JP2016127638A - Control device of vehicle and control method of vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To control behaviour of a vehicle optimally by controlling a target control moment optimally on the basis of reliability of a roll angle of the vehicle.SOLUTION: A control device 200 comprises: a target yaw rate calculation portion 220 that acquires a target yaw rate; a subtraction portion 254 that acquires a first roll angle to be determined from a vehicle model and a second roll angle to be determined from a sensor; a feedback roll angle calculation portion 284 that calculates a feedback roll angle on the basis of a difference between the first roll angle and the second roll angle; a feedback yaw rate calculation portion 260 that acquires a feedback yaw rate, as a yaw rate generated by the vehicle, to be compared with the target yaw rate; and a target control moment calculation portion 290 that calculates a target control moment on the basis of a difference between the target yaw rate and the feedback yaw rate, and corrects the target control moment on the basis of the feedback roll angle.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a vehicle control device and a vehicle control method.

  Conventionally, a technique for controlling the roll angle of a vehicle is known. For example, Patent Document 1 below describes that the roll state of a vehicle is optimally controlled regardless of the running state by controlling the braking force of each wheel so that the roll of the vehicle has a target roll angle. ing. Patent Document 1 describes that the target yaw moment is controlled using feedforward and feedback control in order to set the roll angle to the target roll angle.

  In Patent Document 2 below, rudder capable of counter steer control, which is control for steering the rudder angle of the front wheel of a vehicle in a direction opposite to the turning direction when it is determined that there is a risk of rollover. A configuration is described in which an angle control unit and a driving force control unit capable of controlling acceleration of a vehicle when it is determined that there is a risk of rollover.

JP 2002-114140 A JP 2008-265560 A

  However, as described in Patent Document 1, when trying to control the roll state by brake braking, when changing lanes frequently or when continuously correcting rudder on a slippery road surface, There is a case where a reverse moment is generated by a brake with respect to the vehicle behavior. In this case, there is a problem that the operation timing of the master cylinder or the hydraulic valve is delayed, the intervention timing of the yaw moment control is shifted from the steering timing of the driver, and an unintended change in vehicle behavior or an uncomfortable feeling is caused.

  Further, in the technique described in Patent Document 2, since on / off control is performed according to a threshold value, it is assumed that the vehicle behavior changes suddenly when the mode is changed, and the driver feels uncomfortable. Further, since the control mode is not switched stepwise according to the transition state of the lateral acceleration, it is difficult to achieve both the turning performance and the stable performance.

  In addition, when trying to control the yaw moment by motor drive, response delay due to mechanical elements is reduced compared to braking, but when the difference between the control target yaw moment and the actual vehicle yaw moment is increased on a low μ road surface. Since the control is always performed in the direction in which the motor torque is output, the vehicle turns more than expected by the driver, causing a sense of discomfort. If the difference between the target yaw moment and the actual yaw moment is overshooting or hunting, the vehicle will vibrate more than the yaw moment control by brake braking, and the yaw moment fed back to the controller will also vibrate accordingly. And ride comfort will deteriorate.

  Therefore, the present invention has been made in view of the above problems, and an object of the present invention is to optimally control the control target moment based on the reliability of the roll angle of the vehicle, thereby It is an object of the present invention to provide a new and improved vehicle control apparatus and vehicle control method capable of optimally controlling and suppressing the occurrence of a sense of incongruity and a decrease in riding comfort.

  In order to solve the above problems, according to an aspect of the present invention, a target yaw rate acquisition unit that acquires a target yaw rate of a vehicle, a first roll angle obtained from a vehicle model, and a second roll angle obtained from a sensor are obtained. When the difference is small based on the difference between the roll rate acquisition unit that performs and the first roll angle and the second roll angle, the distribution of the first roll angle is increased, and the difference is large As a yaw rate generated by the vehicle, a feedback roll angle calculation unit that calculates a feedback roll angle from the first roll angle and the second roll angle by increasing the distribution of the second roll angle, A feedback yaw rate acquisition unit for acquiring a feedback yaw rate for comparison with the target yaw rate, the target yaw rate and the feedback yaw rate; It calculates a control target moment based on the difference between the rate control apparatus for a vehicle and a control target moment calculating unit that corrects the control target moment based on the feedback roll angle is provided.

  The control target moment calculation unit calculates the control target moment by adding a steady damping control moment and a transient inertia compensation moment, and calculates the damping control moment and the inertia compensation moment based on the feedback roll angle. It may be corrected.

  The control target moment calculator corrects the damping control moment and the inertia compensation moment so that the damping control moment or the inertia compensation moment decreases as the absolute value of the feedback roll angle increases. There may be.

  Further, the control target moment calculation unit sequentially changes the control target moment into a moment that supports turning, a moment that does not support turning, and a moment in a direction opposite to turning as the absolute value of the feedback roll angle increases. Further, the damping control moment and the inertia compensation moment may be corrected.

  Further, a rollover adjustment gain calculating unit that calculates a gain multiplied by each of the damping control moment and the inertia compensation moment based on the feedback roll angle may be provided.

  Further, it is determined whether or not a steering operation is turned back based on a steering angle and a yaw rate, and the control target moment calculation unit determines that the feedback roll is turned on when it is determined that the steering operation is turned back. The control target moment may be corrected based on the angle.

  Further, a yaw rate acquisition unit that acquires a first yaw rate obtained from a vehicle model and a second yaw rate obtained from a yaw rate sensor is provided, and the feedback yaw rate acquisition unit calculates a difference between the first yaw rate and the second yaw rate. Based on this, when the difference is small, the distribution of the first yaw rate is increased, and when the difference is large, the distribution of the second yaw rate is increased, and the feedback yaw rate is changed to the first and second yaw rates. It may be calculated from

  In order to solve the above problem, according to another aspect of the present invention, a step of obtaining a target yaw rate of a vehicle, a first roll angle obtained from a vehicle model, and a second roll angle obtained from a sensor are obtained. And, based on the difference between the first roll angle and the second roll angle, if the difference is small, increase the distribution of the first roll angle, and if the difference is large, In order to compare the target yaw rate with the step of calculating the feedback roll angle from the first roll angle and the second roll angle by increasing the distribution of the roll angle of 2, and the yaw rate generated by the vehicle Obtaining a feedback yaw rate, and calculating a control target moment based on a difference between the target yaw rate and the feedback yaw rate , The vehicle control method and a step of correcting the control target moment based on the feedback roll angle is provided.

  As described above, according to the present invention, when the vehicle is controlled by feeding back the deviation between the target yaw rate of the vehicle and the actual yaw rate actually generated by the vehicle, the influence of sensor noise is reliably suppressed. It becomes possible to improve drivability.

It is a mimetic diagram showing a vehicle concerning one embodiment of the present invention. It is a schematic diagram which shows turning control by steering. It is a schematic diagram which shows the structure of a control apparatus. It is a schematic diagram for demonstrating the process performed with a control apparatus. It is a schematic diagram for demonstrating the process performed in a rollover determination part. It is a schematic diagram which shows the gain map at the time of a feedback yaw rate calculating part calculating weighting gain (kappa). It is a schematic diagram which shows the gain map at the time of a roll model reliability determination part calculating weighting gain (tau). It is a flowchart which shows the process which calculates weighting gain (tau). It is a flowchart which shows the process which sets sgn_TurnSt. It is a schematic diagram which shows the map which calculates the steady gain for rollover correction | amendment. It is a flowchart which shows the process for calculating the steady gain for rollover correction | amendment. 12 is a flowchart illustrating a process of calculating a steady gain for rollover correction in step S52 of FIG. It is a flowchart which shows the process which calculates the steady gain ConstGainForRollOver for rollover correction | amendment when (phi) _ref> = 0. It is a flowchart which shows the process which calculates the steady gain ConstGainForRollOver for rollover correction | amendment when (phi) _ref <0. It is a schematic diagram which shows the map which calculates the transient gain for rollover correction | amendment. It is a flowchart which shows the process for calculating the transient gain for rollover correction | amendment. It is a flowchart which shows the process which calculates the transient gain for rollover correction | amendment in step S112 of FIG. It is a flowchart which shows the process which calculates the transient gain TransGainForRollOver for rollover correction | amendment when (phi) _ref> = 0. It is a flowchart which shows the process which calculates the transient gain TransGainForRollOver for rollover correction | amendment when (phi) _ref <0. It is a characteristic view for demonstrating the effect by the control which concerns on this embodiment. It is a characteristic view for demonstrating the effect by the control which concerns on this embodiment.

  Exemplary embodiments of the present invention will be described below in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

  First, with reference to FIG. 1, the structure of the vehicle 1000 which concerns on one Embodiment of this invention is demonstrated. FIG. 1 is a schematic diagram showing a vehicle 1000 according to the present embodiment. As shown in FIG. 1, a vehicle 1000 includes driving force generators (motors) 108, 110, 112, driving front wheels 100 and 102, rear wheels 104 and 106, front wheels 100 and 102, and rear wheels 104 and 106, respectively. 114, wheel speed sensors 116, 118, 120, 122 for detecting the respective wheel speeds of the front wheels 100, 102 and the rear wheels 104, 106, a steering wheel 124, a steering angle sensor 130, a power steering mechanism 140, a yaw rate sensor 150, a lateral An acceleration sensor 160 and a control device (controller) 200 are included.

  The vehicle 1000 according to the present embodiment is provided with motors 108, 110, 112, and 114 for driving the front wheels 100 and 102 and the rear wheels 104 and 106, respectively. Therefore, the driving torque can be controlled by each of the front wheels 100 and 102 and the rear wheels 104 and 106. Accordingly, the yaw rate can be generated by the torque vectoring control by driving each of the front wheels 100 and 102 and the rear wheels 104 and 106 independently of the yaw rate generation by the steering of the front wheels 100 and 102. In particular, in this embodiment, the yaw rate is generated independently of the steering system by controlling the torque of the rear wheels 104 and 106 individually. The driving torque of the rear wheels 104 and 106 is controlled by controlling the motors 112 and 114 corresponding to the rear wheels 104 and 106 based on a command from the control device 200.

  The power steering mechanism 140 controls the steering angle of the front wheels 100 and 102 by torque control or angle control according to the operation of the steering wheel 124 by the driver. The steering angle sensor 130 detects the steering angle θh input by the driver operating the steering wheel 124. The yaw rate sensor 150 detects the actual yaw rate γ of the vehicle 1000. Wheel speed sensors 116, 118, 120 and 122 detect vehicle speed V of vehicle 1000.

  Note that the present embodiment is not limited to this embodiment, and the motors 108 and 102 for driving the front wheels 100 and 102 are not provided, and only the rear wheels 104 and 106 are independently driven by the motors 112 and 114. The generated vehicle may be used. Further, the present embodiment is not limited to torque vectoring based on driving force control, and can also be realized in a 4WS system for controlling the steering angle of the rear wheels.

  FIG. 2 is a schematic diagram showing turning control (steering control) by steering performed by the vehicle 1000 according to the present embodiment. In the turning control by the steering, the turning of the vehicle 1000 is supported by generating a driving force difference between the rear wheels 104 and 106 according to the operation of the steering wheel 130 by the driver. In the example shown in FIG. 2, the vehicle 1000 is turning to the left by the steering of the driver (driver). Further, due to the difference in driving force between the rear wheels 104 and 106, a forward driving force is generated on the right rear wheel 106, and the driving force on the left rear wheel 104 is suppressed with respect to the right rear wheel 106 or backwards. By generating a driving force, a driving force difference is generated on the left and right, and a moment is generated in a direction that supports a counterclockwise turn.

  In the present embodiment, in vehicle motion control, the vehicle turning characteristics are estimated from the difference between the yaw rate model value obtained from the vehicle model and the actual yaw rate obtained from the yaw rate sensor. When the difference is small, the tire grip performance is sufficiently ensured. The control target yaw rate corresponding to the situation is calculated.

  When the difference between the yaw rate model value and the actual yaw rate is large, it is determined that the tire grip performance is difficult to secure and the vehicle response limit is likely to be reached (low μ), and the steering and yaw rate values are matched by a sign function. Thus, the steering direction and the direction in which yaw is generated are discriminated. If the product of the respective encoded values becomes negative, it is determined that the turn-back steering is performed, and the turning support mode, the left / right equal torque mode, the reverse moment applying mode are determined according to the amount of leaning of the vehicle body. Are switched in stages to achieve both vehicle turning performance and rollover suppression. Details will be described below.

  FIG. 3 is a schematic diagram illustrating the configuration of the control device 200. The control device 200 includes an in-vehicle sensor 210, a control target yaw rate calculation unit 220, a vehicle model 230, a filter processing unit 240, subtraction units 250, 252, and 254, a feedback yaw rate calculation unit 260, a roll model 270, a rollover determination unit 280, and a control. A target moment calculation unit 290 and a motor required torque calculation unit 295 are provided.

  The rollover determination unit 280 includes a roll model reliability determination unit 282, a feedback roll angle calculation unit 284, subtraction units 286 and 287, sign functions 288 and 289, and a rollover adjustment gain calculation unit 285.

  FIG. 4 is a schematic diagram for explaining processing performed in the control device 200. FIG. 5 is a schematic diagram for explaining processing performed by the rollover determination unit 280. 4 and 5 show the components of the control device 200 in the same manner as in FIG. 3, and also show in detail the processing performed by each component. Below, the process performed with the control apparatus 200 is demonstrated based on FIGS. 3-5. The in-vehicle sensor 210 includes the steering angle sensor 130, the yaw rate sensor 150, the acceleration sensor 160, and the wheel speed sensors 116, 118, 120, and 122 described above.

  The control target yaw rate calculation unit 220 calculates the steering control target yaw rate γTgt from the following expression (1) representing a general two-wheel model. The steering control target yaw rate γTgt corresponds to the vehicle yaw rate γ in the equation (1) of the plane two-wheel model, and is calculated by substituting each value for the right side of the equation (1).

In the equations (1) to (3), variables and constants are as follows.
<Variable>
γ: vehicle yaw rate V: vehicle speed δ: tire steering angle θh: steering wheel steering angle <constant>
l: vehicle wheel base l _f : distance from vehicle center of gravity to center of front wheel l _r : distance from vehicle center of gravity to center of rear wheel m: vehicle weight k _f : cornering power (front)
k _r : Cornering power (rear)
Gh: Conversion gain from steering wheel steering angle to tire steering angle (steering gear ratio)

  The control target yaw rate γTgt (γ on the left side of the equation (1)) is calculated from the equation (1) using the vehicle speed V and the tire steering angle δ as variables. Since the tire steering angle δ in equation (1) cannot be directly sensed, it is calculated from equation (2) by multiplying the steering angle θh by the conversion gain Gh. A steering gear ratio is used as the conversion gain Gh. In addition, the constant A in the equation (1) is a constant representing the characteristics of the vehicle, and is obtained from the equation (3). The control target yaw rate γTgt is input to the subtraction unit 250.

  In calculating the control target yaw rate γTgt, the tire steering angle δ may be calculated from a steering model.

  On the other hand, the vehicle model 230 and the filter processing unit 240 obtain the yaw rate generated by the vehicle 1000 by calculation or actual measurement. The vehicle model 230 calculates the yaw rate model value γ_clc from the following vehicle model equations (4) to (6) in which the side slip angle, yaw, and roll of the vehicle 1000 are coupled. In addition, the vehicle model 230 calculates the first roll angle φ1_clc as the lean amount of the vehicle body from the following vehicle model equation. The yaw rate model value γ_clc is input to the feedback yaw rate calculation unit 260. Further, the first roll angle φ1_clc is input to the subtraction unit 254 and the feedback roll angle calculation unit 284.

  In the above model formula, γ corresponds to γ_clc of the present embodiment, and φ corresponds to the first roll angle φ1_clc of the present embodiment.

I is the yaw inertia of the vehicle 1000, β is the side slip angle of the vehicle, K _cf is the canvas last coefficient of the front wheel, K _cr is the canvas last coefficient of the rear wheel, h is the roll arm length (vehicle center of gravity height−roll center height), I _XZ is the product of inertia, g is the acceleration of gravity, I _φ is the moment of inertia for the rolling motion, C _φ is the damping coefficient for the rolling motion, K _φ is the roll stiffness, ∂φ _f / ∂φ is the camber angle of the front wheel, ∂φ _r / ∂φ is the camber angle of the rear wheel, ∂α _f / ∂φ is the roll steer amount of the front wheel per unit roll angle, and ∂α _r / ∂φ is the roll steer amount of the rear wheel per unit roll angle. .

  The filter processing unit 240 performs filter processing to remove noise from the actual yaw rate γ detected by the yaw rate sensor 150, and outputs the yaw rate γ_fil obtained as a result of the filter processing to the feedback yaw rate calculation unit 260.

  The subtractor 252 subtracts the yaw rate γ_fil from the yaw rate model value γ_clc to obtain a difference γ_diff between the yaw rate model value γ_clc and the yaw rate γ_fil. The difference γ_diff is input to the feedback yaw rate calculation unit 260.

As described above, the yaw rate model value γ_clc and the yaw rate γ_fil are input to the feedback yaw rate calculation unit 260. The feedback yaw rate calculation unit 260 calculates a feedback yaw rate γF / B fed back from the vehicle 1000 as a comparison target with the control target yaw rate γTgt. At this time, the feedback yaw rate calculation unit 260 calculates the weighting gain κ based on the difference γ_diff between the yaw rate model value γ_clc and the actual yaw rate γ_fil. Then, the feedback yaw rate calculation unit 260 weights the yaw rate model value γ_clc and the yaw rate γ_fil with the weighting gain κ based on the following equation (7) to calculate the feedback yaw rate γF / B. The calculated feedback yaw rate γF / B is input to the subtraction unit 250.
γF / B = κ × γ_clc + (1−κ) × γ_fil (7)

  FIG. 6 is a schematic diagram illustrating a gain map when the feedback yaw rate calculation unit 232 calculates the weighting gain κ. As shown in FIG. 6, the value of the weighting gain κ varies between 0 and 1 according to the reliability of the vehicle model 230. A difference (deviation) γ_diff between the yaw rate model value γ_clc and the yaw rate γ_fil is used as an index for improving the reliability of the vehicle model 230. As shown in FIG. 6, the gain map is set so that the value of the weighting gain κ increases as the absolute value of the difference γ_diff decreases. The feedback yaw rate calculation unit 260 performs the map process of FIG. 6 on the difference γ_diff and calculates a weighting gain κ according to the reliability of the vehicle model 230.

  In FIG. 6, TH1_P is a threshold value (+ side) for switching the weighting gain κ, TH2_P is a switching threshold value (+ side) for the weighting gain κ, TH1_M is a switching threshold value (−side) for the weighting gain κ, TH2_M represents a switching threshold value (− side) of the weighting gain κ. The threshold value relationship on the + side is TH1_P <TH2_P, and the threshold value relationship on the − side is TH1_M> TH2_M.

  A region A1 of the gain map shown in FIG. 6 is a region where the difference γ_diff approaches 0, a region where the S / N ratio is small, and a region where the tire characteristics are linear (dry road surface), and is calculated from the vehicle model 216. The yaw rate model value γ_clc is highly reliable. Accordingly, the feedback yaw rate γF / B is calculated with the weighting gain κ = 1 and the distribution of the yaw rate model value γ_clc as 100% from the equation (7). Thereby, the influence of the noise of the yaw rate sensor 150 included in the yaw rate γ_fil can be suppressed, and the sensor noise can be excluded from the feedback yaw rate γF / B. Therefore, it is possible to suppress the vibration of the vehicle 1000 and improve the riding comfort.

  In particular, in the driving support control, the amount by which the vehicle 1000 turns based on the estimated travel path from the straight traveling state before the vehicle 1000 enters the corner is foresight controlled. Therefore, not only when the vehicle 1000 is turning but also when the vehicle 1000 is traveling straight, it is possible to move the vehicle 1000 straight without causing vibrations by eliminating the influence of sensor noise.

  As described above, the region where the reliability of the yaw rate model value γ_clc is high can be designated from the difference γ_diff and the traveling state. As shown in FIG. 6, when the vehicle is traveling on a dry road surface (high μ) and the turning amount is small (such as turning at a low curvature), the difference γ_diff is weighted so that the weighting gain κ becomes 1. Associating the gain κ is assumed as an example of coefficient setting by a map. The planar two-wheel model described above assumes a region where the relationship between tire slip angle and lateral acceleration (tire cornering characteristics) is linear. In the region where the cornering characteristics of the tire are nonlinear, the yaw rate and lateral acceleration of the actual vehicle become nonlinear with respect to the steering angle and slip angle, and the two-wheel model and the yaw rate sensed by the actual vehicle deviate. For this reason, if a model that takes into account the nonlinearity of the tire is used, control based on the yaw rate becomes complicated. However, according to the present embodiment, the reliability of the yaw rate model value γ_clc is easily determined based on the difference γ_diff. Is possible.

  Further, the gain map area A2 shown in FIG. 6 is an area where the difference γ_diff is large, which corresponds to a wet road surface, a snow road, or a turn with a high G, and the limit of tire slipping. It is an area. In this region, the reliability of the yaw rate model value γ_clc calculated from the vehicle model 216 becomes low, and the difference γ_diff becomes larger. Therefore, the feedback yaw rate γF / B is calculated with the weighting gain κ = 0 and the distribution of the yaw rate γ_fil as 100% according to the equation (7). Thus, feedback accuracy is ensured based on the yaw rate γ_fil, and the yaw rate feedback control reflecting the behavior of the actual vehicle is performed. Therefore, the turning of the vehicle 1000 can be optimally controlled based on the yaw rate γ_fil. In addition, since the tire is slipping, even if the signal of the yaw rate sensor 150 is affected by noise, the driver does not feel it as vibration of the vehicle 1000, and a decrease in riding comfort can be suppressed. For the setting of the low μ region A2 shown in FIG. 6, the region where the weighting gain κ = 0 may be determined from the design requirements, and the steering stability performance and the ride when the vehicle 1000 actually travels on the low μ road surface. You may decide experimentally from comfort etc.

  Further, the gain map area A3 shown in FIG. 6 is an area (nonlinear area) in which a transition from a linear area to a limit area is performed, and the yaw rate model value γ_clc is considered in consideration of tire characteristics of the actual vehicle 1000 as necessary. And the distribution of the yaw rate γ_fil (weighting gain κ) are linearly changed. In the transition from the region A1 (high μ region) to the region A2 (low μ region) or in the region transitioning from the region A2 (low μ region) to the region A1 (high μ region), the torque associated with the sudden change of the weighting gain κ In order to suppress fluctuations and fluctuations in the yaw rate, the weighting gain κ is calculated by linear interpolation.

  A gain map region A4 shown in FIG. 6 corresponds to a case where the yaw rate γ_fil is larger than the yaw rate model value γ_clc. For example, when an incorrect parameter is input to the vehicle model 216 and the yaw rate model value γ_clc is erroneously calculated, control can be performed using the yaw rate γ_fil based on the map of the region A4. Further, according to the map of the region A4, even when the yaw rate model value γ_clc is temporarily smaller than the yaw rate γ_fil due to the phase delay of the yaw rate γ_fil accompanying the filter processing, the control using the yaw rate γ_fil is performed. It can be carried out. It should be noted that the range of the weighting gain κ is not limited to 0 to 1, but the configuration of the present invention can also be changed so that an arbitrary value can be taken as long as the range is established as vehicle control. Enter into a possible category.

The subtraction unit 250 receives the control target yaw rate γTgt from the control target yaw rate calculation unit 230, and receives the feedback yaw rate γF / B from the feedback yaw rate calculation unit 232. The subtraction unit 250 subtracts the feedback yaw rate γF / B from the control target yaw rate γTgt to obtain a difference ΔγTgt between γTgt and γF / B. That is, the difference ΔγTgt is calculated from the following equation (8).
ΔγTgt = γTgt−γF / B (8)
The difference ΔγTgt is output to the control target moment calculator 290 as the yaw rate correction amount.

  The control target moment calculator 290 calculates the control target moment MgTgt based on the difference ΔγTgt in order to set the difference ΔγTgt to 0, that is, to make the control target yaw rate γTgt and the feedback yaw rate γF / B coincide. At this time, the control target moment calculator 290 corrects the control target moment MgTgt with the rollover adjustment gain.

  As described above, the first roll angle φ1_clc is estimated from the steering angle θh and the vehicle speed V based on the equations (4) to (6). In addition, the lateral acceleration Gy detected by the lateral acceleration sensor 160 is input to the roll model 270. The roll model 270 calculates the second roll angle φ2_clc as the vehicle body tilt amount based on the lateral acceleration Gy. The second roll angle φ2_clc is calculated from the following equation (9).

  The rollover determination unit 280 performs rollover risk determination based on the first roll angle φ1_clc and the second roll angle φ2_clc. Specifically, the rollover determination unit 280 is calculated based on the first roll angle φ1_clc calculated from the vehicle model 230, the steering angle θh, and the vehicle speed V, and the second G calculated from the lateral acceleration sensor 160. A roll angle deviation (φ_diff) is obtained from the roll angle φ2_clc, and the “roll model reliability” is determined based on the deviation (φ_diff). Therefore, the subtraction unit 254 calculates a difference φ_diff between the first roll angle φ1_clc calculated from the vehicle model 230 and the second roll angle φ2_clc calculated from the detection value Gy of the lateral acceleration sensor. The roll model reliability determination unit 282 of the rollover determination unit 280 uses the difference φ_diff as an index for determining the reliability of the roll model 270, and calculates the weighting gain τ according to the model reliability, thereby determining the behavior of the actual vehicle. The influence of noise on the detection value of the lateral acceleration sensor 160 to be detected is eliminated. The determination of the reliability of the roll model 270 is performed in the same manner as the determination of the reliability of the vehicle model 230 described above.

  As described above, the difference between the first roll angle φ1_clc calculated from the roll model 270 and the second target roll angle calculated from the detection value Gy of the lateral acceleration sensor is used as an index for determining the reliability of the roll model 270, and the feedback yaw rate Based on the same procedure as the determination of the vehicle model reliability described in the calculation, the weighting gain (τ) for determining the model reliability is also calculated for the roll. Thereby, the influence of the noise by the sensor which detects the behavior of the actual vehicle such as lateral acceleration can be eliminated.

  Note that the second roll angle φ2_clc is calculated using a suspension stroke amount of the vehicle 1000, a detection value of the vertical G sensor, a parameter obtained by integrating a direct measurement of the roll rate, or a value directly detected by the tilt angle sensor. You may do it.

  FIG. 7 is a schematic diagram illustrating a gain map when the roll model reliability determination unit 282 calculates the weighting gain τ. As shown in FIG. 7, the value of the weighting gain τ varies between 0 and 1 according to the reliability of the roll model 270. As an index for improving the reliability of the roll model 270, a difference φ_diff between the first roll angle φ1_clc and the second roll angle φ2_clc is used. As shown in FIG. 7, the gain map is set so that the value of the weighting gain τ increases as the absolute value of the difference φ_diff decreases. The roll model reliability determination unit 282 performs the map process of FIG. 7 on the difference φ_diff and calculates a weighting gain τ according to the reliability of the roll model 270.

  In FIG. 7, TH1_P is a switching threshold value (+ side) for weighting gain τ, TH2_P is a switching threshold value (+ side) for weighting gain τ, TH1_M is a switching threshold value (−side) for weighting gain τ, TH2_M represents a switching threshold value (− side) of the weighting gain τ. The threshold value relationship on the + side is TH1_P <TH2_P, and the threshold value relationship on the − side is TH1_M> TH2_M.

  FIG. 8 is a flowchart showing a process for calculating the weighting gain τ. First, in step S10, it is determined whether or not φ_diff ≧ 0. If φ_diff ≧ 0, the process proceeds to step S12 to determine whether or not φ_diff ≦ TH1_P. If φ_diff ≦ TH1_P, the process proceeds to step S14 to set τ = MAX_GAIN. On the other hand, if φ_diff> TH1_P in step S12, the process proceeds to step S16, and it is determined whether φ_diff ≧ TH2_P. If φ_diff ≧ TH2_P, the process proceeds to step S18, and τ = MIN_GAIN. If φ_diff <TH2_P in step S16, the process proceeds to step S20, and the weighting gain τ is calculated from the following equation (10).

  If φ_diff <0 in step S10, the process proceeds to step S22, and it is determined whether φ_diff ≧ TH1_M. If φ_diff ≧ TH1_M, the process proceeds to step S24, where τ = MAX_GAIN. On the other hand, if φ_diff <TH1_M in step S22, the process proceeds to step S26, and it is determined whether φ_diff ≦ TH2_M. If φ_diff ≦ TH2_M, the process proceeds to step S28, and τ = MIN_GAIN. If φ_diff> TH2_M in step S26, the process proceeds to step S30, and the weighting gain τ is calculated from the following equation (11).

  As described above, according to the process of FIG. 8, the weighting gain τ can be calculated according to the map of FIG.

Then, the feedback roll angle calculation unit 284 weights the first roll angle φ1_clc and the second roll angle φ2_clc with the weighting gain τ based on the following equation (12), and calculates the reference value φ_ref of the roll angle used in the control. . The calculated φ_ref is input to the rollover adjustment gain calculation unit 285.
φ_ref = τ × φ1_clc + (1−τ) × φ2_clc (12)

  Thus, the steering roll angle θh, the vehicle speed V, the first roll angle φ1_clc estimated from the vehicle model, and the lateral acceleration and the second roll angle φ2_clc estimated from the roll model are apportioned by the weighting gain τ representing the model reliability. The roll angle reference value φ_ref is calculated.

  Therefore, the rollover risk can be determined based on the model value in the high μ region. On the other hand, in areas where the model value and the behavior of the actual vehicle deviate, such as in the low μ region and the region where the lateral G is high, the side force of the tire is saturated in a state smaller than the high μ region, and skidding occurs more than the vehicle rolls. Taking into account the phenomenon, it is possible to prevent erroneous determination of risk by determining using the roll angle calculated from the actual vehicle behavior (lateral G).

  On the other hand, the tire is steered by the steering torque applied to the vehicle 1000 in accordance with the steering operation of the driver, and a phase delay is caused by interposing the steering mechanism, the tire, and the like until the yaw rate is applied to the vehicle. The rollover determination unit 280 uses this mechanism to compare the steering direction with the yaw rate generation direction and estimate the timing of the reverse steering.

  In addition, the rollover determination unit 280 makes a stepwise transition to the turning support mode, the left / right equal torque mode, and the reverse moment application mode according to the lean amount (φ_ref) of the vehicle body during the turn-back steering, and φ_ref exceeds a predetermined threshold value. If it is determined that the vehicle is in an unstable state, the target damping moment is determined to reduce the difference between the left and right driving forces or to generate a moment that acts on the opposite side of the direction of turning by steering. The gain DampGainForRollOver is corrected. As a result, the vehicle behavior can be stabilized especially with respect to the steady behavior of the vehicle 1000.

  The rollover determination unit 280 also calculates a gain TransGainForRollOver that corrects the target inertia compensation moment. If it is determined that φ_ref exceeds a predetermined threshold value, the output of the target inertia compensation moment is stepwise. Restrict. Thereby, the vehicle behavior can be stabilized particularly with respect to the instantaneous behavior of the vehicle 1000.

  Then, a target damping moment MgDampTgt and a target inertia compensation moment MgTransTgt are calculated from the yaw rate correction amount ΔγTgt, and a sum of the respective target moments multiplied by a correction gain is added to the vehicle according to the magnitude of φ_ref. The control target moment MgTgt for achieving both the turning performance and the stable performance is calculated.

  For this reason, the yaw rate γ_fil and the steering angle θh are input to the rollover determination unit 280. The subtractor 286 calculates Δθt by performing a dead zone process on the difference between the current value and the previous value of the steering angle θH. In addition, the subtracting unit 287 performs a dead zone process on the difference between the current value and the previous value of the yaw rate γ_fil to calculate Δγt. The encoding processing unit 288 converts Δθt into a value of 1, 0, or −1 by a sign function (sgn), and calculates sgn (Δθt). Also, the encoding processing unit 288 converts Δγt into a value of 1, 0, or −1 by a code function (sgn), and calculates sgn (Δγt).

  That is, sgn (Δθt) and sgn (Δγt) can be calculated from the following equations (13) and (14).

  If the values of sgn (Δθt) and sgn (Δγt) are different from each other, it is determined that the yaw direction is starting to change at the same time as the steering starts to turn, and the switch for judging the turning back (Sgn_TurnSt) is turned on (ON). To. If the values of sgn (Δθt) and sgn (Δγt) are the same, it is determined that the steering and yaw directions coincide with each other, and the switch for determination of turning-off is turned off. As described above, the timing of the steering return can be determined based on the values of sgn (Δθt) and sgn (Δγt).

  In the present embodiment, the difference in the steering operation amount and the difference in the yaw rate are used. However, the same function can be realized by encoding with the steering angular velocity or the yaw acceleration.

  The rollover adjustment gain calculation unit 285 calculates two types of gains based on the value of φ_ref. First, calculation of steady gain for rollover correction will be described. The turning support mode (Gain> 0) according to the timing when the product of the sign function (sgn_TurnSt = sgn (Δθt) × sgn (Δγt)) becomes negative (steering return occurs) and φ_ref exceeds a predetermined threshold. ), The right and left equal torque mode (Gain = 0), and the reverse moment applying mode (Gain <0), the value of “steady gain for rollover correction” is changed in a stepwise manner.

  On the other hand, according to the timing when the product of the sign function (sgn_TurnSt = sgn (Δθt) × sgn (Δγt)) is switched from negative to a value of 0 or more, that is, according to the timing at which the rudder angle and the yaw rate direction match. In order to prevent a sudden change in vehicle behavior due to a sudden change in mode, an average value of “a steady gain for rollover correction” is taken between three samplings, and the gain is gradually changed.

  FIG. 9 is a flowchart showing a process for setting sgn_TurnSt. First, in step S40, it is determined whether or not sgn (Δθt) ≠ sgn (Δγt). If sgn (Δθt) ≠ sgn (Δγt), the process proceeds to step S42, and sgn_TurnSt is turned on. Thereby, sgn_TurnSt is set to 1. On the other hand, if sgn (Δθt) = sgn (Δγt), the process proceeds to step S44, and sgn_TurnSt is turned off. Thereby, sgn_TurnSt is set to 0.

  FIG. 10 is a schematic diagram showing a map for calculating a steady gain for rollover correction. FIG. 11 is a flowchart showing a process for calculating a steady gain for rollover correction. First, in step S50, it is determined whether or not sgn_TurnSt = ON. If sgn_TurnSt = ON, the process proceeds to step S52. In step S52, a steady gain for rollover correction is calculated.

  FIG. 12 is a flowchart showing a process for calculating a steady gain for rollover correction in step S52 of FIG. In the process of FIG. 12, first, in step S60, it is determined whether or not the difference φ_ref ≧ 0. If the difference φ_ref ≧ 0, the process proceeds to step S62, and rollover is performed based on the value of φ_ref that is 0 or more. The correction steady-state gain ConstGainForRollOver is calculated. If the difference φ_ref <0, the process proceeds to step S64, and the rollover correction steady-state gain ConstGainForRollOver is calculated based on the value of φ_ref which is a value smaller than 0.

  FIG. 13 is a flowchart illustrating a process of calculating the rollover correction steady-state gain ConstGainForRollOver when φ_ref ≧ 0. In the case of φ_ref ≧ 0, the rollover correction steady-state gain ConstGainForRollOver is calculated according to the characteristics of the region of φ_ref ≧ 0 in the map shown in FIG. First, in step S70, it is determined whether or not φ_ref ≦ TH1_P. If φ_ref ≦ TH1_P, the process proceeds to step S72 to set ConstGainForRollOver = MAX_GAIN. If φ_ref> TH1_P in step S70, the process proceeds to step S74 to determine whether or not φ_ref ≦ TH2_P. If φ_ref ≦ TH2_P, the rollover correction steady-state gain ConstGainForRollOver is calculated from the following equation (15). To do.

  If φ_ref> TH2_P in step S74, the process proceeds to step S78 to determine whether or not φ_ref ≦ TH3_P. If φ_ref ≦ TH3_P, the process proceeds to step S80, and ConstGainForRollOver = MID_GAIN. If φ_ref> TH3_P in step S78, the process proceeds to step S82 to determine whether or not φ_ref ≦ TH4_P. If φ_ref ≦ TH4_P, the process proceeds to step S84, and rollover correction is performed from the following equation (16). The steady gain ConstGainForRollOver is calculated.

  If φ_ref> TH4_P in step S82, the process proceeds to step S86, where ConstGainForRollOver = MIN_GAIN.

  FIG. 14 is a flowchart showing a process of calculating the rollover correction steady-state gain ConstGainForRollOver when φ_ref <0. When φ_ref <0, the rollover correction steady-state gain ConstGainForRollOver is calculated according to the characteristics of the region of φ_ref <0 in the map shown in FIG. First, in step S90, it is determined whether or not φ_ref ≧ TH1_M. If φ_ref ≧ TH1_M, the process proceeds to step S92 to set ConstGainForRollOver = MAX_GAIN. If φ_ref <TH1_M in step S90, the process proceeds to step S94, and it is determined whether φ_ref ≧ TH2_M. If φ_ref ≧ TH2_M, the rollover correction steady-state gain ConstGainForRollOver is calculated from the following equation (17). .

  If φ_ref <TH2_M in step S94, the process proceeds to step S98 to determine whether or not φ_ref ≧ TH3_M. If φ_ref ≧ TH3_M, the process proceeds to step S100, and ConstGainForRollOver = MID_GAIN. If φ_ref <TH3_M in step S98, the process proceeds to step S102, and it is determined whether or not φ_ref ≧ TH4_M. If φ_ref ≧ TH4_M, the process proceeds to step S104, and rollover correction is performed from the following equation (18). The steady gain ConstGainForRollOver is calculated.

  If φ_ref <TH4_M in step S102, the process proceeds to step S106, where ConstGainForRollOver = MIN_GAIN.

  As described above, according to the processing of FIGS. 13 and 14, when sgn_TurnSt = ON, the rollover correction steady-state gain ConstGainForRollOver can be calculated based on the map of FIG.

  If sgn_TurnSt = OFF in step S50 of FIG. 11, the process proceeds to step S54. In step S54, it is determined whether or not the same situation has continued three or more times. If the same situation has continued three or more times, the process proceeds to step S56. In step S56, ConstGainForRollOver = ConstGainForRollOver (n), and the control returns to the normal control mode (OUT). On the other hand, if it is determined in step S54 that the same situation has not continued three times or more, the process proceeds to step S58. In step S18, the rollover correction steady-state gain ConstGainForRollOver is calculated from the following equation (19). As a result, the rollover correction steady-state gain ConstGainForRollOver gradually changes.

  In the present embodiment, when changing from the steering return state to the normal turning mode, the correction gain is gradually changed by taking the average value of the three sampling data, but the average value of the two sampling data (two parameters). Alternatively, the correction gain may be gradually changed by calculating an average value of data of 4 samplings (4 parameters) or more.

Next, calculation of the transient gain for rollover correction will be described. The product of the sign function (sgn_TurnSt = sgn (Δθt) × sgn (Δγt)) becomes negative (steering return occurs), and the output of the target inertia compensation moment is output in accordance with the timing when φ_ref exceeds a predetermined threshold. Change the value of “Transient gain for rollover correction” to limit in stages.

  On the other hand, at the timing when the product of the sign function (sgn_TurnSt = sgn (Δθt) × sgn (Δγt)) is switched from negative to 0 or more (the steering angle and the yaw rate are the same), a sudden change of the correction gain is prevented. During sampling, the average value of “transient gain for rollover correction” is taken, and the gain is gradually changed.

  FIG. 15 is a schematic diagram showing a map for calculating a transient gain for rollover correction. FIG. 16 is a flowchart showing a process for calculating a transient gain for rollover correction. First, in step S110, it is determined whether or not sgn_TurnSt = 0, and if sgn_TurnSt ≧ 0, the process proceeds to step S112. In step S112, a transient gain for rollover correction is calculated.

  FIG. 17 is a flowchart showing the process for calculating the transient gain for rollover correction in step S112 in FIG. In the process of FIG. 13, first, in step S120, it is determined whether or not the difference φ_ref ≧ 0. If the difference φ_ref ≧ 0, the process proceeds to step S122, and rollover is performed based on the value of φ_ref that is 0 or more. The correction transient gain TransGainForRollOver is calculated. If the difference φ_ref <0, the process proceeds to step S124, and the rollover correction transient gain TransGainForRollOver is calculated based on the value of φ_ref which is a value smaller than zero.

  FIG. 18 is a flowchart showing a process for calculating the rollover correction transient gain TransGainForRollOver when φ_ref ≧ 0. When φ_ref ≧ 0, the rollover correction transient gain TransGainForRollOver is calculated according to the characteristics of the region of φ_ref ≧ 0 in the map shown in FIG. First, in step S130, it is determined whether or not φ_ref ≦ TH1_P. If φ_ref ≦ TH1_P, the process proceeds to step S132 to set TransGainForRollOver = MAX_GAIN. If φ_ref> TH1_P in step S130, the process proceeds to step S134 to determine whether or not φ_ref ≦ TH2_P. If φ_ref ≦ TH2_P, the rollover correction transient gain TransGainForRollOver is calculated from the following equation (20). To do.

  If φ_ref> TH2_P in step S134, the process proceeds to step S138, where TransGainForRollOver = MIN_GAIN.

  FIG. 19 is a flowchart showing a process for calculating the rollover correction transient gain TransGainForRollOver when φ_ref <0. In the case of φ_ref <0, the rollover correction transient gain TransGainForRollOver is calculated according to the characteristics of the region of φ_ref <0 in the map shown in FIG. First, in step S140, it is determined whether or not φ_ref ≧ TH1_M. If φ_ref ≧ TH1_M, the process proceeds to step S142 to set TransGainForRollOver = MAX_GAIN. If φ_ref <TH1_M in step S140, the process proceeds to step S144 to determine whether φ_ref ≧ TH2_M. If φ_ref ≧ TH2_M, rollover correction transient gain TransGainForRollOver is calculated from the following equation (21). To do.

  If φ_ref ≧ TH2_M in step S144, the process proceeds to step S148, where TransGainForRollOver = MIN_GAIN.

  As described above, according to the processing of FIGS. 18 and 19, when sgn_TurnSt = ON, the rollover correction transient gain TransGainForRollOver can be calculated based on the map of FIG.

  As shown in FIGS. 10 and 15, when φ_ref is larger than TH2_P, the rollover correction steady-state gain ConstGainForRollOver becomes 0, and the turning support is not performed in the steady term. Further, when φ_ref becomes larger than TH3_P, the rollover correction steady-state gain ConstGainForRollOver becomes a negative value, and depending on the setting of the rollover correction transient gain TransGainForRollOver at that time, the target damping moment MgDampTgt and the target inertia compensation moment MgTransTgt are set. The added value becomes a negative value. Therefore, the control target moment MgTgt can be generated in a direction to suppress the turning. Thereby, it is possible to perform control so as to sequentially change to a moment that supports turning, a moment that does not support turning, and a moment in the opposite direction to turning as φ_ref increases.

  If sgn_TurnSt = OFF in step S110 of FIG. 16, the process proceeds to step S114. In step S114, it is determined whether or not the same situation has continued three or more times. If the same situation has continued three or more times, the process proceeds to step S116. In step S116, TransGainForRollOver = TransGainForRollOver (n) is set, and the normal control mode is restored (OUT). On the other hand, if the same situation does not continue three or more times in step S114, the process proceeds to step S118. In step S118, a rollover correction transient gain TransGainForRollOver is calculated from the following equation (22).

  In the present embodiment, when changing from the steering return state to the normal turning mode, the correction gain is gradually changed by taking the average value of the three sampling data, but the average value of the two sampling data (two parameters). Alternatively, the correction gain may be gradually changed by calculating an average value of data of 4 samplings (4 parameters) or more.

  In the above example, the rollover correction steady-state gain ConstGainForRollOver and the rollover correction transient gain TransGainForRollOver are calculated using φ_ref. However, these correction gains may be calculated from the differential value of φ_ref instead of φ_ref. good.

  The rollover correction steady-state gain ConstGainForRollOver and the rollover correction transient gain TransGainForRollOver calculated by the rollover adjustment gain calculation unit 285 as described above are input to the control target moment calculation unit 290. The control target moment calculator 290 calculates the control target moment MgTgt from the difference ΔγTgt between the control target yaw rate γTgt and the feedback yaw rate γF / B. In addition, the control target moment calculation unit 290 corrects the control target moment MgTgt using the rollover correction steady-state gain ConstGainForRollOver and the rollover correction transient gain TransGainForRollOver, thereby ensuring stability during reverse steering and turning during normal steering. Make support control compatible.

The target moment calculator 290 calculates a moment for correcting the vehicle behavior based on the difference ΔγTgt. For this reason, as shown in FIG. 4, the target moment calculation unit 290 calculates a value D1 based on a coefficient applied to the yaw rate in a known plane two-wheel model (yaw motion) represented by the following equation (23). A value T1 based on a coefficient applied to the yaw acceleration in the equation (23), a damping control moment calculating unit (steady term calculating unit) 290a that calculates “target damping moment MgDampTgt” by multiplying ΔγTgt, and a differential value of ΔγTgt. The inertia compensation moment calculation unit (transient term calculation unit) 290b for calculating “target inertia compensation moment MgTransTgt” by multiplying by. Here, the coefficient D1 is d [gamma] corresponds to a value based on the formula ₂ ^/ V hanging on γ in _{(23) (l f 2 K} f -l r 2 K r), the coefficient T1 is the formula (23) This corresponds to a value based on I (yaw inertia of the vehicle) applied to / dt. Further, the target moment calculation unit 290 includes an addition unit 290c that adds “target damping moment MgDampTgt” and “target inertia compensation moment MgTransTgt”.

  The damping control moment calculation unit (steady term calculation unit) 290a multiplies the yaw rate correction amount ΔγTgt by a constant D1 derived from the coefficient applied to the yaw rate in Expression (23) in which the planar two-wheel model is arranged with respect to the yaw moment. Thus, the damping moment basic amount MgDampBasis that improves the convergence performance when the vehicle turns is calculated. That is, the damping moment basic amount MgDampBasis is calculated from the following equation (24). ΔγTgt is the difference between the control target yaw rate γTgt and the feedback yaw rate γF / B, and corresponds to the yaw rate correction amount (target value). The constant D1 is a calculation coefficient for the damping moment.

  Further, the damping control moment calculation unit (steady term calculation unit) 290a multiplies the damping moment basic amount MgDampBasis by the damping moment correction gain (rollover correction steady gain ConstGainForRollOver) by the following equation (25) to obtain the target damping moment. Calculate MgDampTgt.

  The inertia compensation moment calculation unit (transient term calculation unit) 290b is an expression in which the two-wheel model is arranged with respect to the yaw moment, and multiplies the yaw rate correction amount by a constant T1 derived from the coefficient applied to the yaw acceleration. Thus, the basic amount MgTransBasis of the inertia compensation moment for compensating the yaw inertia when the vehicle turns is calculated. That is, the inertia compensation moment basic amount MgTransBasis is calculated from the following equation (26). ΔγTgt is the difference between the control target yaw rate γTgt and the feedback yaw rate γF / B, and corresponds to the yaw rate correction amount (target value). The constant T1 is an inertia compensation moment calculation coefficient.

  Then, the inertia compensation moment calculation unit (transient term calculation unit) 290b multiplies the inertia compensation moment basic amount MgTransBasis and the inertia compensation moment correction gain (rollover correction transient gain TransGainForRollOver) by the following equation (27): A target inertia compensation moment MgTransTgt is calculated. The target inertia compensation moment MgTransTgt is a function for limiting the output of the inertia compensation moment to suppress rollover when the turning back of the steering is detected and when φ_ref reaches a value equal to or greater than a predetermined threshold value, It also has a turning support function during normal steering.

  Thereafter, the adder 290c of the control target moment calculator 290 adds the target damping moment MgDampTgt and the target inertia compensation moment MgTransTgt according to the following equation (28) to calculate the control target moment MgTgt.

  The control target moment MgTgt calculated by the control target moment calculation unit 290 as described above is input to the motor required torque calculation unit 295. The motor request torque calculator 295 calculates the motor request torque based on the control target moment MgTgt.

  Next, the vehicle behavior when the control of the present embodiment is performed will be described based on FIGS. FIG. 20 and FIG. 21 are characteristic diagrams for explaining the effect of the control according to the present embodiment, and an example in which the road surface friction coefficient μ changes during lane change traveling (the vehicle speed is constant) is shown as an example. This is a comparison of the vehicle behavior in general control (only the turning support mode) and the vehicle behavior by the control of the present embodiment by simulation. Here, FIG. 20 shows the vehicle behavior in general control (only the turning support mode), and FIG. 21 shows the vehicle behavior by the control of the present embodiment. 20 and 21, as the vehicle behavior, the steering angle θh, the actual yaw rate γ, the lateral acceleration Gy, the rollover correction steady gain ConstGainForRollOver, the rollover correction transient gain TransGainForRollOver, the vehicle yaw moment, the roll angle φ, the roll rate. The change of φ_dot is shown. 20 and 21, the road surface friction coefficient is switched from high μ to low μ at time t1.

  In the general control shown in FIG. 20, when attention is paid to the actual yaw rate γ at the timing of the lane change when the steering angle θh is changed, overshoot accompanying steering and overshooting due to steering reversal in both the high μ range and the low μ range. A shoot has occurred. Similarly, when attention is paid to the yaw moment of the vehicle at the timing of the lane change at which the steering angle θh has changed, overshoot accompanying steering and overshoot accompanying turning back occurs in both the high μ range and the low μ range. . At the timing of the lane change when the steering angle θh changes, the vehicle yaw moment and the roll rate φ_dot vibrate greatly, and this vibration occurs in the high μ region and the low μ region. It turns out that it is increasing more.

  On the other hand, in the control according to the present embodiment shown in FIG. 21, a vehicle that is generated in the vehicle at the time of steering return or at the time of steering input at low μ by correcting the control target moment MgTgt by the roll angle φ generated in the vehicle body. It can be seen that the yaw moment, roll angle φ, and roll rate φ_dot are suppressed. Therefore, according to the control according to the present embodiment, it is possible to greatly stabilize the vehicle behavior during turning.

  As described above, according to the present embodiment, the reference value φ_ref is calculated according to the reliability of the roll model, and the rollover correction steady-state gain ConstGainForRollOver and the rollover correction transient gain TransGainForRollOver are calculated according to the value of the reference value φ_ref. And the control target moment MgTgt is corrected. As a result, when the reference value φ_ref is large and the vehicle 1000 rolls greatly, the control target moment MgTgt is optimized by reducing the rollover correction steady-state gain ConstGainForRollOver and the rollover correction transient gain TransGainForRollOver. Can be controlled. Therefore, the roll of the vehicle 1000 can be optimally controlled.

  The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field to which the present invention pertains can come up with various changes or modifications within the scope of the technical idea described in the claims. Of course, it is understood that these also belong to the technical scope of the present invention.

200 Control Device 220 Control Target Yaw Rate Calculation Unit 252 Subtraction Unit 254 Subtraction Unit 260 Feedback Yaw Rate Calculation Unit 284 Feedback Roll Angle Calculation Unit 285 Rollover Adjustment Gain Calculation Unit 290 Control Target Moment Calculation Unit

Claims (8)

A target yaw rate acquisition unit for acquiring a target yaw rate of the vehicle;
A roll rate acquisition unit that acquires a first roll angle obtained from a vehicle model and a second roll angle obtained from a sensor;
Based on the difference between the first roll angle and the second roll angle, the distribution of the first roll angle is increased when the difference is small, and the second roll angle when the difference is large. A feedback roll angle calculator that calculates a feedback roll angle from the first roll angle and the second roll angle,
As a yaw rate generated by the vehicle, a feedback yaw rate acquisition unit that acquires a feedback yaw rate for comparison with the target yaw rate;
A control target moment calculating unit that calculates a control target moment based on a difference between the target yaw rate and the feedback yaw rate, and corrects the control target moment based on the feedback roll angle;
A vehicle control device comprising: The control target moment calculating unit calculates the control target moment by adding a steady damping control moment and a transient inertia compensation moment,
The vehicle control device according to claim 1, wherein the damping control moment and the inertia compensation moment are corrected based on the feedback roll angle.   The control target moment calculator corrects the damping control moment and the inertia compensation moment so that the damping control moment or the inertia compensation moment decreases as the absolute value of the feedback roll angle increases. The vehicle control device according to claim 2.   The control target moment calculation unit sequentially changes the control target moment into a moment that supports turning, a moment that does not support turning, and a moment in a direction opposite to turning, as the absolute value of the feedback roll angle increases. The vehicle control device according to claim 2, wherein the damping control moment and the inertia compensation moment are corrected.   5. The apparatus according to claim 2, further comprising a rollover adjustment gain calculation unit that calculates a gain multiplied by each of the damping control moment and the inertia compensation moment based on the feedback roll angle. Vehicle control device.   Based on the steering angle and the yaw rate, it is determined whether or not the steering operation has been turned back, and when the control target moment calculation unit determines that the steering operation has been turned back, the feedback roll angle The vehicle control device according to claim 1, wherein the control target moment is corrected based on the control target moment. A yaw rate acquisition unit for acquiring a first yaw rate obtained from a vehicle model and a second yaw rate obtained from a yaw rate sensor;
The feedback yaw rate acquisition unit, based on the difference between the first yaw rate and the second yaw rate, increases the distribution of the first yaw rate when the difference is small, and the first yaw rate when the difference is large. 7. The vehicle control device according to claim 1, wherein the feedback yaw rate is calculated from the first and second yaw rates by increasing the distribution of the yaw rate of 2. Obtaining a target yaw rate of the vehicle;
Obtaining a first roll angle determined from a vehicle model and a second roll angle determined from a sensor;
Based on the difference between the first roll angle and the second roll angle, the distribution of the first roll angle is increased when the difference is small, and the second roll angle when the difference is large. Increasing the distribution of the first roll angle and calculating the feedback roll angle from the second roll angle and the second roll angle;
Obtaining a feedback yaw rate for comparison with the target yaw rate as a yaw rate generated by the vehicle;
Calculating a control target moment based on a difference between the target yaw rate and the feedback yaw rate, and correcting the control target moment based on the feedback roll angle;
A vehicle control method comprising:

Images