on the predicted travel position of the host vehicle recognized by the front recognition device 31 and the travel state of the host vehicle (the travel posture with respect to the white line) The corresponding output value is calculated as a feedback output value Ifb and output to the steering torque calculator 20d.
Ifb = Gfb1. (Xr-xl-xv)
+ Gfb2 · ((θtr−θ) + (θtl−θ)) / 2 (23)
In the first term of the equation (23), xv is an x coordinate in the z coordinate of the forward gazing point of the vehicle. In the present embodiment, for example, as shown in FIG. 11, the forward gazing point is predicted that the host vehicle exists after a preset preview time T (for example, 1.2 seconds) has elapsed. Is a point. The z coordinate zv at the forward gazing point is calculated by, for example, zv = T · V. In addition, it is good also as a point of the distance set ahead simply.

Accordingly, the x coordinate xv of the forward gazing point can be calculated by the following equation (24), for example.
xv = xi + V · (β + θ) · T
+ (1/2) · ((dθ / dt) + (dβ / dt)) · V · T ² (24)
Here, xi is the current x coordinate of the vehicle, and (dβ / dt) is the vehicle slip angular velocity, which can be calculated by the following equation (25), for example.
(Dβ / dt) = (Gy / V) − (dθ / dt) (25)
The vehicle slip angle β can be calculated by integrating the vehicle slip angular velocity (dβ / dt).

  In the equation (23), xr is the x coordinate of the right white line in the z coordinate of the forward gazing point, and xl is the x coordinate of the left white line in the z coordinate of the forward gazing point.

  Therefore, as shown in FIG. 11, the first term of the equation (23) is an arithmetic term for the x-coordinate deviation between the forward gazing point and the center point of the left and right white lines, and the host vehicle recognized by the forward recognition device 31. This is a calculation term based on the shape of the left and right white lines based on the predicted travel position.

Further, as shown in FIG. 12, θtr is the inclination of the host vehicle with respect to the right white line based on the image information from the front recognition device 31, and θtl is the left side based on the image information from the front recognition device 31, as shown in FIG. This is the inclination of the vehicle with respect to the white line. These θtr and θtl are, for example, the coefficient of the primary term calculated by the quadratic least square method with respect to the white line candidate point obtained from the image information (that is, x = A · z ² + B · z + c) is used.

  Therefore, the second term of the equation (23) is a calculation term for the traveling posture of the host vehicle with respect to the white line recognized by the front recognition device 31, as shown in FIG.

  Next, feedback control executed by the feedback control unit 20c will be described with reference to the flowchart of FIG.

  First, in step S401, control gains Gfb1 and Gfb2 are read from the control gain setting control unit 20a.

  Next, in S402, the z coordinate zv of the forward gazing point is calculated.

  Next, proceeding to S403, the x coordinate xv of the forward gazing point is calculated by, for example, the above-described equation (24).

  Next, in S404, the left white line x-coordinate xl and the right white line x-coordinate xr at the forward gazing point are calculated.

  Next, proceeding to S405, the inclination θtl with respect to the left white line of the own vehicle, the inclination θtr with respect to the right white line of the own vehicle, and the yaw angle θ of the own vehicle are calculated.

  Then, the process proceeds to S406, where the feedback output value Ifb is calculated by the above equation (23) and output to the steering torque calculator 20d. Thus, the feedback control unit 20c is provided as feedback control means.

The steering torque calculator 20d receives the feedforward output value Iff from the feedforward controller 20b and the feedback output value Ifb from the feedback controller 20c. Then, for example, the steering torque Ti is calculated by the following equation (26), and is output to the motor drive unit 21.
Ti = Gt · (Iff + Ifb) (26)
Here, Gt is a preset conversion coefficient.

In the steering control unit 20 configured as described above, steering control is executed as shown in FIG.
First, in S101, the control gain setting control unit 20a performs setting control of the control gains Gff, Gfb1, and Gfb2 according to the flowcharts shown in FIGS.

  Next, in S102, the feedforward control unit 20b calculates the feedforward output value Iff according to the flowchart shown in FIG.

  Next, proceeding to S103, the feedback control unit 20c calculates the feedback output value Ifb according to the flowchart shown in FIG.

  Then, the process proceeds to S104, where the steering torque calculation unit 20d calculates the steering torque Ti by the above-described equation (26) and outputs it to the motor drive unit 21.

  As described above, according to the embodiment of the present invention, the control gain is variably set according to the recognized obstacle data, and the left and right white lines recognized by the front recognition device 31 using the variably set control gain. An output value corresponding to the recognized shape is calculated as a feedforward output value Iff, an output value corresponding to the recognized shape of the left and right white lines recognized by the front recognition device 31 and the traveling state of the host vehicle is calculated as a feedback output value Ifb, A value obtained by adding these is defined as a control amount. At this time, in particular, the driver's intention to avoid the recognized obstacle is taken into consideration, and the control amount for the lane keeping by the driving support device is set to be subtracted. Therefore, even in a complicated traffic environment where an obstacle exists. Even so, steering intentions such as avoidance steering of the driver and steering assistance along the target route of the driving lane are appropriately combined to perform natural and stable steering assistance control without giving the driver a sense of incongruity or fear. Is possible.

  Specifically, in a driving environment as shown in FIG. 15, there is an obstacle such as a pedestrian on the left side of the driving lane and an obstacle such as a vehicle on the right side of the driving lane. In the example where priority is given to avoiding the vehicle, the host vehicle takes the travel line indicated by A.

  An example of each control gain Gff, Gfb1, Gfb2 set in the traveling environment shown in FIG. 15 will be described with reference to the time chart of FIG.

  First, until entering the area I in FIG. 15, the distance to the left and right obstacles is long, and the control gains Gff, Gfb1, and Gfb2 set by the driving support device remain unchanged, and normal lane keeping is performed. Steering support control is executed.

  After that, when entering the area II, the distance to the obstacle is reduced, and the control gains Gff, Gfb1, Gfb2 are subtracted. In particular, the driver's pedestrian (has a higher risk than the vehicle) It is allowed to deviate from the center of the driving lane to the right with respect to the intention of avoiding.

  Next, when entering the area III and avoiding pedestrians, the subtraction from the control gains Gff and Gfb1 continues to allow the driver to deviate from the center of the driving lane to the right as it is intended by the driver. The intention to avoid the driver's pedestrian (having a higher degree of danger than the vehicle) is reflected as it is, but after avoiding the obstacle, the posture of the host vehicle is moved to the ideal center position of the driving lane. Therefore, the subtraction of the control gain Gfb2 is stopped (dGfb2_l = dGfb2_r = 0 in S216).

  Next, in a state where the vehicle enters the IV region and finishes avoiding the pedestrian, the control returns to the center of the traveling lane. Therefore, subtraction of the control gains Gff and Gfb1 is stopped (dGff_l = dGff_r = dGfb1_l = dGfb1_r = 0 in S221), and the control is again performed to prevent sudden change of the vehicle posture of the host vehicle. Only the subtraction of the gain Gfb2 is executed and the steering assist control is performed.

  Then, when the vehicle enters the region V and reaches the center of the traveling lane, the subtraction of the control gain Gfb2 is also stopped to perform the control by the steering assist control as it is (dGfb2_l = dGfb2_r = 0 in S225).

  Although the embodiment of the present invention has been described with an example in which the steering system is configured by the normal electric power steering apparatus 1, it goes without saying that the present invention can also be applied to a steering system of a steer-by-wire mechanism.

  Further, although the embodiment of the present invention has been described with an example using image information from a stereo camera, image information from a monocular camera may be used.

  Furthermore, in the embodiment of the present invention, the control gain change in the control gain setting control unit 20a of the steering control unit 20 is delayed to some extent through a filter with a predetermined time constant so that the driver does not feel uncomfortable. It is desirable to change it.

  Further, the setting of the control gain subtraction amounts dGff_l, dGff_r, dGfb1_l, dGfb1_r, dGfb2_l, dGfb2_r of the control gain setting control unit 20a of the steering control unit 20 in the embodiment of the present invention is merely an example. , Risk risk_kind_l preset for each obstacle type, distance risk_xl from the preset reference position of the obstacle to the extreme end projecting inside the lane, It is calculated using three of the speed risk_vl in the width direction, but any one, any two, or further obstacle information (relative speed etc. with the own vehicle) is used. May be calculated. In the embodiment of the present invention, the gain Gff related to the feedforward control may be set to separate gains on the left and right by the sign of the curvature (Rr + Rl) / 2 of the white line imaged by the camera. Similarly, in the first control gain Gfb1 related to feedback control, the shape of the left and right white lines based on the predicted travel position of the host vehicle recognized by the forward recognition device 31 and the travel state of the host vehicle (with respect to the white line) Depending on the sign of the value of (xr−xl−xv) corresponding to the (traveling posture), the left and right gains may be set. Similarly, in the gain Gff2 related to the feedforward control, separate left and right gains may be set according to the inclination of the white line and the vehicle ((θtr−θ) + (θtl−θ)).

  In the embodiment of the present invention, control gain Gff related to feedforward control and first and second control gains Gfb1 and Gfb2 related to feedback control are calculated and controlled. Alternatively, steering assist control may be executed by setting any two control gains.

DESCRIPTION OF SYMBOLS 1 Electric power steering apparatus 4 Steering wheel 10L, 10R Wheel 12 Electric motor 20 Steering control part (steering control means)
20a Control gain setting control unit (control gain setting means)
20b Feed forward control unit (feed forward control means)
20c Feedback control unit (feedback control means)
20d Steering torque calculation unit 21 Motor drive unit 31 Front recognition device (front environment recognition unit)
32 Vehicle speed sensor 33 Steering angle sensor 34 Steering torque sensor 35 Yaw rate sensor 36 Lateral acceleration sensor

Claims (5)

In a vehicle driving support device comprising a front environment recognition means for recognizing a white line in front and an obstacle and a steering control means capable of setting a steering angle independently of a driver input,
The steering control means includes
Feedforward control means for calculating an output value corresponding to the recognized shape of the left and right white lines recognized by the forward environment recognition means as a feedforward output value;
Having at least one of feedback control means for calculating an output value corresponding to the recognized shape of the left and right white lines recognized by the front environment recognition means and the traveling state of the host vehicle as a feedback output value;
Control gain setting means for variably setting the weight of the calculation term in the calculation of the feedforward output value and the weight of the calculation term in the calculation of the feedback output value according to the obstacle recognized by the forward environment recognition means A vehicle driving support device.   2. The feedback control means according to claim 1, wherein the feedback control means calculates the feedback output value based on at least one of the recognized travel position of the recognized vehicle on the left and right white lines and the travel attitude of the vehicle. Vehicle driving support device.   The control gain setting means variably sets the weighting according to the driver's intention to avoid the recognized obstacle, the distance from the own vehicle to the obstacle, and the risk level for each obstacle set in advance. The driving support device for a vehicle according to claim 1 or 2.   4. The degree of danger for each obstacle is set by at least one of the type of each obstacle, the speed in the traveling lane width direction, and the protruding position to the center of the traveling lane. Vehicle driving support device.   The variable setting of the weighting according to the obstacle in the control gain setting means is performed by setting the obstacle present on the left side of the traveling lane and the obstacle existing on the right side of the traveling lane independently. The vehicle driving support device according to any one of claims 1 to 4.

Images