Vehicle driving assistance device

Abstract

To allow for realizing reduction of the convergence time in feedback control when the own vehicle is driven along a target traveling route.SOLUTION: A driving assistance device includes a neutral point learning calculation unit configured to set a neutral point learning correction value that corrects a neutral point of a steering angle indication value output from a steering support control unit. The neutral point learning calculation unit calculates a first estimated lateral acceleration based on a steering angle and the speed of the own vehicle, calculates a second estimated lateral acceleration based on the speed of the own vehicle, and a lane curvature and an opposite lane curvature recognized by a traveling environment recognition unit, calculates a lateral acceleration difference from a difference between the first estimated lateral acceleration and the second estimated lateral acceleration, calculates a steering angle difference based on the lateral acceleration difference and the speed of the own vehicle, and sets the steering angle difference as the neutral point learning correction value.SELECTED DRAWING: Figure 4A

Description

[Technical field]

[0001] The present invention relates to a driving assistance device for a vehicle. [Background technology]

[0002] In recent years, driving assistance devices that assist the driver in driving operations have been put into practical use in vehicles, such as automobiles, in order to reduce the burden of driving operations on the driver and to improve safety. One of these driving assistance functions is automatic steering control. Known examples of automatic steering control include lane departure prevention (LDP) control and active lane keeping (ALK) control. The LDP control and ALK control use a forward recognition sensor, such as a stereo camera, to recognize the lane markings that separate the left and right sides of the lane in which the vehicle is traveling, and perform steering control so that the vehicle travels along the lane without deviating from the markings.

[0003] In automatic steering control, the steering device is driven so that the vehicle travels along the lane. The steering angle of the steering device is detected by a steering angle sensor, and feedback control is performed so that the steering angle converges to a target steering angle.

[0004] In the feedback control of the steering angle, if the neutral point of the steering angle is shifted, the detection value of the steering angle sensor will deviate from the actual steering angle, resulting in the feedback control amount being set to an excessive or insufficient steering angle for the curvature of the road.

[0005] Therefore, it is necessary to constantly learn the center point of the steering angle sensor. For example, Patent Document 1 (JP 2018-75946 A) discloses a technology in which, when it is determined that the vehicle is traveling straight along a road, the steering angle detected by the steering angle sensor at that time is learned as the neutral point. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] JP 2018-75946 A Summary of the Invention [Problem to be solved by the invention]

[0007] Conventionally, driving assistance devices control the steering angle by outputting a command steering torque (target steering torque) to an electric motor of an electric power steering (EPS) device. However, recent driving assistance devices tend to switch the control of the electric motor of the EPS device from steering torque control to steering angle control.

[0008] Responsiveness is improved by switching the electric motor of the EPS device from steering torque control to steering angle control. Steering angle control is performed based on the neutral point of the steering angle. Therefore, it is necessary to learn the neutral point of the steering angle with high accuracy.

[0009] In the technology disclosed in the cited document 1, the neutral point of the steering angle is learned only under the condition that the vehicle is traveling straight along a straight road. Under these conditions, when the straight road is short or the road has a series of curves, the driving assistance device has few opportunities to learn the neutral point of the steering angle, making it difficult to learn the neutral point with high accuracy.

[0010] As a result, in lane departure prevention (LDP) control executed by the driving support device, a difference occurs between the left and right in the maximum amount when the vehicle deviates from the left and right marking lines due to the shift in the neutral point of the steering angle. Also, in active lane keeping (ALK) control executed by the driving support device, a difference occurs between the left and right in the convergence of feedback control for driving the vehicle along the target traveling path due to the shift in the neutral point of the steering angle.

[0011] The driving assistance control device executes the above-mentioned LDP control and ALK control not only on straight roads but also on curved roads. In the technology disclosed in the cited document 1, when the vehicle is traveling on a curved road, the controllability of the LDP control and ALK control may deteriorate because the neutral point of the steering angle is not learned.

[0012] The present invention aims to provide a vehicle driving assistance device that is capable of learning the neutral point of the steering angle not only when the vehicle is traveling on a straight road but also when traveling on a curved road, and that can shorten the convergence time in feedback control when the vehicle is traveling along a target course. [Means for solving the problem]

[0013] The present invention relates to a driving assistance device for a vehicle including a steering angle detection unit that detects a steering angle of the host vehicle, a vehicle speed detection unit that detects the host vehicle speed, a driving environment recognition unit that recognizes the driving environment in front of the host vehicle, a steering drive unit that steers left and right steering wheels, and a steering assist control unit that outputs a steering angle command value to the steering drive unit so that the host vehicle travels along the lane curvature recognized by the driving environment recognition unit. The device further includes a neutral point learning calculation unit that sets a neutral point learning correction value that corrects a neutral point of the steering angle command value output from the steering assist control unit, and the neutral point learning calculation unit calculates a first estimated lateral acceleration calculation unit that calculates a first estimated lateral acceleration received by the host vehicle based on the steering angle and the host vehicle speed detected by the vehicle speed detection unit. a second estimated lateral acceleration calculation unit that calculates a second estimated lateral acceleration to be applied to the host vehicle based on the host vehicle speed detected by the vehicle speed detection unit and the lane curvature and lane to lane curvature recognized by the driving environment recognition unit; a lateral acceleration difference calculation unit that calculates a lateral acceleration difference from a difference between the first estimated lateral acceleration calculated by the first estimated lateral acceleration calculation unit and the second estimated lateral acceleration calculated by the second estimated lateral acceleration calculation unit; a steering angle difference calculation unit that calculates a steering angle difference based on the lateral acceleration difference calculated by the lateral acceleration difference calculation unit and the host vehicle speed detected by the vehicle speed detection unit; and a neutral point learning correction value setting unit that sets the steering angle difference calculated by the steering angle difference calculation unit as the neutral point learning correction value. Effect of the Invention

[0014] According to the present invention, the neutral point learning calculation unit, which sets a neutral point learning correction value for correcting the neutral point of the steering angle command value output from the steering assist control unit, calculates a first estimated lateral acceleration based on the steering angle and the vehicle speed, calculates a second estimated lateral acceleration to be received by the host vehicle based on the vehicle speed and the lane curvature and lane curvature recognized by the driving environment recognition unit, calculates a lateral acceleration difference from the difference between the first estimated lateral acceleration and the second estimated lateral acceleration, calculates a steering angle difference based on the lateral acceleration difference and the host vehicle speed, and sets the steering angle difference as a neutral point learning correction value for correcting the neutral point of the steering angle command value output from the steering assist control unit. This makes it possible to learn the neutral point of the steering angle not only when the host vehicle is traveling on a straight road but also on a curved road, and shortens the convergence time in feedback control when the host vehicle is traveling along a target traveling course. [Brief description of the drawings]

[0015] [Figure 1] Schematic diagram of the main components of a vehicle equipped with a driving assistance device [Diagram 2] Functional block diagram of the camera unit and steering assist unit [Diagram 3] Functional block diagram of actual steering angle calculation unit [Figure 4A] Functional block diagram of neutral point learning calculation unit (part 1) [Figure 4B] Functional block diagram of neutral point learning calculation unit (part 2) [Diagram 5] FIG. 11 is an explanatory diagram showing the relationship between the lane yaw angle and the steering angle with respect to the lane curvature. [Figure 6A] FIG. 2 is an explanatory diagram showing a state in which the vehicle is traveling on a straight road; [Figure 6B] FIG. 2 is an explanatory diagram showing a state in which the vehicle is traveling on a curved road. [Figure 7] FIG. 1 is an explanatory diagram showing the behavior of a vehicle traveling on a straight road having a transverse gradient; [Figure 8] An explanatory diagram showing the two-wheel model [Figure 9]A time chart showing lateral acceleration based on steering angle and lateral acceleration based on lane curvature when driving on a straight road and a curved road. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In Fig. 1, a vehicle (own vehicle) M has left and right front wheels FL, FR and left and right rear wheels RL, RR. The left and right front wheels FL, FR are steered wheels. If the own vehicle M is a four-wheel drive vehicle, the wheels FL, FR, RL, RR are driven wheels.

[0017] The left and right front wheels FL, FR are connected to a steering mechanism 1, such as a rack and pinion mechanism, via tie rods 2. The base of a steering shaft 3 is connected to this steering mechanism 1. A handle 4 is fixed to the tip of the steering shaft 3. An electric power steering (EPS) motor 5 is connected to the steering shaft 3 via a transmission mechanism (not shown).

[0018] The vehicle M is also equipped with a driving support device 10. The driving support device 10 has a steering support unit 11. The steering support unit 11 and a lane recognition unit 24 of a camera unit 21, which will be described later, are configured with a microcontroller having a CPU, a RAM, a ROM, a rewritable non-volatile memory (flash memory or EEPROM), and peripheral devices. The RAM is provided as a work area for the CPU, and various data in the CPU are temporarily stored. The ROM stores programs and fixed data required for executing each process in the CPU. The CPU is also called an MPU (Microprocessor) or a processor. Instead of the CPU, a GPU (Graphics Processing Unit) or a GSP (Graph Streaming Processor) may be used. Alternatively, the CPU, the GPU, and the GSP may be selectively combined.

[0019] The steering assist unit 11 executes steering assist control in accordance with a program stored in a ROM. The steering assist control includes lane departure prevention (LDP) control and active lane keeping (ALK) control.

[0020] A lateral acceleration sensor 12 as a lateral acceleration detection section, a vehicle speed sensor 13 as a vehicle speed detection section, a yaw rate sensor 14 as a yaw rate detection section, and a steering angle sensor 15 as a steering angle detection section are connected to the input side of the steering assist unit 11. The lateral acceleration sensor 12 detects the lateral acceleration αy generated when the host vehicle M travels on a curved road or the like. The vehicle speed sensor 13 detects the vehicle speed (host vehicle speed) V of the host vehicle M. The yaw rate sensor 14 detects the yaw rate γ generated in the vehicle body. The steering angle sensor 15 detects the steering angle β [rad] from the rotation angle of the steering shaft 3. In addition, an EPS motor 5 as a steering drive section is connected to the output side of the steering assist unit 11.

[0021] Furthermore, a camera unit 21 serving as a driving environment recognition unit is connected to the input side of the steering assist unit 11. As shown in FIG. 2, the camera unit 21 has an on-vehicle camera 22 and also has an image processing unit 23 and a lane recognition unit 24 built in. The on-vehicle camera 22 is configured as a stereo camera consisting of a main camera 22a and a sub-camera 22b. Both cameras 22a and 22b are fixed with a predetermined baseline length. Both cameras 22a and 22b are horizontally installed at equal intervals from the center in the vehicle width direction to the left and right, for example, above the rearview mirror at the front of the vehicle and close to the windshield.

[0022] Each of the cameras 22a and 22b is provided with a color image sensor equipped with a color CCD or color CMOS. The color image sensor provided in each of the cameras 22a and 22b captures a three-dimensional color image of the driving environment ahead in the traveling direction, such as the lane (driving lane) in which the vehicle M is traveling and the left and right dividing lines that divide it. The image processing unit 23 converts a pair of analog images captured by each of the cameras 22a and 22b into a digital image with a predetermined brightness gradation. The image processing unit 23 generates reference image data from the output signal of the main camera 22a. The image processing unit 23 generates comparison image data from the output signal of the sub-camera 22b. Then, based on the parallax between the reference image data and the comparison image data, distance data (distance from the vehicle to the object) of the same object in both images is obtained.

[0023] The lane recognition unit 24 sets, on the left and right of the lane, demarcation lines that demarcate the left and right of the driving lane, recognized by a method such as pattern matching, on the virtual road plane generated based on the reference image data and the comparison image data transmitted from the image processing unit 23, on the left and right sides with the center in the vehicle width direction of the vehicle M as a reference. Then, the distance between the inner edges of the left and right demarcation lines (lane width) is detected based on the distance data. This lane information (distance information from the left and right demarcation lines to the vehicle M, lane width information) is read by the steering assist unit 11.

[0024] The steering assist unit 11 recognizes the curvature (lane curvature) 1 / Rl of the center of the lane width of the traveling lane based on the lane information transmitted from the lane recognition unit 24, sets a target traveling path on this lane curvature 1 / Rl, and performs steering control so that the host vehicle M travels along this target traveling path. Note that when the traveling lane is straight, the lane curvature 1 / Rl is 0.

[0025] Incidentally, when the steering assist unit 11 executes steering control, the steering angle β is set based on the neutral point (β=0 [rad]), which is the position of the steering shaft 3 when the host vehicle M is traveling straight. The steering angle sensor 15 detects the steering angle β based on the neutral point. Therefore, if the neutral point is misaligned, the travel trajectory of the host vehicle M is likely to deviate significantly from the target travel path, and the convergence of the feedback control deteriorates.

[0026] In particular, the steering assist unit 11 executes steering control not only on a straight road but also on a curved road. Therefore, even when the host vehicle M is traveling on a curved road, it is necessary to constantly learn the neutral point that is the reference for the steering angle β.

[0027] The steering assist unit 11 always learns the neutral point that is the reference for the steering angle β, regardless of whether the road is straight or curved. The steering assist unit 11 estimates the lateral acceleration based on two different basic parameters. Then, the neutral point of the steering angle is estimated from the difference between these two lateral accelerations. The two different parameters are the steering angle β detected by the steering angle sensor 15 and the lane curvature 1 / Rl recognized by the camera unit 21 (see FIG. 5). Then, the neutral point of the steering angle is estimated from the difference between the lateral acceleration calculated based on the steering angle β and the lateral acceleration calculated based on the lane curvature 1 / Rl.

[0028] 2, the steering assist unit 11 includes an actual steering angle calculation unit 11a and a neutral point learning calculation unit 11b as functions for estimating the neutral point of the steering angle. The driving assist unit 11 also includes a steering assist control unit 11c that executes steering assist control. The steering assist control unit 11c drives the EPS motor 5 to execute appropriate steering control so that the host vehicle M travels in the travel lane without deviating from the left and right dividing lines.

[0029] <Actual steering angle calculation unit 11a> As shown in Fig. 3, the actual steering angle calculation unit 11a includes a transverse gradient estimation unit 31 and an actual steering angle estimation unit 32. The actual steering angle calculation unit 11a estimates an actual steering angle (actual steering angle) βr by subtracting the counter steering angle from the steering angle β detected by the steering angle sensor 15. The actual steering angle calculation unit 11a includes the transverse gradient estimation unit 31 and the actual steering angle estimation unit 32 as a function of estimating the actual steering angle βr.

[0030] The transverse gradient estimation unit 31 estimates the transverse gradient θs [rad] of the lane on which the host vehicle M is traveling. Fig. 6A shows a state in which the host vehicle M is traveling on a straight road on a horizontal road surface without any external disturbance. In Fig. 6A, if the neutral point of the steering angle sensor 15 is set correctly, the steering angle β detected by the steering angle sensor 15 will be 0 [rad] (neutral point).

[0031] Fig. 6B shows a state in which the host vehicle M is traveling on a curved road on a horizontal road surface without any disturbances. In Fig. 6B, if the neutral point of the steering angle sensor 15 is set correctly, the steering angle β detected by the steering angle sensor 15 is a steering angle at which the host vehicle M travels along the curved road.

[0032] 7, when the host vehicle M is traveling on a straight road (1 / Rl=0) having a cross gradient θs, the host vehicle M experiences a drift phenomenon (a phenomenon in which the host vehicle M drifts downward in a slope) as indicated by the dashed arrow. The steering assist unit 11 adds a counter steering angle to the feedback control to offset the lateral force component Ff acting on the host vehicle M due to the drift phenomenon. This allows the host vehicle M to travel along a straight road.

[0033] When the host vehicle M is traveling straight on a straight road having a cross gradient θs, the steering assist unit 11 may erroneously determine that the steering angle command value output to the EPS motor 5 indicates the neutral point, and may perform learning correction on the steering angle β detected by the steering angle sensor 15. As a result, a deviation occurs in the neutral point when the host vehicle M is traveling on a flat road with a cross gradient θs of approximately 0 [deg].

[0034] The cross slope estimation unit 31 estimates the cross slope θs of the travel lane. As shown in FIG. 7, the cross slope θs [rad] is expressed as follows: θs=sin[(V·γ-αy) / G] -1 …(1) Here, V is the vehicle speed detected by the vehicle speed sensor 13, γ is the yaw rate detected by the yaw rate sensor 14, αy is the lateral acceleration detected by the lateral acceleration sensor 12, and G is the gravitational acceleration. Incidentally, the disturbance received by the vehicle M can be calculated from the transverse gradient θs (=m·Gsinθs).

[0035] The transverse gradient θs calculated by the transverse gradient estimation unit 31 is read by the actual steering angle estimation unit 32. The actual steering angle estimation unit 32 subtracts a transverse gradient component based on the transverse gradient θs calculated by the transverse gradient estimation unit 31 from the steering angle β detected by the steering angle sensor 15. That is, when the vehicle M is traveling on a road surface having a transverse gradient θs, the steering assist unit 11 adds a counter steering angle βan that counters the transverse gradient θs to the feedback control in order to make the vehicle M travel along the travel lane. Therefore, the actual steering angle estimation unit 32 first calculates the counter steering angle βan, and then calculates the actual steering angle βr excluding the counter steering angle βan.

[0036] Figure 8 shows a two-wheel model that is commonly used in vehicle dynamics models. The lateral force Ff generated in both front wheels FR, FL when the host vehicle M is traveling on a road surface with a transverse gradient θs is expressed as follows: 2Ff=-lr / (lf+lr)m·Gsinθs …(2) Here, lr is the distance from the center of gravity of the host vehicle M to the rear wheel axle, and lf is the distance from the center of gravity of the host vehicle M to the front wheel axle.

[0037] The counter steering angle βan is a steering angle required to offset the lateral force 2Ff and move the host vehicle M straight. In this embodiment, the correlation between the counter steering angle βan and the lateral force 2Ff is obtained in advance for each vehicle model by simulation or the like. A map showing the correlation between the counter steering angle βan and the lateral force 2Ff is stored in advance in the non-volatile memory of the steering assist unit 11.

[0038] Then, the actual steering angle estimation unit 32 calculates the actual steering angle βr by subtracting the counter steering angle βan from the steering angle β detected by the steering angle sensor 15 (βr=β-βr). This actual steering angle βr is read by the neutral point learning calculation unit 11b. Incidentally, when the host vehicle M is traveling on a road surface having a transverse gradient θs, the camera unit 21 tilts together with the host vehicle M. Therefore, the image captured by the camera unit 21 is hardly affected by the transverse gradient θs.

[0039] <Neutral point learning calculation unit 11b> As shown in Figures 4A and 4B, this neutral point learning calculation unit 11b includes a first estimated lateral acceleration calculation unit 33, a second estimated lateral acceleration calculation unit 34, first and second limiter processing units 35a, 35b, first and second low pass filter (LPF) processing units 36a, 36b, a lateral acceleration difference calculation unit 37, a steering angle difference calculation unit 38, and a neutral point learning correction value setting unit 39.

[0040] The first estimated lateral acceleration calculation unit 33 sets the turning curvature 1 / Rm [1 / m] (see FIG. 5) of the host vehicle M by referring to a curvature conversion map based on the actual steering angle βr and the host vehicle speed V. This curvature conversion map stores turning curvature 1 / Rm [1 / m] calculated for each vehicle model by simulation or the like based on the actual steering angle βr and the host vehicle speed V. Then, the square of the turning curvature 1 / Rm and the host vehicle speed V (V 2 [(m / s) 2 ]) to calculate the first estimated lateral acceleration a1 (a1 = (1 / Rm) V 2 [m / s 2 ]).

[0041] The second estimated lateral acceleration calculation unit 34 multiplies the host vehicle speed V detected by the vehicle speed sensor 13 by the lane curvature 1 / Rl recognized by the camera unit 21 to calculate the lane yaw rate ωl (ωl=(1 / Rl)·V [rad / s]). Furthermore, it differentiates the lane yaw angle Ψ, which indicates the direction of the host vehicle M with respect to the lane, recognized by the camera unit 21, to estimate the lane yaw rate ωΨ [rad / s] (see FIG. 5). The lane yaw rate ωl and the lane yaw rate ωΨ are added together (ωl+ωΨ). Then, it multiplies this value (ωl+ωΨ) by the host vehicle speed V to calculate the second estimated lateral acceleration a2 (a2=(ωl+ωΨ)·V [m / s 2 ]).

[0042] The first and second limiter processing sections 35a, 35b perform limit processing on the first and second estimated lateral accelerations a1, a2 calculated by the first and second estimated lateral acceleration calculation sections 33, 34, thereby suppressing unnecessary excessive outputs.

[0043] The first and second LPF processing units 36a, 36b remove high-frequency noise components equal to or higher than a preset cutoff frequency from the first and second estimated lateral accelerations a1, a2 that have been subjected to limit processing by the first and second limiter processing units 35a, 35b.

[0044] The lateral acceleration difference calculation unit 37 calculates a lateral acceleration difference Δa from the difference between the first and second estimated lateral accelerations a1, a2 (Δa=a1-a2). Furthermore, the lateral acceleration difference Δa is subjected to a low-pass filter (LPF) process with a low cutoff frequency to extract only the offset component.

[0045] The lateral acceleration difference Δa is the difference between the first estimated lateral acceleration a1 due to the turning curvature 1 / Rm and the second estimated lateral acceleration a2 due to the lane curvature 1 / Rl. The turning curvature 1 / Rm is calculated from the actual steering angle βr. The lane curvature 1 / Rl is recognized based on the image captured by the camera unit 21.

[0046] The component of the counter steering angle βan that offsets the transverse gradient θs is removed from the actual steering angle βr. Therefore, the lateral acceleration component due to the counter steering angle βan shown by the dashed line in Fig. 9 is not added to the first estimated lateral acceleration a1 calculated from the actual steering angle βr shown by the solid line in Fig. 9. On the other hand, the second estimated lateral acceleration a2 shown by the dashed line in Fig. 9 is based on the lane curvature 1 / Rl recognized from the image of the camera unit 21, and does not include the factor of the transverse gradient θs.

[0047] Therefore, as shown in Fig. 9, the lateral acceleration difference Δa itself shows the same value whether the vehicle is traveling on a straight road or a curved road. When traveling on a curved road, the lateral acceleration according to the curve curvature is simply added to both estimated lateral accelerations a1 and a2 as an offset amount. In Fig. 9, the lateral acceleration generated when the vehicle M is traveling on a left curve is expressed as positive, and the lateral acceleration generated when the vehicle M is traveling on a right curve is expressed as negative.

[0048] The steering angle difference calculation unit 38 first calculates the steering angle difference Δβ based on the lateral acceleration difference Δa. 2 ] is the square of the vehicle speed V (V2 [(m / s) 2 ]) to calculate the curvature difference Δ1 / R (Δ1 / R=Δa / V2[1 / m]). Next, a steering angle conversion map is referenced based on this curvature difference Δ1 / R and the vehicle speed V to set the steering angle difference Δβ[rad]. This steering angle conversion map stores the steering angle difference Δβ calculated for each vehicle type by simulation or the like based on the curvature difference Δ1 / R and the vehicle speed V.

[0049] A neutral point learning correction value setting unit 39 converts the steering angle difference Δβ [rad] set by the steering angle difference calculation unit 38 into a steering angle difference Δβ [deg]. Then, the steering angle difference Δβ converted into degrees is set as a neutral point learning correction value kβ [deg] (kβ←Δβ).

[0050] The neutral point learning correction value kβ set by neutral point learning correction value setting unit 39 is read by the steering assist control unit 11c. When outputting a steering angle command value to the EPS motor 5, the steering assist control unit 11c takes the neutral point learning correction value kβ into account in the steering angle command value. By taking the neutral point learning correction value kβ into account in the steering angle command value, deviations of the neutral point caused by each sensor involved in steering angle control, including the steering angle sensor 15, are instantly corrected.

[0051] This allows the steering assist control unit 11c to provide a highly accurate steering angle command based on the steering angle neutral point to the EPS motor 5. As a result, the steering assist control unit 11c can achieve a reduction in the convergence time required for feedback control when the host vehicle M is caused to travel along the lane curvature 1 / Rl set based on the image captured by the camera unit 21.

[0052] Also, the transverse gradient θs is calculated based on the vehicle speed V, the yaw rate γ, and the lateral acceleration αy, as shown in the above-mentioned formula (1). Also, the lateral acceleration difference Δa is calculated without taking into account the factor of the transverse gradient θs, even when the vehicle is traveling on a road having the transverse gradient θs. Since the neutral point learning correction value kβ is added to the steering angle command value output from the steering assist control unit 11c, the deviation of the transverse gradient θs calculated by formula (1) is also corrected as a result.

[0053] In addition, the driving environment recognition unit is not limited to the camera unit 21, but may be a monocular camera or LiDAR (Light Detection And Ranging), or a combination of a monocular camera and LiDAR, as long as it can recognize the dividing lines that separate the left and right sides of the lane in which the vehicle M is traveling. [Explanation of symbols]

[0054] 1...Steering mechanism, 1 / Rl…lane curvature, 1 / Rm…turning curvature, 2…Tie rod, 3...Steering shaft, 4…Handle, 5...EPS motor, 10...Driving assistance device, 11...Steering assist unit, 11a...Actual steering angle calculation section, 11b...neutral point learning calculation unit, 11c...Steering assistance control unit, 12...Lateral acceleration sensor, 13...Vehicle speed sensor, 14...Yaw rate sensor, 15...Steering angle sensor, 21...Camera unit, 22…In-vehicle camera, 22a…Main camera, 22b…Sub camera, 23...image processing unit, 24…Lane recognition unit, 31...cross slope estimation section, 32...Actual steering angle estimation unit, 33,34...first and second estimated lateral acceleration calculation units, 35a, 35b...first and second limiter processing units, 36a, 36b...first and second LPF processing units, 37...Lateral acceleration difference calculation unit, 38...Steering angle difference calculation unit, 39...neutral point learning correction value setting unit, Ff…lateral force, FL, FR…Left and right front wheels, M…own vehicle, RL, RR...Left and right rear wheels, V...Vehicle speed, a1, a2...first and second estimated lateral acceleration, kβ: neutral point learning correction value, Ψ…yaw angle relative to lane, αy...lateral acceleration, β...Steering angle, βan…counter steering angle, βr…Actual steering angle, γ: yaw rate, Δ1 / R…curvature difference, Δa...lateral acceleration difference, Δβ: steering angle difference, θs: cross slope, ωl: lane yaw rate, ωΨ…Yaw rate relative to lane

Claims

1. A steering angle detection unit that detects a steering angle of the host vehicle; a vehicle speed detection unit that detects a vehicle speed of the host vehicle; a driving environment recognition unit that recognizes a driving environment in front of the host vehicle; A steering drive unit that steers the left and right steering wheels; a steering assist control unit that outputs a steering angle command value to the steering drive unit so that the host vehicle travels along the lane curvature recognized by the traveling environment recognition unit; A driving assistance device for a vehicle comprising: a neutral point learning calculation unit that sets a neutral point learning correction value that corrects a neutral point of the steering angle command value output from the steering assist control unit, The neutral point learning calculation unit is a first estimated lateral acceleration calculation unit that calculates a first estimated lateral acceleration applied to the host vehicle based on the steering angle and the host vehicle speed detected by the vehicle speed detection unit; a second estimated lateral acceleration calculation unit that calculates a second estimated lateral acceleration applied to the host vehicle based on the host vehicle speed detected by the vehicle speed detection unit and the lane curvature and lane-to-lane curvature recognized by the driving environment recognition unit; a lateral acceleration difference calculation unit that calculates a lateral acceleration difference from a difference between the first estimated lateral acceleration calculated by the first estimated lateral acceleration calculation unit and the second estimated lateral acceleration calculated by the second estimated lateral acceleration calculation unit; a steering angle difference calculation unit that calculates a steering angle difference based on the lateral acceleration difference calculated by the lateral acceleration difference calculation unit and the host vehicle speed detected by the vehicle speed detection unit; a neutral point learning correction value setting unit that sets the steering angle difference calculated by the steering angle difference calculation unit as the neutral point learning correction value; A vehicle driving assistance device comprising:

2. A cross slope estimation unit that estimates a cross slope of a travel lane; an actual steering angle estimating unit that calculates an actual steering angle by subtracting a counter steering angle against the transverse gradient when the host vehicle travels on the driving lane having the transverse gradient from the steering angle detected by the steering angle detecting unit; Further comprising: The first estimated lateral acceleration calculation unit calculates the first estimated lateral acceleration based on the actual steering angle and the host vehicle speed.

2. The vehicle driving support device according to claim 1.

3. The first estimated lateral acceleration calculation unit sets a turning curvature of the host vehicle based on the actual steering angle and the host vehicle speed, and calculates the first estimated lateral acceleration by multiplying the turning curvature by the square of the host vehicle speed.

3. The vehicle driving support device according to claim 2.

4. A lateral acceleration detection unit detects a lateral acceleration of the host vehicle, The cross-slope estimation unit calculates the cross-slope based on a value obtained by multiplying the host vehicle speed by a yaw rate detected by a yaw rate detection unit and subtracting the lateral acceleration detected by the lateral acceleration detection unit from the product.

3. The vehicle driving support device according to claim 2.

5. The second estimated lateral acceleration calculation unit calculates a lane yaw rate by multiplying the vehicle speed by the lane curvature, calculates a lane yaw angle indicating a direction of the vehicle with respect to the lane obtained based on the driving environment recognized by the driving environment recognition unit, and calculates the second estimated lateral acceleration by multiplying the vehicle speed by a value obtained by adding the lane yaw rate and the lane yaw rate.

2. The vehicle driving support device according to claim 1.

Images