JPH11189063A - Road surface friction coefficient estimating device for vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make properly catching possible of a road surface condition (road surface friction coefficient) for a vehicle in running, so that for instance, various running controls in an automobile can be properly performed. SOLUTION: This device has a parameter calculation means 200 calculating a first parameter βpp indicating an irregularity condition of a road surface and a second parameter αh indicating slip easiness of the road surface from a running condition during steady running of a vehicle, a road surface index calculation means 202 calculating a road surface index from a value of the first/second parameter βpp, αh, and a road surface friction coefficient calculation means 204 continuously accumulation obtaining the calculated road surface index evaluated to calculate a road surface friction coefficient. In the parameter calculation means 200, based on a car speed, right/left wheel rotational speed difference, and actual longitudinal/cross acceleration, the second parameter αh is calculated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device for estimating a friction coefficient of a road on which a vehicle travels, and more particularly to a vehicle suitable for use in a vehicle having a torque transfer control device for controlling transfer of torque between left and right. Road friction coefficient estimating apparatus.

[0002]

2. Description of the Related Art In recent years, various traveling controls have been developed in order to enhance the traveling performance of an automobile and to assist an artificial operation relating to traveling. Such running control includes, for example, engine output control called traction control, differential limiting control between left and right wheels and front and rear wheels, and torque distribution control (power transmission between left and right wheels and front and rear wheels). Control) has already been developed.

For example, a differential mechanism is provided between left and right wheels, which are driving wheels of an automobile, to allow a differential that occurs during turning or the like. In this differential mechanism, one of the left and right wheels is provided. If one of the wheels slips on the sand, for example, it may rotate, so that the other wheel hardly rotates, and a state where the driving torque cannot be transmitted to the road surface may occur.

Therefore, in such a case, a differential limiting mechanism (LSD = Limited Slip Diff) capable of limiting the differential has been developed. Such a left and right wheel differential limiting mechanism includes a type that is proportional to the rotational speed difference between the left and right wheels and a type that is proportional to the input torque. The right and left wheel rotation speed difference proportional type uses VC that utilizes the viscosity of the liquid.
(Viscous coupling) type LSD, etc.
There is an advantage that the running stability of the vehicle can be improved. On the other hand, the input torque proportional type includes a mechanical type such as a friction type such as a general LOM (lock automatic) type LSD, and has an advantage that the turning performance of the vehicle can be improved.

However, in the various types of differential limiting mechanisms described above, the differential control characteristics are determined by physical properties and the like, and the differential control characteristics can always be adjusted so that differential control can always be appropriately performed. Has not become. There is also a system called an electronically controlled LSD in which the LSD is electronically controlled. However, even in such a system, the torque transfer between the wheels is limited only from a high speed side to a low speed side. It is considered that the traveling performance cannot be sufficiently improved during the turning traveling of the vehicle.

The present applicant has proposed, for example, Japanese Unexamined Patent Publication No. Hei 5-131855 (hereinafter, referred to as Japanese Patent Application Laid-Open No. Hei 5-131855) so that torque can be distributed between left and right wheels in various running states of a vehicle without causing a large torque loss or energy loss. Kaihei 7-108840
JP-A-7-108841, JP-A-7-108842,
Japanese Patent Laying-Open Nos. Hei 7-108843 and Hei 7-156681 have proposed a torque transfer control device for left and right wheels for a vehicle as disclosed in Japanese Patent Application Laid-Open Nos. 7-108843 and 7-156681.

In this torque transfer control device between left and right wheels, when two rotating bodies arranged coaxially are slid in contact with each other at different rotating speeds, the rotating body having a higher rotating speed has a lower rotating speed. This utilizes the characteristic that torque is transmitted to the rotating body. That is, for example, this device is provided with a transmission mechanism that changes the rotation speed input to the differential device or the rotation speed of one of the wheel shafts to high and low speeds and outputs the same, and receives the respective outputs of the transmission mechanism to make a difference. A plurality of speed change interlocking members that rotate at a rotation speed different from the driving device or one of the wheel shafts, a constant speed interlocking member that rotates at the same speed as the other wheel shaft of the left and right wheels, and these speed change interlocking members It is provided with a plurality of torque transmission couplings such as a wet multi-plate clutch provided between the transmission and the speed interlocking member.

In such a device, even if the left and right wheels are rotating at a constant speed, the torque transmission coupling requires:
Since the rotational speed is different between the speed change interlocking member side and the constant speed interlocking member side, if the torque transmission coupling is actuated by engaging a wet multi-plate clutch or the like, the speed change interlocking member side and the constant speed interlocking member side can be changed. The torque is transmitted from the higher speed to the lower speed. If the degree of speed change by the speed change mechanism is set to be greater than or equal to a certain value, torque can be transmitted from the inner wheel having a lower rotation speed to the outer wheel having a higher rotation speed during turning.

Further, in a torque transmission coupling such as a wet multi-plate clutch, for example, the switching of engagement of each wet multi-plate clutch and the control of the degree of engagement are performed so that the transmission torque to one wheel axle is controlled. Can be increased or decreased, and the transmission torque to the other wheel axle can be increased or decreased. Therefore, since the transmission torque capacity can be variably controlled, the torque can be transmitted to the left and right wheels in a desired direction with a desired transmission torque capacity.

Such a device can be applied whether the left and right wheels are driving wheels or driven wheels, and if the left and right wheels are driving wheels, the distribution of the driving force from the engine to the left and right wheels is adjusted. If the left and right wheels are driven wheels, the wheels on the torque transmitting side will receive the driving force by the torque transmission, and the wheels on the torque transmitting side will receive the braking force.

In any case, the magnitude of the driving force or braking force exerted between each of the left and right wheels and the road surface is imbalanced in the left and right directions, thereby generating a yaw moment in the vehicle and reducing the behavior of the vehicle. Can be controlled. Further, such torque transmission control can be considered not only between the left and right wheels but also between the front and rear wheels.

[0012]

By the way, in the above-mentioned apparatus for performing torque transmission control (power transmission control) between the left and right wheels and between the front and rear wheels of the vehicle, the torque transmission control corresponding to the running condition of the vehicle is performed. Need to be done. That is,
Performing torque transmission control according to whether the vehicle is traveling straight or turning, accelerating or decelerating, whether each wheel is slipping, etc., improves the running performance of the vehicle. .

However, in order to perform more appropriate control, it is necessary to grasp the road surface condition on which the vehicle travels. In other words, it is considered that the control effect of the torque transmission control is different depending on whether or not the traveling road surface is slippery. To properly obtain the required control effect, the road surface condition is grasped and the torque based on the road surface condition is determined. It is necessary to perform transmission control (power transmission control).

Such a road surface condition can be expressed as a so-called road surface friction coefficient (road surface friction coefficient). The road surface condition (road surface friction coefficient) can be grasped only by the torque transmission control (power transmission control) described above. Instead, it is effective in various traveling controls in automobiles. Therefore, there is a demand for the development of a technology capable of estimating the road surface friction coefficient (road surface μ) of a running road while the vehicle is running.

In addition, in order to reliably reflect the estimation result of the road surface friction coefficient in the traveling control, it is necessary to ensure that the road surface friction coefficient can be estimated even when a device related to the traveling control provided in the vehicle is operated. . In order to estimate the road surface friction coefficient of the traveling road while the vehicle is traveling, it is sufficient if there are parameters corresponding to the road surface friction coefficient that can be detected during traveling of the vehicle. As one of such parameters, a slip state of the drive wheels with respect to the traveling road surface can be considered. That is, when a constant driving force is applied from the drive wheels to the road surface, the smaller the road surface friction coefficient, the more the drive wheels slip on the road surface. Therefore, by paying attention to this characteristic, the road surface friction coefficient can be estimated.

[0016] However, it is not easy to grasp how much driving force each driving wheel intends to exert on the road surface. On the other hand, in the case where the above-described torque transfer control device is provided between the left and right wheels, when the torque is transferred between the left and right wheels, the driving force transmitted between the wheels and the road surface is correspondingly changed. Since there is a change between the left and right wheels, the rotational speed difference between the left and right wheels with respect to this torque movement can be considered as one of the parameters indicating the road surface friction coefficient of the traveling road.

However, the difference between the rotational speeds of the left and right wheels depends not only on the torque movement but also on various other factors.
Therefore, it is desired to be able to accurately detect a parameter relating to the rotational speed difference between the left and right wheels, which corresponds to the road surface friction coefficient of the traveling road, while also considering these various rotational speed difference generation factors. Further, the road surface friction coefficient of the traveling road affects not only the easiness of slipping of the wheels on the road surface but also the unevenness of the road surface.

Therefore, it is desired to accurately grasp the unevenness of the road surface and to estimate the road surface friction coefficient of the traveling road from such a surface. By the way, Japanese Patent Application Laid-Open No. Hei.
In the official gazette, the wheel acceleration Gv obtained from the average wheel speed is described.
ΔG between the acceleration and the vehicle acceleration G detected by the acceleration sensor
Is compared with a threshold value ΔGs, if ΔG> ΔGs, a slip is determined, and a low μ road determination is made based on the vehicle body acceleration G.

Japanese Patent Application Laid-Open No. Hei 5-238404 discloses a sum of a longitudinal acceleration detected by a longitudinal acceleration sensor and a lateral acceleration detected by a lateral acceleration sensor (first vector sum), and a wheel speed sensor. The sum (second vector sum) of the longitudinal slip acceleration obtained by subtracting the longitudinal acceleration from the wheel acceleration obtained by time-differentiating the detected wheel speed and the lateral slip acceleration obtained from the information of the vehicle speed sensor and the yaw rate sensor. ), A technology for estimating the road surface friction coefficient is disclosed.

Japanese Patent Application Laid-Open No. 5-338457 discloses a technique for estimating a road surface friction coefficient from each information of a steering angle, a steering torque and a wheel speed. Further, Japanese Patent Application Laid-Open No. 7-128221 discloses that a slip ratio is obtained by dividing a rotation speed difference between a driving wheel and a driven wheel by a rotation speed of a driven wheel, and the slip ratio is calculated from the slip ratio and the driving force of the tire. A technique for estimating a road surface state is disclosed.

However, these techniques cannot reliably estimate the road surface friction coefficient even when a device related to travel control mounted on an automobile is operated, and secure sufficient estimation accuracy depending on the traveling state of the vehicle. It is difficult. Japanese Patent Application Laid-Open No. 5-346394 discloses a technique for estimating a road surface friction coefficient based on a relationship between a change amount of an operation limitation amount and a change amount of a rotation speed difference in a differential device having a differential limiting device. Is disclosed. However, this technique can estimate the road surface friction coefficient even if the vehicle does not reach the traveling limit, but it is difficult to secure sufficient estimation accuracy depending on the traveling state of the vehicle.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems, and enables a road condition (a coefficient of road surface friction) on which a vehicle travels to be accurately grasped so as to more appropriately perform, for example, various traveling controls in an automobile. I was able to
An object of the present invention is to provide a road surface friction coefficient estimating device for a vehicle.

[0023]

According to a first aspect of the present invention, there is provided a road surface friction coefficient estimating apparatus for a vehicle, wherein a vehicle state detecting means detects a running state of the vehicle, and a parameter calculating means detects the vehicle state. The first parameter indicating the unevenness of the road surface and the second parameter indicating the ease of road surface slippage are calculated from the running state of the vehicle during steady running detected by the detection means.

Then, the road surface index calculating means is provided with the first
And an index corresponding to each state of a third parameter that represents the road surface state in a unified manner from the values of the second parameter and the road surface friction coefficient calculating means. The road surface friction coefficient calculating means continues the index calculated by the road surface index calculating means. The road surface friction coefficient is calculated by accumulatively calculating and evaluating the road surface friction coefficient. Especially,
In the parameter calculation means, the calculation of the second parameter is performed by the vehicle speed detected or estimated by the vehicle speed detection means, the left and right wheel rotation speed difference detected by the left and right wheel rotation speed difference detection means, and the longitudinal acceleration detection means. This is performed based on the detected actual longitudinal acceleration and the actual lateral acceleration detected by the lateral acceleration detecting means.

At this time, the left and right wheel rotational speed difference detecting means:
In order to detect the rotation speed difference between the left and right wheels provided with a torque movement control device for controlling the movement of the torque between the pair of left and right wheels, the calculation of the second parameter and the estimation of the road surface friction coefficient are performed using this method. This is performed based on the difference between the left and right wheel rotational speeds of the left and right wheels provided with the torque movement control device. The vehicle road surface friction coefficient estimating apparatus of the present invention according to claim 2 is:
The vehicle state detecting means detects a running state of the vehicle, and the parameter calculating means calculates a first parameter indicating a road surface unevenness state and a road surface slip from the running state of the vehicle during steady running detected by the vehicle state detecting means. The second parameter indicating the ease is calculated.

Then, the road surface index calculating means performs the first
And an index corresponding to each state of a third parameter that represents the road surface state in a unified manner from the values of the second parameter and the road surface friction coefficient calculating means. The road surface friction coefficient calculating means continues the index calculated by the road surface index calculating means. The road surface friction coefficient is calculated by accumulatively calculating and evaluating the road surface friction coefficient. Especially,
The parameter calculation means performs gain correction, high-pass processing, and low-pass processing according to the theoretical lateral acceleration calculated by the lateral acceleration calculation means on the actual lateral acceleration detected by the lateral acceleration detection means, The above first parameter is calculated.

[0027]

Embodiments of the present invention will be described below with reference to the drawings. Here, a vehicle road surface friction coefficient estimating apparatus as the present embodiment will be described, and a vehicle left-right wheel power transmission control apparatus (torque transfer control apparatus) including the vehicle road surface friction coefficient estimating apparatus will be described below. These items will be described in order.

Here, the power transmission control device between the left and right wheels for a vehicle is also abbreviated as "torque transfer control device" or simply "device", but the vehicle road surface friction coefficient estimating device is referred to without omission. Or, it is referred to as a road surface μ estimation device. [Table of Contents] 1. Outline of the system of the torque transfer control device 1.1 Concept of hardware configuration of the torque transfer control device 1.2 Hardware configuration of the torque transfer control device 1. Overview of control of the three torque transfer control devices 2. 2. Control details of this device 2.1 Input calculation processing 2.2 Drift determination logic 2.3 Vehicle motion control logic 2.3.1 Target ΔN tracking control 2.3.2 Acceleration turning control 2.3.3 Tack-in correspondence control 2 3.3.4 Steering transient response control 2.4 Estimation of road surface μ 2.4.1 Estimation of road surface μ at steady turning 2.4.2 Estimation of road surface μ at non-linear turning 2.4.3 Estimation of road surface μ at starting 2 4.4.4 Output value setting 2.5 Actuator drive Operation of this device and effects of this device 3.1 Operations of this device 3.2 Effects of this device 3.2.1 Effect of device for estimating road surface friction coefficient 3.2.2 Vehicle road surface friction coefficient (road surface μ ) Effect of response control [Embodiment] 1. System overview of this device 1.1 Concept of hardware configuration of this device First, the concept of hardware configuration of this device will be described.
The power transmission control device for left and right wheels for a vehicle, when two rotating bodies disposed coaxially are slid in contact with each other at different rotating speeds, the rotating body having the higher rotating speed has the lower rotating speed. This utilizes the characteristic that torque is transmitted to the rotating body.

For example, if the left wheel rotates at a higher speed than the right wheel, torque can be easily transmitted from the left wheel to the right wheel, and the right wheel is easier than the left wheel. If the motor rotates at a high speed, torque can be easily transmitted from the right wheel to the left wheel. Therefore, even if the left and right wheels rotate at a constant speed, if the left wheel side can realize a state of rotating at a higher speed than the right wheel side, torque can be transmitted from the left wheel side to the right wheel side. Even if the wheels rotate at a constant speed, torque transmission from the right wheel side to the left wheel side can be performed if a state in which the right wheel side rotates faster than the left wheel side can be realized.

That is, for example, if a speed change mechanism for shifting the rotation speed VL of the left wheel side at high speed is provided on the left wheel side, or a speed change mechanism for shifting the rotation speed VR of the right wheel side at low speed is provided on the right wheel side, Even if the wheels are rotating at a constant speed, between the left wheel side member receiving the output of the speed change mechanism (speed increasing mechanism) and the right wheel side member rotating at the same speed VR as the right wheel, or the speed change mechanism ( Between the right wheel side member receiving the output of the speed reduction mechanism) and the left wheel side member rotating at the same speed VL as the left wheel, a state where the rotation speed on the left wheel side is higher than that on the right wheel side can be realized.

If the right wheel side has a symmetrical functional configuration, a state where the right wheel side is larger than the left wheel side can always be realized. Depending on the setting of the speed ratio of such a speed change mechanism, when the vehicle turns, originally,
Although the turning inner wheel rotates at a lower speed than the turning outer wheel, the rotating member on the inner wheel side can be shifted at a higher speed than the rotating member on the outer wheel side even when the vehicle turns.

By providing a torque transmission coupling between the left wheel side rotating member and the right wheel side rotating member provided with the speed difference in this way, and by appropriately operating the torque transmitting coupling, a constant torque is provided. Under running conditions, torque can be constantly transmitted from the left wheel side to the right wheel side and from the right wheel side to the left wheel side. Of course, if the gear ratio is set by the speed change mechanism so that the driving torque on the inner wheel side is transmitted to the outer wheel side even when turning at the maximum steering angle, from the left wheel side to the right wheel side under all running conditions, Torque can be constantly transmitted from the right wheel side to the left wheel side.

In a coupling of variable torque transmission capacity such as a wet multi-plate clutch mechanism, the amount of transmission torque can be generally adjusted according to the engagement pressure (pressing force P) and the like. Incidentally, the transmission mechanism and the coupling interposed between the right wheel side and the left wheel side are provided directly between the right wheel side and the left wheel side. By providing these transmission mechanisms and couplings between the wheel side (right wheel side or left wheel side), power transmission between the left wheel side and right wheel side via an input part of the differential ( Torque transfer).

The power transmission (torque transfer) between the left and right wheels of the vehicle according to such a principle can be applied whether the left or right wheel is a driving wheel or a driven wheel.
The distribution of the driving force (torque) from the engine to the left and right wheels will be adjusted. If the left and right wheels are driven wheels, the torque transmitting wheels will receive the driving force by torque transmission. Thus, the wheel on the side that transmits the torque receives the braking force. In any case, the magnitude of the driving force or the braking force exerted between each of the left and right wheels and the road surface is imbalanced in the left and right directions, thereby generating a yaw moment in the vehicle and thereby changing the behavior of the vehicle. Can be controlled.

1.2 Hardware Configuration of the Apparatus Next, the hardware configuration of the power transmission control apparatus for left and right wheels for a vehicle based on the above theory will be described with reference to FIGS. 1.2.1 Configuration of the vehicle power transmission system according to the present apparatus
As shown in the figure, it is provided on the rear wheel of a four-wheel drive vehicle.

In FIG. 1, reference numeral 2 denotes an engine,
The output of the engine 2 is transmitted to a differential gear mechanism (= center differential, hereinafter, referred to as a center differential) 8 via a transmission 4 and an intermediate gear mechanism 6. The output of the center differential 8 is transmitted to the axle 1 via a front differential gear mechanism (= front differential, hereinafter referred to as front differential) 10 on the one hand.
2L, 12R to the left and right front wheels 14, 16, while the bevel gear mechanism 18, the propeller shaft 20
The transmission is transmitted from the axles 26L, 26R to the left and right rear wheels 28, 30 via a bevel gear mechanism 22, and a rear wheel differential gear device (= rear differential, hereinafter, referred to as a rear differential) 24. The rotary propulsion force distribution adjusting mechanism (or torque adjusting means, hereinafter, referred to as a torque transfer mechanism) 50 of the power transmission control device between the left and right wheels is provided in the rear differential 24.

The center differential 8 is composed of differential pinions 8A and 8B and side gears 8C and 8D meshed with the differential pinions 8A and 8B, as in the conventional well-known one. The torque is applied to the side gear 8
C, 8D, from the side gear 8C to the front wheel side,
The differential is transmitted from the side gear 8D to the rear wheel side while being allowed.

Here, from the side gear 8C to the front differential 10 on the front wheel side via the front wheel output shaft 32, from the side gear 8D, the rear wheel output shaft 34 and the bevel gear mechanism 1
8, torque is transmitted from the propeller shaft 20 to the rear wheel side. The center differential 8 restricts (or restricts) the differential between the front-wheel output portion and the rear-wheel output portion to distribute the output torque (rotational propulsion) of the engine between the front and rear wheels. A viscous coupling unit (VCU) 36 is provided as a controllable differential limiting means (ie, a limited slip differential (LSD)).

The VCU 36 is interposed between the front wheel output shaft 32 and the rear wheel output shaft 34, and limits the differential between the front wheel side and the rear wheel side with a force according to the differential state. by doing,
It is possible to avoid a situation in which only the light load side of the front and rear wheels idles and the rotational torque is not transmitted to the heavy load side. 1.2.2 Configuration of Rotational Propulsion Force Distribution Adjusting Mechanism of Present Device The present left and right wheel power transmission control device includes a rotational propulsion force distribution adjusting mechanism (torque moving mechanism) 50 provided in a differential carrier 51, It comprises a hydraulic unit 38 and an electronic control unit (hereinafter referred to as ECU) 42 as control means (or rotational propulsion force distribution control means). Here, the rear differential 24 and the rear differential 24 and the axle 26
The configuration of the torque transfer mechanism 50 inserted between the L and 26R will be described with reference to FIG.

As shown in FIG. 2, an input shaft 52 is connected to the rear end of the propeller shaft 20, and a drive pinion gear 54 is connected to the input shaft 52 so as to rotate integrally. The drive pinion gear 54 meshes with a crown gear 56 provided on the outer periphery of a differential case (differential case) 58. The output of the engine is transmitted from the input shaft 52 to the rear differential 24 via the drive pinion gear 54 and the crown gear 56. It is being conveyed.

The rear differential 24 has two pairs of pinions provided in a differential case 58, that is, differential pinions 60A and 60B, and side gears 62 and 64 meshed with these differential pinions 60A and 60B, as in the conventional case. The rotational torque input from the differential pinions 60A and 60B is
The differential gears are transmitted to the side gears 62 and 64 and transmitted from the side gear 62 to the left-wheel rotating shaft 66 and from the side gear 64 to the right-wheel rotating shaft 68 while allowing the respective differentials to be allowed. . Also, left and right rotating shafts 66,
1, 68 is connected to axles 26L, 26R connected to the left and right rear wheels 28, 30 as shown in FIG.

The torque transfer mechanism 50 of the present embodiment is provided between a differential case 58 of the rear differential 24 for distributing driving force between left and right driving wheels of the rear wheels and a right wheel side rotating shaft 68.
Transmission mechanism 70 and transmission capacity variable control torque transmission mechanism 90
The transmission of rotational propulsion between the left wheel side and the right wheel side, that is, power transmission (torque movement), is performed via the differential case 58.

The speed change mechanism 70 includes a speed increase mechanism 70A for increasing the rotational speed of the input portion of the rear differential 24, that is, the rotational speed of the differential case 58, and outputting it to one of the left and right wheels (here, the right wheel side).
Since it is integrally provided with a speed reduction mechanism 70B for decelerating and outputting to one side (right wheel side), it is also referred to as an acceleration / deceleration mechanism. The transmission capacity variable control type torque transmission mechanism 90 uses a wet hydraulic multi-plate clutch mechanism (hereinafter, also referred to as a clutch) capable of adjusting the transmission capacity according to the control oil pressure.
A clutch (left clutch) 90L provided between the output side of the speed reduction mechanism 70B of the transmission mechanism 70 and the right wheel side to transmit torque to the left wheel side, and the output side of the speed increasing mechanism 70A of the transmission mechanism 70 and the right wheel. A clutch (right clutch) 90 </ b> R provided between the right side and the right side to transmit torque is integrally formed. Such a transmission capacity variable control type torque transmission mechanism 90 is also referred to as an integral coupling or simply a coupling.

The acceleration / deceleration mechanism 70 will be described. The acceleration / deceleration mechanism 70 includes a hollow intermediate shaft 72 connected to rotate integrally with the differential case 58, a hollow intermediate shaft 74 connected to the right clutch 90R, It is interposed between a hollow intermediate shaft 76 connected to the left clutch 90L. Each of the intermediate shafts 72, 74, 76 is a hollow shaft, and the intermediate shafts 72, 74 are mounted on the outer periphery of the right wheel side rotation shaft 68 so as to be able to relatively rotate. 7
4 is also provided on the outer circumference so that it can also be rotated relative to each other.

Gears 78A, 80A, 82A are provided on these intermediate shafts 72, 74, 76, respectively, and a counter shaft 84 is provided on the outer periphery of these intermediate shafts 72, 74, 76. This counter shaft 84
Is provided with a triple gear 86. Triple gear 86
Is composed of gears 78B, 80B, 82B,
8B is the gear 78A of the intermediate shaft 72, the gear 80B is the gear 80A of the intermediate shaft 74, and the gear 82B is the gear 8 of the intermediate shaft 76.
2A, respectively.

The acceleration / deceleration mechanism 70 is provided with such a gear 78
Intermediate shafts 72, 74, 76 having A, 80A, 82A
, Counter shaft 84, gears 78B, 80B, 8
And a triple gear 86 having 2B. Note that the counter shaft 84 has intermediate shafts 72, 74, 76.
A plurality (three in this case) of which the phase is shifted from that of the drive pinion 54 is provided on the outer periphery. Although this does not include a ring gear, it has the same arrangement as a triple planetary gear mechanism in which the gears 78A, 80A, and 82A are sun gears and the gears 78B, 80B, and 82B are planetary pinions.

Each counter shaft 84 is fixed to a wall 51A provided on the differential carrier 51.
Therefore, the gears 78B, 80B, and 82B only rotate about the counter shaft 84 as the axis. This allows
The intermediate shafts 72, 74, 76 are supported in the radial direction by gears 78A, 80A, 82A and gears 78B, 80B, 82.
Through the engagement with B, it is also performed by the plurality of counter shafts 84 fixed to the wall 51A as described above.

The gears 78A, 80A, 8
Assuming that the number of teeth of 2A is Z ₁ , Z ₂ , and Z ₃ respectively, Z ₂
<Z ₁ <Z ₃ is set. Also, the gear 78
B, 80B, and 82B are represented by Z ₄ , Z ₅ , and Z ₆ , respectively.
Then, the relation of Z ₆ <Z ₄ <Z ₅ is established.

Thus, the transmission mechanism (acceleration / deceleration mechanism) 70
Then, gear 78A, gear 78B, gear 80A, gear 80
The combination of gears B constitutes a speed increasing mechanism 70A for increasing the rotation input to the rear differential 24 and outputting the rotation to the right wheel side. The combination of the gears 78A, 78B, 82A, and 82B is input to the rear differential 24. A deceleration mechanism 70B that decelerates the applied rotation and outputs the decelerated rotation to the right wheel side is configured.

By the way, the transmission capacity variable control type torque transmission mechanism 90 to which the output of the acceleration / deceleration mechanism 70 is input is provided.
That is, the left clutch 90L and the right clutch 90R are
As shown in the figure, the differential gear 51 is installed in a space on the right wheel side of the acceleration / deceleration mechanism 70 in the differential carrier 51. These hydraulic multi-plate clutches 90L, 90R are connected to clutch plates 90AL, 90AR, which are connected to a clutch case 92 so as to rotate integrally with the right wheel side rotation shaft 68, and to rotate integrally with the intermediate shafts 74, 76. Clutch plates 90BR, 90B
L, and two pistons (not shown) for applying hydraulic pressure (clutch pressure) to the clutches 90L and 90R, respectively. The drive hydraulic pressures of the two hydraulic pistons are adjusted through the hydraulic unit 38 by electronic control of the controller 42. The engagement state of the clutches 90L and 90R is adjusted.

The left clutch 90L has a right wheel side clutch plate 90AL which rotates integrally with the right wheel side rotation shaft 68, and an intermediate shaft 76.
And a clutch plate 90BL on the output side of the speed reduction mechanism 70B that is coupled so as to rotate integrally. Since the clutch plate 90BL rotates integrally with the gear 82A reduced by the reduction mechanism 70B together with the intermediate shaft 76, the clutch plate 90B is rotated unless the speed ratio of the left wheel to the right wheel increases.
L rotates at a lower speed than the right wheel side clutch plate 90AL that rotates integrally with the right wheel side rotation shaft 68.

Therefore, if the clutch 90L is engaged, even if the right wheel is rotating at a lower speed than the left wheel during a right turn, the clutch plate 90AL is
The torque is transmitted to the BL, that is, from the right wheel to the input side of the rear differential, and the amount of torque from the engine to the right wheel is reduced, and the amount of torque to the left wheel is increased. Can be done.

The right clutch 90R has a right wheel side clutch plate 90AR which rotates integrally with the right wheel side rotation shaft 68, and a speed increasing mechanism 70A which is connected so as to rotate integrally with the intermediate shaft 74.
And an output side clutch plate 90BR. The clutch plate 90BR includes the intermediate shaft 74 and the speed increasing mechanism 70A.
As a result, the clutch plate 90BR rotates at a higher speed than the right wheel side clutch plate 90AR which rotates integrally with the right wheel side rotation shaft 68 unless the speed ratio of the right wheel to the left wheel increases. I do.

Therefore, if the clutch 90R is engaged, even if the left wheel is rotating at a lower speed than the right wheel during a left turn, the clutch plate 90BR is shifted from the clutch plate 90BR side to the right wheel side clutch plate 90R.
The torque is transmitted to the AR side, that is, from the input portion side of the rear differential to the right wheel side, so that the amount of torque from the engine to the right wheel side is increased, and the amount of distribution to the left wheel side is increased. Can be reduced.

Therefore, the hydraulic pressure adjusting units for the clutches 90L and 90R in the hydraulic unit 38 are also controlled through the ECU 42 so that the torque distribution to the right and left rear wheels is in a desired state. In this case, the ECU 42 controls required parts of the hydraulic unit 38 based on engine information, wheel speed information, steering wheel angle information (steering angle information), information on the lateral acceleration and longitudinal acceleration of the vehicle body, and the like.

For example, when it is desired to distribute the drive torque from the input shaft 52 to the left wheel rotating shaft 66 more, the left clutch 90L can be engaged with an appropriate control pressure in accordance with the degree of distribution (distribution ratio). When it is desired to distribute the drive torque from the input shaft 52 to the right wheel rotating shaft 68 more, the right clutch 90R may be engaged with an appropriate control pressure in accordance with the degree of distribution (distribution ratio).

Also, the clutches 90L and 90R are set so that they are not completely engaged at the same time.
When one of the left and right clutches 90L, 90R is completely engaged, the other is not engaged. That is, the operation mode of the clutches 90L and 90R is
There is a mode in which only the right clutch 90R is engaged, and a neutral mode in which neither is engaged.

As described above, in the torque transfer mechanism 50, the distribution of the left and right torques can be adjusted by moving the torque. Therefore, the torque loss is extremely reduced as compared with the case where the distribution of the right and left torques is adjusted by simply braking one wheel. There is a feature that the torque distribution can be adjusted over a wider range, and a yaw moment can be generated in the vehicle without discomfort, for example.

1.2.3 Configuration of Hydraulic Unit According to the Present Apparatus Here, the configuration of the hydraulic unit 38 will be described with reference to FIG. As shown in FIG. 3, the hydraulic unit 38 includes a pressure accumulating unit 101 for accumulating hydraulic oil and a hydraulic oil accumulated in the pressure accumulating unit 101 by appropriately adjusting the pressure of the clutch 90L, 90R.
And a control pressure output unit 102 for supplying the oil pressure to an oil chamber (not shown).

The accumulator 101 includes an accumulator 103
And a motor pump 104 for pressurizing the working oil in the accumulator 103 to a predetermined pressure, and a pressure switch 105 for monitoring the differential oil pressure pressurized by the motor pump 104. Further, the control pressure output unit 102 is connected to the motor pump 1
An electromagnetic proportional pressure control valve (abbreviated as a proportional valve) 106 for adjusting the pressure of the hydraulic oil in the accumulator 103 whose pressure has been adjusted through the pressure valve 04, and a hydraulic oil adjusted by the proportional valve 106 are connected to either the left or right clutch 90L, 90R. And an electromagnetic directional control valve (directional switching valve) 107 for switching between supply to an oil chamber (not shown).

The hydraulic unit 38 is provided with the ECU 4
2, the ECU 42 includes a wheel speed sensor (wheel speed detecting means) 48A, a steering wheel angle sensor (ie, steering wheel turning angle detecting means for detecting a turning angle of the steering wheel) 48B, and a longitudinal acceleration sensor (front and rear). G sensor) 4
8C, a lateral acceleration sensor (lateral G sensor) 48D, a throttle position sensor (TPS) 48E, and sensors such as a pressure switch 105 are connected.

The ECU 42 controls the motor pump 104 and the proportional valve of the hydraulic unit 38 in accordance with the running state of the vehicle, that is, the vehicle speed, the steering state, the motion state of the vehicle body, etc., based on the information from these sensors. 106 and the direction switching valve 107 are controlled. This proportional valve 1
06 and differential limiting control through the control of the direction switching valve 107,
That is, the details of the torque transfer control will be described later.

In FIG. 3, reference numeral 108 denotes a battery,
Reference numeral 09 denotes a motor relay, which controls the motor pump 104 by controlling the supply of electric power from the battery 108 through the motor relay 109, and manages the accumulated pressure by the accumulator 101 based on the detection information of the pressure switch 105. The operation of the motor pump 104 is controlled through the relay 109 while being controlled. Reference numeral 110 denotes an indicator lamp that indicates whether or not differential limiting control by the hydraulic unit 38, that is, torque movement control is being performed.

Since the differential limiting control through the hydraulic unit 38 needs to be linked with the engine output control, the ECU 42 outputs a control command to the hydraulic unit 38 and controls the engine output control. The output reduction information is also sent to an engine ECU (not shown). The ECU 42 includes a CPU (not shown), a ROM, a RAM,
It has an interface.

1.3 Outline of Control of the Present Apparatus Here, an outline of control of the present apparatus will be described with reference to a control block diagram showing a functional configuration related to control of the present apparatus in FIG. As shown in FIG. 4, the process according to the present control includes a sensor input process for receiving a sensor input, a calculation process for calculating various values based on these sensor input values, and a motion of the vehicle based on the calculation process result. Control amount calculation processing for calculating each control amount of control and drive processing for driving each actuator based on each calculated control amount can be divided.

Of these, in the sensor input processing, four wheel speed sensors 48A, steering wheel angle sensors 48B, longitudinal acceleration sensors (longitudinal G sensors) 48C, lateral acceleration sensors (lateral G sensors) 48D, and throttle position sensors (TP)
S) Receive a sensor input from 48E or the like. In the arithmetic processing, an actual measurement value and a theoretical value of a speed difference (rotational speed difference) between the left and right wheels of the rear wheel are calculated. The actual measurement value (actual rotation speed difference) is based on the wheel speed values from the four wheel speed sensors 48A, and the theoretical value (target value, theoretical rotation speed difference) is based on the steering angle from the steering wheel angle sensor 48B and 4 Each is calculated based on the vehicle speed (vehicle speed) obtained from the wheel speed value from the wheel speed sensor 48A of the wheel. Further, the calculated before and after G (gb) and calculated lateral G (gy) are calculated based on the detection values from the front and rear G sensor 48C and the lateral G sensor 48D. In the arithmetic processing, drift determination and road surface μ estimation are further performed.

In the control amount calculating process, each control amount of the motion control of the vehicle is calculated based on each calculation result.
In this control, the control amount (target ΔN follow-up control amount) of the target rotation speed difference follow-up control (target ΔN follow-up control) related to the control during normal turning, and the control amount (acceleration turn control amount) of the acceleration turning control during the acceleration turn And the control amount of the tuck-in corresponding control at the time of tack-in of the vehicle (tack-in corresponding control amount)
And a control amount (steering transient response control amount) of the steering transient response control relating to the steering transition time are provided, and the respective control amounts are added to output the added control amount (output control amount). ing.

Among the functions for performing the control amount calculation process (control amount calculation means), the function for setting the control amount related to the target ΔN follow-up control (or the function for setting the control amount and the control amount obtained by this setting). The control signal output function) is controlled based on the ΔN tracking control means (target rotation speed difference tracking control means), the function for setting the control amount of the acceleration turning control (or the setting function and the control amount obtained by this setting). A function of outputting a signal) and a function of setting a control amount of the tack-in correspondence control (or a function of outputting a control signal based on this setting function and a control amount obtained by this setting) by a tack-in correspondence control. Means, a function for setting a control amount (transient control amount) of the steering transient response control relating to the steering transition (or a control amount based on this setting function and the control amount obtained by this setting). The control signal output function) is referred to as steering transient response control means.

Further, a means for calculating a control amount (vehicle behavior corresponding control amount) corresponding to the vehicle behavior, such as a target rotation speed difference follow-up control means, an acceleration turning control means, and a tack-in correspondence control means, is provided. Means for calculating a control amount (transient control amount) based on a driving operation state such as a steering wheel operation or a throttle operation is also referred to as a transient control amount means.

The target ΔN tracking control has a function (target value calculating means) for calculating or storing a target value of a yaw angle or a difference between left and right wheel rotational speeds corresponding to a turning state of the vehicle. It has a function of calculating a control amount according to the target value at the time (steady turning control means). In the drive processing, a proportional valve output for outputting a command signal to the proportional valve 106 to adjust the amount of torque movement, and a command signal to a directional valve (direction switching valve) 107 for setting the direction of torque movement. A direction valve output and an indicator display output for outputting a display command signal to the indicator lamp 110 are performed.

2. Here, the contents of the above-described torque control will be further described in the order of input arithmetic processing, drift determination logic, vehicle motion control logic, road surface μ estimation, and actuator drive. 2.1 Input Calculation Processing In the input calculation processing, as shown in FIG.
rl, rear right wheel speed vrr, steering wheel angle θh, vehicle body speed vb, steering wheel angular speed dθh, front left wheel speed vfl,
A detection signal relating to the front right wheel speed vfr is received from each sensor, and a previous calculated value (torque movement amount ta, road surface μ determination coefficient γ) and a pressure switch, an idle switch,
In response to detection signals from the lateral G sensor, TPS (throttle position sensor) and the like, the following numerical processing is performed.

2.1.1 Rear wheel left / right speed difference (dvrd) First, the rear wheel left / right speed difference (dvrd) is calculated by the actual rear wheel left / right speed difference dvr generated during turning and torque transfer control.
d (= vrl−vrr), which is calculated by calculating the difference between the rear left wheel speed vrl and the rear right wheel speed vrr. 2.1.2 Digital filter value (d
vrf) Since the actual speed difference dvrd is used to determine the operating state of the torque transfer control, the actual speed difference dvrd is filtered by a digital filter to remove noise effects. Here, filter processing is performed as shown in equation (2.1.2.2) for performing smoothing processing as in equation (2.1.2.1).

Dvrf1 = (dvrd + odvrd) / 2 (2.1.2.1) dvrf = LPF [dvrd] = LPF [dvrf1, dvrf] (2.1.2.2) where dvrd: holds the previous dvrd Value dvrfl: Smoothed value 2.1.3 Average rear wheel speed (vr) By averaging rear left wheel speed vrl and rear right wheel speed vrr, average rear wheel speed vr [= (vrl + vrr)
/ 2] to determine the operating state of the torque transfer control.

2.1.4 Estimated vehicle speed (vb), turning radius (RR) The present device has a function of estimating the vehicle speed by calculation (vehicle speed calculating device or vehicle speed detecting means). In the speed calculation device (vehicle speed detection means), the estimated vehicle speed vb
Is basically calculated on the basis of the third highest wheel speed v3 of the four left and right front wheels and the right and left rear wheels.

Since the present vehicle is a four-wheel drive vehicle, each wheel becomes a drive wheel. When such a drive wheel transmits a driving force to the road surface, a slip occurs between the drive wheel and the road surface. If the vehicle body speed is obtained, even if it is slight, the value becomes faster than the actual vehicle speed. Therefore, the slowest wheel speed among the four drive wheels most corresponds to the actual vehicle speed. However, it is conceivable that the detected value of the wheel speed may not be an appropriate value due to noise or the like. Therefore, in consideration of the reliability of the detected value, the second slowest wheel speed (ie, the third wheel speed) among the four drive wheels is considered. Thus, the estimated vehicle speed vb is obtained by adopting the faster wheel speed v3.

When the vehicle is traveling straight, the wheel speed and the vehicle speed correspond at a fixed ratio. For example, the vehicle speed (simple calculated vehicle speed) v obtained by multiplying the wheel rotation speed by the wheel outer peripheral length is obtained.
bd can be the vehicle speed vb. Therefore, the present apparatus has a function of calculating the estimated vehicle body speed vb from the third wheel speed (ie, the third-highest wheel speed) v3 (straight vehicle speed estimation means).

However, at the time of turning, the wheel speed changes between the turning inner wheel and the turning outer wheel, and such a change in the wheel speed between the inner wheel and the outer wheel differs depending on the turning radius and the vehicle speed. For this reason, at the time of turning, correction according to the turning radius and the like is required. That is, at the time of turning, the third highest wheel speed is the inner wheel of the rear wheel, and the inner wheel side is considered to be the simply calculated vehicle speed vbd.
Obtained from geometric relationships.

Therefore, the third wheel speed vbd (= v3) as the vehicle speed (simple calculated vehicle speed) estimated (calculated) by the above-mentioned straight-ahead vehicle speed estimating means, and a steering wheel angle sensor (a steering wheel turning angle detecting means) From the steering wheel angle (steering wheel turning angle) θh detected at 48B and a constant unique to the vehicle body, that is, the vehicle wheelbase, tread width, stability factor, steering wheel gear ratio, etc., the vehicle width during turning of the vehicle The vehicle speed at the center of the direction is calculated and estimated. The function of calculating the vehicle speed at the center in the vehicle width direction at the time of turning is referred to as turning vehicle speed calculating means.

That is, the turning radius RRi on the inner wheel side can be calculated by the following equation (2.1.3.1) based on the vehicle speed vbd on the inner wheel side.

RRi = (1 + A * vbd ² ) * Lw / δ = (1 + A * vbd ² ) * Lw * GR / θh (2.1.3.1) where δ: actual steering angle (= θh / GR) [ A: Stability factor Lw: Wheel base Lt: Tread width GR: Handle gear ratio The ratio between the vehicle body speed vbd and the vehicle body speed vb is determined by the turning radius RRi on the inner wheel side. And the turning radius RR at the center of the vehicle body, and the turning radius RR is calculated using the turning radius RRi by the following equation (2.1.
4.1), the vehicle speed vb is divided into the following formula (2.1.
As in 4.2) to (2.1.4.4), it can be obtained from the vehicle speed vbd and the steering wheel angle θh.

RR = RRi + Lt / 2 (2.1.4.1) Right turn vb = (RRi + Lt / 2) / RRi * vbd (2.1.4.2) Straight ahead vb = vbd (2.1. 4.3) When turning left vb = (RRi−Lt / 2) / RRi * vbd (2.1.4.4) Note that the turning radius RR at the center of the vehicle body is such a vehicle speed vb
Based on the following equation (2.1.4.5) can be expressed.

RR = (1 + A * vb ² ) * Lw / δ = (1 + A * vb ² ) * Lw * GR / θh (2.1.4.5) Further, a wheel related to the third wheel speed v3 (= vbd) Is greatly slipped, the third wheel speed v3 deviates significantly from the actual vehicle speed. In such a case,
Since the wheel acceleration of the wheel at the third wheel speed v3 greatly differs from the actual longitudinal acceleration (longitudinal G) of the vehicle, the comparison between the actual wheel acceleration and the actual wheel acceleration results in a large wheel acceleration. Slips and the third wheel speed v3 (= vb
It can be determined that d) cannot be adopted as the vehicle speed.

Therefore, in the present vehicle speed calculation apparatus, the longitudinal G by the longitudinal acceleration sensor 48C provided on the vehicle and the third G
Based on the wheel acceleration d (v3) / dt of the wheel at the wheel speed, it is determined whether or not the wheel is largely slipping.
When determining such a slip, the longitudinal acceleration sensor 48
The vehicle speed is estimated using the front and rear G by C, and the vehicle speed based on the front and rear G is adopted instead of the vehicle speed based on the wheel speed.

When a slip occurs in the tire, the wheel speed v3
Rapidly increases, and the wheel acceleration d (v3) /
If the difference between dt and the longitudinal G by the longitudinal acceleration sensor 48C becomes larger than a predetermined amount, it is considered that the tire is slipping in a non-linear region. At this time, the estimated front and rear G vehicle body speed vbs is calculated, and the estimated front and rear G vehicle body speed vbs is used instead of the estimated vehicle body speed (wheel speed corresponding vehicle speed) vbd (= v3) based on the wheel speed v3.

The vehicle speed will be described later (item,
2.2.2) is calculated as in the following equation. vbs == gx _SL · t + vb _SL where vb _SL is the vehicle speed vb _SL when the tire slips. gx _SL is the front and back G detected when the tire slips. Due to the increase, the difference between the wheel speed v3 and the estimated front and rear G vehicle speed vbs, that is, the longitudinal slip coefficient dvvbs of the tire described later increases, but as the slip converges, the wheel speed v3 decreases and the front and rear G estimated. As it approaches the vehicle speed vbs,
The longitudinal slip coefficient dvvbs of the tire decreases.

Therefore, the tire longitudinal slip coefficient dvvb
Based on s, it is possible to estimate the slip state of the tire, that is, whether the tire is in a linear region where the tire is not slipping or in a nonlinear region where the tire is slipping. Here, when the longitudinal slip coefficient dvvbs converges below a certain value, the tire returns to the linear region where the tire does not slip.
From the adoption of the estimated vehicle speed vbs, the vehicle speed is returned to the estimated vehicle speed (vehicle speed corresponding to wheel speed) vbd (= v3) based on the wheel speed v3.

2.1.5 Back-and-forth acceleration (gx) First, as shown in the following equation (2.1.5.1), the acceleration is calculated from a change in the simply calculated vehicle speed vbd calculated at a predetermined cycle. Since the acceleration gxd fluctuates greatly, it is processed by a low-pass filter [see (2.1.5.2)] to obtain the longitudinal acceleration gx.

Gxd = vbd-ovbd (2.1.5.1) where ovbd is a simple calculated vehicle speed vbd before one cycle or a predetermined cycle gx = LPF [gxd] (2.1.5.2) 2.1. 6. Reference lateral acceleration (gy) The reference lateral acceleration (gy) can be calculated from the radius RRi and the estimated vehicle speed vb, assuming the centrifugal force acting on the vehicle when turning, and the radius RRi is calculated from the steering wheel angle θh as described above. Therefore, the reference lateral acceleration (gy) can be calculated by the following equation (2.1.6.1).
Is calculated from the steering wheel angle θh and the estimated vehicle speed vb. This reference lateral acceleration (gy) is also referred to as a calculated lateral G.

Gy = vb ² / RR = vb ² * θh / [(1 + A * vb ² ) * Lw * GR] (2.1.6.1) 2.1.7 Rear wheel reference rotational speed difference (dvhf) The wheel reference rotation speed difference dvhf is determined by a turning radius RR when turning.
The rotational speed difference of the rear wheel can be geometrically calculated from the relationship shown in FIG. 6 according to the relationship shown in FIG. 6. First, using the relationship of Expression (2.1.4.5), the following expression (2.1.7.1) is used. Estimated vehicle speed v
b, rotational speed difference dvhr as a function of steering wheel angle θh
Ask for. The above-mentioned rear left wheel speed vrl and rear right wheel speed vrr have been subjected to low-pass filter processing, and in order to match the phases thereof, the rotation speed difference dvhr is processed by a low-pass filter [see (2.1.7.2)]. , The rear wheel reference rotational speed difference dvrf is obtained. The function of calculating the rear wheel reference rotation speed difference dvrf is referred to as target value calculation means.

Dvhr = Lt * vb / RR = Lt * vb * θh / [(1 + A * vb ² ) * Lw * GR] (2.1.7.1) dvhf = LPF [dvhr] (2.1.7.2) 2.1.8 Front Wheel Reference Rotation Speed Difference (dvhff) The front wheel reference rotation speed difference dvrfff is determined by the turning radius R at the time of turning.
The rotational speed difference of the front wheels can be calculated geometrically from the relationship shown in FIG. 6 according to R and the steering angle δ.
First, as shown in the following equation (2.1.8.1), a rotation speed difference dvh is obtained from a function of the estimated vehicle speed vb and the steering wheel angle θh, and this is processed by a low-pass filter [ (See 2.1.8.2)], front wheel reference rotational speed difference dvrf
Get f.

Dvrf = Lt * vb * cos (θh / GR) / RR = Lt * vb * cos (θh / GR) * [θh / [(1 + A * vb ² ) * Lw * GR] (2.1. 8.1) dvrfff = LPF [dvrf] (2.1.8.2) 2.1.9 Front wheel left / right speed difference (dvfd) The difference between front left wheel speed vfl and front right wheel speed vfr is calculated, and the vehicle is turning. And the actual speed difference dvf between the left and right rear wheels
d (= vfl-vfr) is obtained.

2.1.10 Amount of Torque Movement (taf: Temporary Delay Value) Since the torque movement has a time delay from when its command value is output until it appears as actual vehicle behavior, the torque movement command value A low-pass filter is applied to ta to adjust the phases (see (2.1.10.1)) to obtain a torque movement amount taf. taf = LPF [ta] (2.1.10.1) 2.2 Drift determination logic In this control, it is determined whether or not the vehicle is about to drift, and the result of this determination is determined by the vehicle through the torque movement control of the left and right wheels. It is used for motion control. Therefore, in this control,
Drift determination is performed by each processing as shown in FIG. The function of determining whether the vehicle is in the drift state or the non-drift state is referred to as drift determination means (turning state determination means).

In other words, the present drift determination means determines that drift occurs when the tire slips or slips. Tire sideslip can be determined when the relationship between the calculated lateral G and the actual lateral G becomes non-linear, and the tire longitudinal slip is determined by the estimated vehicle speed vb and the longitudinal G estimated vehicle speed v described later.
It can be determined when the relationship with bs becomes nonlinear. Normally, when the vehicle drifts, skidding or vertical skidding is involved, and therefore, in this control, drift determination is performed in consideration of both.

2.2.1 Tire Side Slip Coefficient (dgy) In this control, as described above, the calculated lateral G, that is, the reference lateral acceleration gy is calculated from the steering wheel angle θh and the estimated vehicle speed vb. On the other hand, the actual lateral acceleration (actual lateral G) rgy is detected by the lateral G sensor. When the vehicle is traveling without skidding, the relationship between the calculated lateral G and the actual lateral G becomes linear. Therefore, in order to make a drift determination,
The calculated horizontal G and the actual horizontal G are compared.

However, since the calculated lateral G (gy) is calculated from the input information such as the steering wheel angle θh, a phase delay occurs before the lateral G occurs in the vehicle according to the steering wheel. In the control, the calculated horizontal G is filtered by a low-pass filter to perform phase adjustment (see (2.2.1.1)). gyf = LPF [gy] (2.2.1.1) Also, due to the influence of tires and differences in gear ratio, etc., an error exists between the calculated lateral G (gy) and the actual lateral G (rgy) even in the linear region. Therefore, the actual horizontal G (rgy) is corrected by the coefficient k as shown in the following equation (2.2.1.2) to perform coefficient adjustment.

Rgyh = k * rgy (2.2.1.2) Thus, it is possible to compare the calculated horizontal G (gyf) with the matched phase and the actual horizontal G (rgyh) with the matched coefficient. Then, the calculated horizontal G calculated by the following equation (2.2.1.3)
(Gyf) and the actual lateral G (rgyh) based on the dimensionless value (tire slip coefficient) dgy, which is non-linear between the calculated lateral G and the actual lateral G, ie, occurs in the lateral direction of the tire. Determine non-linearity.

FIG. 8 is a diagram showing an example of the correspondence between the actual lateral G (rgy) and the calculated lateral G. If there is no side slip of the tire, the actual lateral G (rgy) and the computed lateral G Is a linear relationship, but actually, when the tire exceeds the grip limit of the tire, a skid or the like occurs, and the actual lateral G does not increase like the calculated lateral G. On a high μ road, the linearity is maintained up to an area where the lateral G is high as shown by a curve B, but on a low μ road, the linearity cannot be maintained in an area where the lateral G is low as shown by a curve C.

The tire side slip coefficient dgy is calculated by the following equation (2.2.
Define as 1.3). dgy = | (gyf-rgyh) / rgyh | (2.2.1.3) However, in the calculation of the side slip coefficient dgy of such a tire, a calculation start condition such as the following equation (2.2.1.4) and A clear condition such as the following equation (2.2.1.5) is provided. This is due to the size of the actual horizontal G (rgyh), the calculated horizontal G and the actual horizontal G
If the magnitude of the difference (gyf-rgyh) does not increase beyond a certain value, there is no danger that the vehicle will drift. In such a case, the calculation of the side slip coefficient dgy is not performed, and the calculation frequency is reduced. It is decreasing.

| Rgyh | <a [m / s ² ] and | gyf−rgyh | <b [m / s ² ] where a and b are constants, and dgy = 0. (2.2.1.4) In general, the actual lateral G does not increase like the computed lateral G beyond the linear region between the computed lateral G and the computed lateral G.
1.3) can be transformed as follows.

Gyf = (1 + dgy) rgyh (2.2.1.3.a) When the linear region is escaped, dgy gradually increases from 0, and the relation of the above equation (2.2.1.3.a) is For example, it can be shown as a straight line D in FIG. Therefore, theoretically, it can be determined that the linearity is broken when the sideslip coefficient dgy becomes a value other than 0. However, in practice, even if the actual horizontal G and the calculated horizontal G are phase-matched and coefficient-matched, perfect matching is always achieved. Since it is difficult to perform the operation, the slip coefficient dgy often occurs (besides 0) even in the actual linear region. Therefore, in the present control, as shown in FIG. 9, if the sideslip coefficient dgy is equal to or less than the first predetermined value (dgy1), the area is a linear area, and if the sideslip coefficient dgy is equal to or more than the second predetermined value (dgy2), the area is a completely nonlinear area. The side slip coefficient dgy is the first
When the value is between the predetermined value and the second predetermined value, it is assumed that the degree of nonlinearity increases as the value approaches the second predetermined value.

2.2.2 Tire coefficient of longitudinal slip (dvvb)
s) In this control, as described above, the estimated vehicle speed vb is calculated based on the third highest wheel speed v3 of the four wheels, but if the tire slips significantly, the vehicle speed based on such wheel speed v3 is calculated. vb becomes larger than the actual vehicle speed. Therefore, when the occurrence of tire slip is estimated, based on the vehicle speed at this time and the front and rear G instead of the wheel speed, the front and rear G
The estimated vehicle speed vbs is calculated.

The estimated vehicle speed vbs before and after G is
Based on the front and rear G of the vehicle detected by the sensor, the vehicle speed vb _SL when the tire slips and the detected value of the front and rear G (gx) _SL are calculated by the following equation (2.2.2.1). Note that t
Is the elapsed time since the slip occurred, and the wheel speed (for example,
It can be estimated that a slip has occurred when the third fastest wheel speed v3) increases sharply.

Vbs == gx _SL · t + vb _SL (2.2.2.1) The longitudinal slip coefficient dvvbs of the tire is the front and rear G estimated vehicle speed vbs calculated as described above, and the third detected simultaneously with this. The following equation (2.2.2.
The calculated value dvvbsd is taken into account in consideration of the noise effect and the like, and this is further filtered with a low-pass filter [see (2.2.2.3)] to determine the tire longitudinal slip coefficient dvvbs.

Dvvbsd = v3−vbs (2.2.2.2) dvvbs = LPF [dvvbsd] (2.2.2.3) As for the estimated front / rear G vehicle body speed vbs, for example, the actual vehicle speed VR is substantially constant. When the vehicle slips on an extremely low μ road and slips on the tires, and then the slips converge, the wheel speed v
3 rapidly increases, and the estimated longitudinal G vehicle speed vbs is calculated.

Immediately after the occurrence of the slip, the wheel speed v3
Increases, the difference between the wheel speed v3 and the estimated front and rear G vehicle body speed vbs, that is, the tire longitudinal slip coefficient dvvbs increases. When the slip converges, the wheel speed v
3 decreases and approaches the estimated front and rear G vehicle speed vbs, so that the longitudinal slip coefficient dvvbs of the tire decreases. Therefore, based on the longitudinal slip coefficient dvvbs of the tire, it is possible to estimate the slip state of the tire, that is, whether the tire is in the linear region where the tire is not slipping or in the nonlinear region where the tire is slipping.

Therefore, in theory, the longitudinal slip coefficient dvvbs
If becomes non-zero, it can be determined that it has become nonlinear,
Actually, the slip occurrence estimation and the longitudinal G estimated vehicle speed v
Since an error also occurs in the estimation of bs, in this control, FIG.
As shown in the figure, the longitudinal slip coefficient dvvbs is equal to the first predetermined value (d
vvbs1) if less than, linear region, longitudinal slip coefficient dvvb
If s is equal to or larger than a second predetermined value (dvvbs2), the longitudinal slip coefficient dvvbs is set to the first predetermined value and the second predetermined value.
When the value is between the predetermined value and the second predetermined value, the degree of nonlinearity is assumed to increase.

2.2.3 Drift determination coefficient (srp) In the present apparatus, drift determination is performed in consideration of both the side slip coefficient dgy and the vertical slip coefficient dvvbs as described above.
Therefore, a value obtained by combining the sideslip coefficient dgy and the longitudinal slip coefficient dvvbs (this is referred to as a drift determination coefficient) srp (= srpd ² ) is calculated by the following equation (2.2.3.1).
Used for drift judgment.

Srp = (a · dgy) ² + (b · dvvbs) ² (2.2.3.1) where a and b are coefficient adjustments to make a circle. This drift determination coefficient srp is shown in FIG. It can be evaluated by such a drift determination circle (friction circle). FIG. 11 shows values (a
Dgy) and the value (b · dvvbs) obtained by adjusting the coefficient of longitudinal slip dvvbs (b · dvvbs) are shown on the abscissa and the ordinate, respectively, and the orthogonal coordinates are shown.

The drift determination circle is a circle centered on the origin of such coordinates, and includes a first radius r ₁ and a second radius r ₂ (r ₁ <r ₂ ). Then, the linear region within a circle having a radius r ₁ (area where the tire is not slipping), the circle outside the radius r ₁ non-linear region (the tire is slipping) and the radius r ₂ of the non-linear region The outside of the circle is set as the drift region.

That is, if the square root (srp ^1/2 ) of the drift determination coefficient srp is within the radius r ₁ (ie, srp ^1/2 ≦ r ₁ ), the linear region is established, and srp ^1/2 is larger than the radius r _1. (That is,
If (srp ^1/2 > r ₁ ), the nonlinear region, and further, srp ^{1/2
If 1/2} is larger than the radius r ₂ (that is, srp ^1/2 > r ₂ ), it is determined to be in the drift region. In the non-linear region, the region of r ₁ <srp ^1/2 ≦ r ₂ is not completely drifted, but has a tendency to drift to a degree corresponding to the drift determination coefficient srp.

For example, FIG. 12 shows the drift determination coefficient srp
Intended to show the drift determination for, srp is the radius r ₁ ² or less (i.e., srp ≦ r ₁ ²⁾ the linear region if,
srp than the radius r ₂ ² large (i.e., srp> r ₂ ²⁾ drift region if, srp is ^{_{r 1 2 <srp ≦ r 2}} 2 regions is not a full drift, drift determination coefficient sr
It is assumed that the degree of drift corresponds to p.

(Drift Corresponding Control Start Condition) If the drift determination coefficient srp is equal to or greater than a predetermined value, and if the countersteer is turned off and the steering wheel angular velocity of the countersteer is higher than a predetermined speed, it is determined that the vehicle is drifting (drift determination). Means or turning state determination means). It is determined that the countersteer is turned off when the steering angle exceeds the neutral position, that is, when the direction of the calculated lateral G is opposite to the direction of the actual lateral G. That is, when all of the following three conditions are satisfied at the same time, it is determined that the vehicle is traveling in drift, and the drift-based control (slip-based control) is started. Note that the function of determining the start of such drift correspondence control (slip correspondence control) is referred to as start determination means. The drift determination coefficient srp is equal to or greater than a predetermined value. Srp> srp0 (2.2.6.1) The direction of the calculated lateral G (gy) and the direction of the actual lateral G (rgyh) are opposite gy · rgyh <0 ··· (2.2.6.2) also handle angular velocity Derutashitah is Δθh ≧ Δθ ₁ is at a predetermined speed [Delta] [theta] ₁ or more (deg / s) ··· (2.2.6.3 ), the three equations conditions are simultaneously Even when the condition is not satisfied, it may be determined that the vehicle is drifting when the drift determination coefficient srp is equal to or greater than a predetermined value. Incidentally, the steering wheel angular velocity Δθh, Δθ _1, respectively Dishitah, referred both d [theta] _1.

(Drift-Compatible Control Termination Condition) When the steering angle returns to the neutral position again, that is, when the calculated lateral G direction is equal to the actual lateral G direction, it is determined that drift traveling has been completed. Stop the drift control. In addition, the function of determining the end of the drift corresponding control (slip corresponding control) is referred to as an end determining unit. The direction of the calculated lateral G (gy) and the direction of the actual lateral G (rgyh) are the same direction. Gy · rgyh> 0 (2.2.6.4) Also, it is determined that the vehicle is not drifting by the above conditional expression. Then, the drift determination coefficient srp is set to zero (sr
p = 0).

2.2.6 Turning side G (for drift, gg
y) By the way, in this control, there is a torque movement control based on a lateral acceleration (turning lateral G) applied to the vehicle at the time of turning, and for example, tack-in correspondence control and acceleration turning control correspond to this.
The turning side G corresponds to the calculated side G and the actual side G described above.
When there is no difference between the calculated lateral G and the actual lateral G when the tire is gripping the road surface (during grip traveling), the computed lateral G and the actual lateral G correspond to the behavior of the vehicle. Thus, the calculated lateral G having a higher processing speed than the actual lateral G can be used as the turning lateral G. However, when the vehicle is drifting, a large difference occurs between the calculated lateral G and the actual lateral G. Therefore, the calculated lateral G cannot be used. In this case, the actual lateral G corresponding to the behavior of the vehicle is used as the turning lateral G. Must be used.

Therefore, the present apparatus normally uses the calculated horizontal G, and uses the calculated horizontal G when the calculated horizontal G cannot correspond to the actual situation. For this reason, when it is determined that the vehicle is drifting under the drift-dependent control start condition, the turning lateral G
Then, the mode is switched from the use of the calculated lateral G to the use of the actual lateral G, and if it is determined that the drift traveling has been completed in the drift corresponding control end condition, the setting is made to return from the adoption of the actual lateral G to the use of the calculated lateral G. ing.

The selection of the horizontal G is performed by selecting the horizontal G selection coefficient dor.
i, and when calculation horizontal G is selected, dori = 0,
When the actual horizontal G is selected, dori = dori1 (constant). The turning side G: ggy corresponding to the drift can be expressed by the following equation by the side G selection coefficient dori. ggy = [dori · rgyh + (dori1-dor)
i) .gy] / dori1 where gy: calculated lateral G, rgyh: actual lateral G. Further, an example of selecting the turning lateral G for such drift will be described with reference to FIG. In FIG. 13, the solid line indicates the calculated horizontal G (gy), the broken line indicates the real horizontal G (rgyh),
As shown in the figure, when the vehicle turns, a lateral G occurs in the vehicle, and there is no difference between the calculated lateral G and the actual lateral G during the grip traveling. G changes suddenly. The reason why the calculated lateral G changes suddenly is that the driver applies the steering wheel operation in the drift state. When the steering wheel operation is applied, the calculated lateral G is calculated based on the steering wheel angle θh as in the equation (2.1.6.1). That is, the calculated lateral G greatly changes. In particular, when the counter steer is turned off during drift, the calculated lateral G becomes the actual lateral G.
Changes in the opposite direction. The calculation is performed only until the calculated lateral G changes in the direction opposite to the actual lateral G and the calculated lateral G becomes the same direction as the actual lateral G, that is, only during the period indicated as “Drift control” in FIG. Instead of the horizontal G, an actual horizontal G input is adopted.

2.3 Vehicle Motion Control Logic As described above, in the present torque movement control device, the target rotation speed difference follow-up control (the target ΔN follow-up control) is set as the control mode.
, Acceleration turning control, tack-in response control, and steering transient response control, each of which will be described in detail here.

2.3.1 Target ΔN Tracking Control The target ΔN tracking control is a control aiming at both the action as the yaw rate feedback control (yaw rate FBC action) and the action as the LSD (LSD action). Control is performed so as to eliminate the difference between the rear wheel reference rotational speed difference (theoretical value, dvhf) obtained as described above according to 2.1.7.2) and the speed difference between the right and left rear wheels (actual speed difference: dvrd). . Therefore, the dashed block B3 in FIGS.
As shown in 1, a plurality of control amounts (high μ road control amount tbh, low μ road control amount tbl) corresponding to μ are set.

These control amounts are described in, for example, JP-A-7-10
The control amount for the high μ road and the control amount for the low μ road are set with different gain characteristics with respect to the rotational speed difference between the left and right wheels, which is obtained by the method disclosed in Japanese Patent No. 8840.
Further, when drift traveling is determined by the drift determination coefficient srp or the like, the drift value is adjusted by the drift determination coefficient srp1 so that the theoretical value dvhf becomes zero, and control is performed so as to eliminate the rotational speed difference between the left and right wheels.

Further, the control amount for the high μ road and the control amount for the low μ road are gain-adjusted by the drift determination coefficients srp2 and srp3, and the control amount suitable for drift traveling is calculated.

2.3.2 Acceleration Turn Control Acceleration turn control is a control for suppressing an increase in the understeer tendency during a sharp turn as described above. This control is necessary only because the stability of the vehicle is non-linear. This is the case. That is, as shown in FIG.
When the relationship between the turning G and the steering ratio is out of the linear region (see the broken line), the turning radius of the vehicle increases. This is because during a sharp turn, the steering characteristics of the vehicle increase toward the understeer side.

As described above, during a sharp turn, the target ΔN
In the following control, the torque is moved to the turning outer wheel to generate a moment in the turning direction to increase the cornering force of the front wheels. However, the target ΔN following control has a slight reaction delay due to feedback control. Therefore,
At the time of such a sharp turn, acceleration turning control for moving the torque to the turning outer wheel side is performed to generate or increase the yaw moment in the turning direction, and to increase the cornering force of the front wheels in a region where the longitudinal acceleration is large. This is to prevent understeering.

(1) Acceleration turning control amount (teh, tel) In this control, as shown in the block B32 in FIGS. 14 and 15, when the turning lateral G (ggy) is equal to or more than a predetermined value, the acceleration turning control is performed. Are set as basic control amounts tehd and teld. This control amount is output on condition that the tack-in correspondence control is not being performed. FIG. 17 is a map for a high μ road (map for high road frictional resistance), and FIG. 18 is a map for low μ road (map for low road frictional resistance). )), That is, the basic control amount for high μ road (control amount corresponding to high road surface friction resistance) tehd and the basic control amount for low μ road (control amount corresponding to low road surface friction resistance) teld. Set.

As shown in FIGS. 17 and 18, the horizontal G (gg
In the region where y is small, the same turning lateral G (ggy)
In contrast, the map for the low road frictional resistance has a larger control amount than the map for the high road frictional resistance, and in the region where the lateral G (ggy) is large, the high lateral frictional resistance is similar to the turning lateral G (ggy). The control map gives a larger control amount than the low road friction resistance map. Further, as shown in FIGS. 17 and 18, each map is provided with a dead zone region in which the control amount is 0 in a region where the turning lateral G (ggy) is small, thereby stabilizing the control. Note that the chain line in FIG. 17 shows the characteristics of the low road surface frictional resistance map (see FIG. 18).
The chain line in the middle indicates the characteristics of the high road frictional resistance map (see FIG. 17).

In this embodiment, the turning lateral G
When (ggy) is equal to or more than a predetermined value, it is determined that the vehicle is turning sharply. However, even if the turning lateral G (ggy) is detected as small, even if the control amount is calculated so as to increase the rotational propulsion force of the turning outer wheel. Good. When it is determined that the vehicle is drifting, the turning side G
Is used to calculate the control amount.

When it is determined that the vehicle is drifting, the acceleration / turn control amounts teh and tel are gain-adjusted by the drift correction coefficient srp5 to calculate a control amount suitable for drifting. For example, during drift running, the acceleration turning control amount te
You may be comprised so that h and tel may be set to zero.

2.3.3 Tack-in Correspondence Control As described above, when the vehicle is decelerating and turning, the steering characteristic of the vehicle tends to be over-steered as the cornering force of the front wheels increases, contrary to the acceleration turning. Become. As described above, during deceleration turning, the target ΔN
In the follow-up control, the torque is moved to the turning inner wheel side to generate a yaw moment in the turning restraining direction, and thereby,
Although oversteering is suppressed, the target ΔN tracking control has a slight reaction delay due to feedback control.

Therefore, at the time of deceleration turning, tack-in correspondence control for generating or increasing the yaw moment in the turning suppression direction is performed by moving the torque to the turning inner wheel side, and the cornering force of the front wheels is reduced to suppress oversteering. I do. Thus, the turning behavior of the vehicle is controlled to avoid the tuck-in of the vehicle and the spin due to the tuck-in.

[0129] These control amounts are determined, for example, by the method disclosed in JP-A-7-108840.
The control amount tdh for the high μ road and the control amount tdl for the low μ road are obtained. Further, similarly to the acceleration turning control, when it is determined that the vehicle is drifting, the control amount is calculated using the actual side G as the turning side G.

When it is determined that the vehicle is drifting, the control amounts tdh and tdl corresponding to the tack-in are adjusted to the drift correction coefficient sr.
The gain is adjusted by p5 to calculate a control amount suitable for drift traveling. For example, during drift driving, the tack-in control amounts tdh and tdl may be set to zero. Further, in the tack-in correspondence control, similarly to the acceleration turning control, in the area where the turning lateral G (ggy) is small, the low road surface friction resistance map (the low μ road map) is higher than the high road surface friction resistance map (the high road friction resistance map). A larger control amount than (μ road map) may be given. Alternatively, a dead zone is provided in a small area of the turning lateral G (ggy), and the control amount for the high road surface friction resistance map (high μ road map) is larger than the low road surface friction resistance map (low μ road map). May be given.

2.3.4 Steering Transient Response Control Steering transient response control is a control performed at the time of steering transition.
As shown in a block B33 in FIGS. 14 and 15, control is performed so as to be proportional to the change in the steering angle, that is, the steering angular velocity. In other words, the ECU 42 has a function of setting a steering transient response control amount (transient control amount), that is, a means steering transient response control amount setting means (transient control amount calculating means). Control amount (steering angular velocity proportional control amount) t
c can be set. For this reason, first, the basic control amounts tchd, tcld according to the steering angular velocity dθh
And a correction according to the vehicle speed, a correction according to the turning or turning back of the steering wheel, and a drift correction coefficient of 4
Drift correction is performed by (srp4), and control is performed based on the control amounts tch and tcl thus obtained.

[0132] These control amounts are determined, for example, by the method disclosed in Japanese Patent Laid-Open No. 7-108840.
The control amount tch for the high μ road and the control amount tcl for the low μ road are obtained. If it is determined that the vehicle is traveling in drift, the control amounts tch and tcl corresponding to the tack-in are changed to the drift correction coefficient sr
The control amount suitable for drift traveling is calculated by adjusting the gain by p4. For example, during the drift traveling, the tack-in control amounts tch and tcl may be set to zero.

2.4 Estimation of Road Surface μ (Description of Road Surface Friction Coefficient Estimation Apparatus for Vehicle) In the torque transfer control, the control effect differs depending on whether the running road is slippery, that is, the state of the road surface frictional resistance. Therefore, the present device (vehicle power transmission control device between left and right wheels) is provided with a vehicle road surface friction coefficient estimating device, and the vehicle road surface friction coefficient estimating device uses a road surface friction coefficient (e.g., (Hereinafter also referred to as road surface μ).

The road friction coefficient estimating apparatus for a vehicle according to the present invention performs estimation of the road surface μ in three stages: μ estimation at the time of steady turning, μ estimation at the time of starting, and μ estimation at the time of non-linearity. In other words, each of the steady turning, starting and non-linear stages exists in a region shown in FIG. 19 with respect to the turning side G and the vehicle speed. Note that the μ estimation at the start of
Set the initial value for The non-linear time is a case where the vehicle is non-linear with respect to the steering of the steering wheel. Here, in each of these cases, the road surface μ
A determination coefficient (a coefficient of road surface friction, that is, a coefficient representing the degree of road surface μ) γ is obtained, and an output gain value (output value) of each control amount is determined from the road surface μ determination coefficient γ value. The vehicle road surface friction coefficient estimating device is also referred to as a road surface μ estimating device, a road surface μ determining device, or a road surface friction coefficient detecting device and a road surface μ detecting device.

2.4.1 Estimation of Road Surface μ during Steady Turn The optimal value of the amount of torque movement in vehicle motion control differs depending on the level of the road surface μ (road surface friction coefficient). Therefore,
In the vehicle road surface friction coefficient estimating apparatus, a first road surface unevenness state indicating a road surface unevenness state during steady running (particularly, during steady turning) is provided.
And a second parameter indicating the slipperiness of the road surface are detected, and from the values of the first and second parameters, a third parameter that represents the road surface state in a unified manner is obtained. From the index corresponding to the state of the parameter of 3,
The road friction coefficient is estimated.

In the road friction coefficient estimating apparatus, the actual lateral acceleration (actual lateral G) applied to the vehicle is used as the first parameter indicating the unevenness of the road surface. That is, for the actual lateral acceleration (actual lateral G) applied to the vehicle, the lateral acceleration component generated due to the turning of the vehicle is removed (this is referred to as turning correction). The lateral acceleration component of the vehicle that does not need to be removed is removed, and then the absolute value of the lateral acceleration subjected to the turning correction processing and the high-pass processing is taken, and this is further subjected to the low-pass processing to obtain the actual lateral acceleration caused by the road surface unevenness. The amplitude of the component is extracted and the amplitude βp
p is a first parameter (bad road determination index).

Further, in the present road surface friction coefficient estimating apparatus, the ratio of the increase in the slip ratio S to the driving force (torque) is used as the second parameter indicating the ease of slipping on the road surface. In other words, in the grip area of the wheel (tire),
Road surface μ, which is the ratio of increase in slip ratio S to driving force
The determination value α (here, the average value αh of the road surface μ determination value α is used as described later) is used as the second parameter.

The rough road determination index (amplitude after actual lateral acceleration correction) βpp as the first parameter and the second parameter
From the average value αh of the road surface μ determination value α as a parameter, the road surface condition is estimated by fuzzy inference, and the frequency is obtained by a counter (learning function) to determine the road surface μ. In order to perform such processing, the vehicle road surface friction coefficient estimating apparatus has a functional configuration as shown in FIG.

In FIG. 20, reference numeral 200 denotes a parameter calculating means. The parameter calculating means 200 calculates a rough road determination index (first parameter) βpp indicating the unevenness of the road surface from the running state of the vehicle during steady running. First parameter calculating means (road surface unevenness parameter calculating unit) 2
10 and a second parameter calculating means (slippery parameter calculating unit) for calculating an average value αh (second parameter) of the road surface μ determination value α indicating the ease of road surface slippage from the running state during steady running of the vehicle. 220.

Reference numeral 202 denotes a road surface index calculating means.
The road surface index calculating means 202 uses the values βpp and αh of the first and second parameters to obtain a high μ road degree (high μ road), a medium μ The road degree (medium μ road) and the low μ road degree (low μ road) are determined, and these high μ road, middle μ road, and low μ road are determined.
An index corresponding to each state of the road (here, this index is referred to as a fitness level because fuzzy inference is used) is calculated.

Reference numeral 204 denotes a road surface friction coefficient calculating unit. The road surface friction coefficient calculating unit 204 continuously obtains the index calculated by the road surface index calculating unit 202 and performs cumulative evaluation to obtain the road surface friction coefficient ( Road surface μ judgment coefficient γ)
Is calculated. Hereinafter, these parameter calculation means 20
0, road surface index calculating means 202, road surface friction coefficient calculating means 2
04 will be described in detail. Calculation of first parameter (bad road determination index) βpp The first parameter calculation means (road surface unevenness parameter calculation unit) 210 of the parameter calculation means 200, as shown in FIG. A turning correction unit 211 for correcting the turning of the lateral G, and a turning correction unit 211
, A high-pass processing unit 213, an absolute value processing unit 214, and a low-pass processing unit 215.

The turning correction gain calculating section 212 calculates a correction gain kr from a map such as that shown in FIG. 21, for example, according to the calculated lateral G (= gy). That is, if the value of the calculated lateral G is smaller than the first set value gy1, it is determined that the vehicle is in a straight traveling state, and the correction gain kr is set to 1 (that is, the correction gain kr is 1).
Although the area where the value of the calculated lateral G is equal to or less than the first set value gy1 is a dead zone), when the value of the calculated lateral G becomes larger than the first set value gy1, the correction gain kr increases in accordance with the increase in the value of the calculated lateral G. Is reduced, and the value of the calculated horizontal G is changed to the second set value gy2 (>
gy1), the correction gain kr is fixed to the minimum value kr1 so that the correction gain kr does not become zero.

The turning correction section 211 adjusts the gain of the actual lateral G value rgy by the correction gain kr calculated by the turning correction gain calculating section 212 as described above, so that the vehicle The horizontal G component caused by the turning is removed. The high-pass processing unit 213
A high-pass process is performed in which only components having a frequency equal to or higher than a predetermined frequency (for example, 6 Hz) out of the value krgy of the actual lateral G output from the turning correction unit 211 are passed and components having a frequency lower than the predetermined frequency are cut. Road surface irregularities are input to the vehicle body from wheels,
The lateral G generated due to the road surface unevenness is in a certain frequency band due to tire and suspension characteristics. This frequency band can be grasped according to the vehicle. In general, this frequency band is higher than the frequency of the lateral G component due to causes other than the road surface unevenness. Therefore, by grasping such a frequency band, setting a predetermined frequency, and performing high-pass processing on the predetermined frequency or higher, it is possible to extract only the lateral G component caused by the road surface unevenness.

The absolute value processing section 214 calculates the absolute value of the output value of the high-pass processing section 213. This is because it is simplest and most reliable to evaluate the magnitude of the amplitude of the lateral G component caused by the unevenness of the road surface, and therefore the absolute value of the output value of the high-pass processing unit 213 is evaluated. This is for taking the value and evaluating the horizontal G component. The low-pass processing unit 215 performs a low-pass process on the output value from the absolute value processing unit 214 to suppress the fluctuation of the amplitude of the horizontal G component. Since the output value from the absolute value processing unit 214 has a large amplitude fluctuation, it is difficult to perform a stable evaluation if the amplitude is evaluated as it is. Therefore, only components having a predetermined frequency (for example, 0.5 Hz) or less are passed through low-pass processing, and The first parameter (amplitude after actual lateral acceleration correction) βpp
As a result, the evaluation of the horizontal G component can be facilitated. Calculation of second parameter (road surface μ determination value) α In the second parameter calculation means (slippery parameter calculation unit) 220 of the parameter calculation means 200, the road surface μ
The determination value α is calculated. First, the principle of calculating the road surface μ determination value α will be described with reference to FIGS.

First, FIG. 22 is a diagram showing an example of a change in the magnitude of the driving force with respect to the slip ratio S. Here, the slip ratio S is the difference between the wheel speed vw and the vehicle speed vb (v
w-vb) divided by the vehicle speed vb, and is obtained by the following equation. The driving force is a driving force (torque) transmitted from the wheels to the road surface. Slip ratio S = (wheel speed−vehicle speed) / vehicle speed = (vw−vb) / vb (2.4.1.1) In FIG. 22, a curve Hμ represents a characteristic of a high μ road (a road surface having a high friction coefficient). The curve Lμ is a low μ road (a road surface with a low friction coefficient).
The characteristics of As shown in FIG. 22, in each area of the road surface, in a region where the slip ratio is small, the wheels (tires) grip the road surface (grip region), so that the driving force increases substantially linearly with respect to the slip ratio. When the slip ratio increases, the wheels (tires) no longer grip the road surface (slip region), so that even if the slip ratio increases, the driving force decreases rather.

When attention is paid to the increasing ratio of the driving force to the slip ratio in the grip region, that is, the inclinations α _H and α _L of the characteristic lines Hμ and Lμ, the high μ road (Hμ) and the low μ road (L
μ), the inclinations α _H and α _L are different in this grip region. That is, the gradient α _x becomes smaller as the road becomes higher μ, so that the road surface μ can be estimated based on the gradient α _x . Such a gradient α _x is a _value of a ratio between an increase amount of the slip ratio difference (slip ratio difference due to torque transfer) βs and a drive force increase amount (torque transfer amount) when the driving force increases from T1 to T2. Can be expressed as Note that Tm is the amount of torque movement (Tm = T1−T2).

Α _x = βs / Tm (2.4.1.2) The road surface μ determination value α is calculated based on the increase rate of the driving force with respect to the slip ratio difference βs in the grip region, that is, the gradient α _x . This value corresponds to the value after various corrections.
Can be obtained by dividing the slip ratio difference βs by the torque movement amount (temporary delay value of the torque movement amount whose phase has been adjusted) taf as in the following equation.

Α = βs / taf (2.4.1.3) By the way, the rotational speeds of the left and right wheels generally include a speed difference due to the left and right load movement and an inner / outer wheel speed difference at the time of turning. In a wheel provided with a torque movement control device, a speed difference due to the torque movement is added to this. Therefore, when the torque movement control is performed only on the rear wheels as in the present embodiment, as shown in FIG. 23, the difference between the left and right wheel rotational speeds of the front wheels is the sum of the difference due to the load movement and the difference due to the turning. On the other hand, the difference between the rotational speeds of the left and right wheels of the rear wheels is the sum of the difference due to load movement, the difference due to turning, and the difference due to torque movement.

Therefore, the slip ratio difference βs due to torque movement is calculated by subtracting the left-right rotation speed difference of the front wheel from the left-right rotation speed difference of the rear wheel and dividing the difference by the vehicle speed as in the following equation. Can be. βs = (Rear wheel left / right speed difference−Front wheel left / right speed difference) / vehicle speed (2.4.1.4) However, when calculating the slip ratio difference βs in this way, the left and right front wheels are differential. It is premised that the vehicle rotates while allowing the differential through the mechanism. For example, if an LSD is provided between the left and right wheels of the front wheel, the differential rotation of the front wheel left and right is restricted, and the speed difference between the front wheel left and right becomes small, The slip rate difference βs due to the torque movement calculated by the above equation becomes larger than the actual one. Therefore, considering that a mechanism that affects rotation such as LSD is provided between the left and right wheels of the front wheel, it is necessary to calculate the slip ratio difference βs due to torque movement by a method different from the above equation.

Therefore, in the slipperiness parameter calculating unit 220 of the present vehicle road surface friction coefficient estimating device, the actual rotational speed difference between the left and right wheels of the rear wheel having the torque movement control device is determined by Speed difference due to the load movement of
And the inner and outer wheel speed difference during turning, by focusing on the points in which and the speed difference by the torque transfer has been added, from the slip ratio difference beta ₁ based on the actual rotational speed difference between the left and right wheels, due to the load movement of the left and right By subtracting the slip rate difference ΔSb based on the speed difference and the slip rate difference ΔSa based on the inner and outer wheel speed difference during turning, the slip rate difference βs due to the torque movement is obtained.
Is calculated.

The slip ratio difference ΔSa based on the difference between the inner and outer wheel velocities during turning can be obtained from the geometrical characteristics during turning as shown in FIG. That is, the rotational speed difference ΔVr of the rear wheel (= the difference between the wheel speed Vout of the turning outer wheel and the wheel speed Vin of the turning inner wheel), which is generated as the inner and outer wheel speed difference at the time of turning, is
The tread width of the vehicle is Lt, the turning radius is RR, and the vehicle speed (vehicle speed) is vb.

ΔVr = Vout−Vin = vb · (Lt / RR) (2.4.1.5) Actual lateral acceleration (lateral acceleration detection value) generated in the vehicle
Is gy, the following equation is established. RR = vb ² / gy Therefore, ΔVr = Lt · gy / vb (2.4.1.5a) The slip ratio difference ΔSa based on the inner and outer wheel speed difference during turning is:
It can be obtained by the following equation.

ΔSa = ΔVr / vb = Lt · gy / vb ² (2.4.1.6) Further, since the right and left load movement occurs during the normal turning, the speed difference due to the right and left load movement at the time of the turning is obtained. Can be determined from the geometrical characteristics of the load points as shown in FIG. That is, the load movement amount ΔW at the time of turning is such that the load applied to the rear wheel (entire left and right wheels) is Wr, the difference between the center of gravity of the vehicle and the roll height is hs, the tread width of the vehicle is Lt, the gravitational acceleration is g, and The actual lateral acceleration (lateral acceleration detection value) that occurs can be determined by the following equation, as gy.

ΔW = Wr · hs · gy / Lt · g (2.4.1.7) Here, the loads Fzin, F applied to the inner and outer rings during turning
zout can be expressed by the following equation by the load Wr applied to the entire rear wheel and the load movement amount ΔW. Turning inner wheel: Fzin = (1/2) · (Wr−ΔW) (2.4.1.8) Turning outer wheel: Fzout = (1/2) · (Wr + ΔW) (2.4. 1.9) The driving force Fx for the rear wheels can be obtained from the load Wr applied to the entire rear wheels, the actual longitudinal acceleration (detected longitudinal acceleration value) gx generated in the vehicle, and the running resistance C, as in the following equation. it can.

Fx = Wr · gx / (2 · g) + C (2.4.1.10) Here, the load Fz applied to the rear wheel and the driving force Fx for the rear wheel
Since there is a proportional relationship between the slip ratio S and the slip ratio S, it can be assumed that Fx / Fz = a · S (2.4.1.11) The slip ratio difference ΔSb due to the movement can be obtained by the following equation.

ΔSb = Fx / (a · Fzout) −Fx / (a · Fzin) = − (Fx · ΔW) / [a · (Wr ² · ΔW ² ) ≒ − (Fx · ΔW) / (a · Wr) ² ) ∵ΔW ² ≪Wr (2.4.1.12) Here, by substituting equations (2.4.1.7) and (2.4.1.10) into equation (2.4.1.12), ΔSb ≒ −hs · gx · gy / (2a · g ² · Dt) −hs · C · gy / (a · g · Dt · Wr) = − (A · gx · gy + B · gy) A, B: constants (2.4.1.13) In this way, the slip ratio difference ΔSa based on the inner and outer wheel speed difference during turning is based on the speed difference due to the right and left load movement from the vehicle speed vb and the actual lateral acceleration (actual lateral G). The slip ratio difference ΔSb can be calculated from the actual longitudinal acceleration (actual longitudinal G) and the actual lateral acceleration (actual lateral G), respectively.

The slip ratio difference βs due to the torque movement is represented by the slip ratio difference β ₁ based on the rear wheel rotation speed difference as shown in the following equation.
(= Dvrd / vb) can be obtained by correcting the slip rate difference ΔSa based on the inner and outer wheel speed differences during turning and the slip rate difference ΔSb due to load movement during turning. βs = dvrd / vb− (ΔSa + ΔSb) (2.4.1.14) Therefore, the road surface μ determination value α is obtained by the following equation, where the torque movement amount (control value) is taf.

Α = βs / taf = [dvrd / vb− (ΔSa + ΔSb)] / taf = [(dvrd−Dt · gy) / vb + A · gx · gy + B · gy] / taf (2.4.1.15) ) In order to calculate the road surface μ determination value α, the second
Parameter calculation means (slipperiness parameter calculator) 22
0 is a rear wheel left / right actual speed difference calculation unit 221 that calculates an actual speed difference between left and right rear wheels (actual speed difference), as shown in FIG.
An inner / outer wheel speed difference calculating unit 222 for calculating a rear wheel left / right speed difference based on an inner / outer wheel speed difference during turning; and an inner / outer wheel difference correction amount (for correcting a slip ratio difference based on the actual rear wheel left / right speed difference) An inner / outer wheel difference correction amount calculator 223 for calculating a slip ratio difference based on a rear wheel left / right speed difference due to an inner / outer wheel speed difference, and a load movement correction amount for correcting a slip ratio difference based on an actual rear wheel left / right speed difference. (Slip ratio difference based on load movement of left and right wheels during turning) Load movement correction amount calculation unit 224
And a slip ratio difference calculating unit 225 for driving force difference (torque difference)
The road μ determination value α calculation unit 226 and the average value calculation unit 22
7 is provided.

The rear wheel left / right actual speed difference calculation section 221 receives the rear left wheel speed vrl and the rear right wheel speed vrr from the respective wheel speed sensors of the rear wheels, and calculates the difference between them, that is, the actual speed difference dv.
rd (= vrl−vrr) is calculated. The inner / outer wheel speed difference calculation unit 222 calculates the rear wheel rotation speed difference ΔVr (= the difference between the wheel speed Vout of the turning outer wheel and the wheel speed Vin of the turning inner wheel) generated as the inner / outer wheel speed difference during turning (2.4.1.5). According to a), it is calculated from the vehicle speed (vehicle speed) vb and the actual lateral acceleration (detected lateral acceleration value) gy.

The inner / outer wheel difference correction amount calculation unit 223 converts the slip ratio difference ΔSa based on the rear wheel left / right speed difference due to the inner / outer wheel speed difference during turning into a rear wheel rotation speed difference ΔVr generated as the inner / outer wheel speed difference during turning. The vehicle speed vb is calculated by the above (2.4.1.6), and this is output as an inner / outer wheel difference correction amount. The load movement correction amount calculation unit 224 calculates the difference between the left and right slip ratios ΔSb of the rear wheels generated according to the left and right load movements when turning.
Is calculated from the actual longitudinal acceleration (actual longitudinal G) gx and the actual lateral acceleration (actual lateral G) gy by the above equation (2.4.1.13), and this is output as the load movement correction amount.

The slip ratio difference calculator 225 for the driving force difference
Actual speed difference d calculated by rear wheel left / right actual speed difference calculation section 221
vrd (= vrl-vrr), the vehicle speed vb, the slip ratio difference ΔSa as the inner / outer wheel difference correction amount calculated by the inner / outer wheel difference correction amount calculation unit 223, and the load movement correction amount calculation unit 2
The slip ratio difference βs of the driving force difference is calculated from the slip ratio difference ΔSb as the load movement correction amount calculated in 24 by the above equation (2.4.1.14), and is output.

The road μ determination value α calculation unit 226 calculates the slip ratio difference βs of the driving force difference calculated by the slip ratio difference calculation unit 225 of the driving force difference and the above-described torque movement amount (primary delay value) taf. The road surface μ determination value α is calculated by the above equation (2.4.1.15). The average value calculation unit 227 calculates the road surface μ determination value α calculated by the road surface μ determination value α
Is calculated. That is, the road μ determination value α is
The calculated value is calculated at a predetermined cycle. However, if the calculated value is used as it is, it is affected by noise or the like. Therefore, the reliability of the determination is increased by averaging a predetermined number of calculated values.

Here, the latest N (here, N = 12) road surface μs sampled every predetermined period T ₁ second are used.
The average value of the judgment value α is calculated [see (2.4.1.11)]. αh = [α (n-11) + α (n-10) +... + α (n)] / 12 (2.4.1.16) where α (n) is the current period Obtained road surface μ determination value α α (nk): road surface μ obtained k periods before the current period
Judgment value α Note that the average value calculation section 227 calculates and outputs the average value of a predetermined number of nearest road surface μ judgment values α. Of course, the average value αh is updated every predetermined calculation cycle.

FIG. 27 shows the measured data of the road surface μ judgment value α on the high μ road [FIG. 27A] and the low μ road [FIG. The constants A and B for obtaining the slip ratio difference ΔSb due to
2.0e-5 and B = 3.8e-2. As shown, the magnitude of the road surface μ determination value α is a low μ road [FIG.
7 (B)] is clearly larger than the high μ road [FIG. 27 (A)]. Note that each vertical axis in FIGS. 27A and 27B is a rough road determination index (first parameter) βpp calculated by the road surface unevenness parameter calculation unit 210.

Incidentally, reference numeral 230 denotes a road surface μ judgment condition check unit. The road surface μ judgment condition check unit 230 judges whether or not a road surface μ judgment judgment condition described later is satisfied. In the road surface μ determination condition check unit 230, the traveling state of the vehicle detected by various vehicle state detection means, that is, the torque movement amount taf, the road surface μ determination value α, the brake switch information bksw, the steering wheel angular velocity dθh, and the drift determination coefficient The determination is made based on information such as Srp and reference lateral acceleration gy. The road surface μ determination determination condition also includes a condition as to whether the vehicle is traveling in a steady state (here, a steady turning travel).

In this road surface μ judgment condition check unit 230,
If it is determined that the road surface μ determination determination condition is not satisfied, the first parameter calculation unit (road surface unevenness parameter calculation unit) 2
Bad road determination index (first parameter) β calculated in 10
pp to the road surface friction coefficient calculating means 204 and the second parameter calculating means (slipperiness parameter calculating section) 220
Road surface friction coefficient calculating means 2 for road surface μ determination value α calculated in
The switch function 232, 234, 236 is applied to the signal line so that the output to the signal line 04 (in this embodiment, the output to the average value calculation unit 227 and the output of the road surface μ determination coefficient) γ1 is stopped.
Is provided.

In addition, the road surface index calculating means 202 calculates another parameter that represents the road surface condition from the above two parameters, that is, the average value αh of the bad road determination index βpp and the average value α of the road μ determination value α, Index corresponding to each state of high μ road degree (high μ road), medium μ road degree (medium μ road), and low μ road degree (low μ road) (This index is used because fuzzy inference is used here. Degree).

The road surface friction coefficient calculating means 204 continuously calculates the index (fitness) calculated by the road surface index calculating means 202 and performs cumulative evaluation.
The value indicating the road surface friction coefficient (road surface μ
(Determination coefficient) γ1 is calculated. The road surface index calculating means 202 and the road surface friction coefficient calculating means 204 will be further described later.

By the way, as shown in FIG. 34, the first road surface μ determination coefficient γ1 calculated by the steady-state parameter calculating means 200 and a specific parameter calculating means (to be described later) during the specific running other than the steady running of the vehicle. Means for calculating parameters from state) second road surface μ calculated in 250)
The determination coefficient γ2 is integrated (selected) by the road surface friction coefficient calculation means 260 to indicate a road surface friction coefficient (road surface μ determination coefficient).
Calculate γ.

That is, during steady running, the road surface μ determination coefficient γ1 calculated by the steady
The road surface μ determination coefficient γ2 calculated by the specific parameter calculation means 250 during the specific operation is selected as the road surface μ determination coefficient γ. Note that FIG.
A block B81 shown in FIG.
0 and a part of the function as the road surface friction coefficient calculating means 260.
Is composed of a part of the function of the block B81 and the selector 264. Road surface μ determination condition The road surface μ determination is performed based on the rough road determination index βpp calculated as described above and the average value αh of the road surface μ determination value α, and the road surface μ determination is performed as follows. Performed when the judgment condition is satisfied.

The torque movement amount taf is equal to a predetermined value tax [N
m] or more [taf ≧ tax (Nm)] When the torque movement amount taf is equal to or less than a predetermined value tax,
Since the amount of torque movement is not stable due to a variation in the pressing force of the clutch unit, a reliable determination is made in a region where the amount of torque movement is stable.

The road surface μ judgment value α is not negative [α ≧
0] is the road surface μ determination value α when the response delay at the time of turning back due to steering wheel steering or the longitudinal direction of the tire becomes nonlinear.
Is negative, so that such a case is excluded. The brake switch bksw is off. During braking (that is, when the brake switch bksw is on), the braking force exerts an influence of a speed difference other than the amount of torque movement, so that this is excluded.

The front and rear G (gb) is equal to a predetermined deceleration g1 (g
(1 is a negative minute value) or more [gb ≧ g1] This is to limit the road surface μ determination to a range in which tack-in correspondence control is not performed. Drift determination coefficient srp predetermined value (e.g., r ₁ ²⁾ that more than a [srp ≧ r ₁ ^2] This is because the entire direction of the tire is limited to the case in the grip region.

The turning side G (gy) is smaller than a predetermined value gy1 [gy <gy1] When the turning side G (gy) increases, that is, the turning side G (g).
If y) is equal to or greater than a predetermined value gy1, the stability factor of the vehicle becomes non-linear with respect to the turning side G (gy), and the theory of road surface μ determination based on a linear region does not hold. It is. The average value αh of the road surface μ determination value α thus calculated and the rough road determination index β
FIG. 28 shows a μ determination region that finally converges with respect to pp. In the region where both the road surface μ judgment average value αh and the bad road judgment index βpp are small, the high μ road is used, and in the region where the bad road judgment index βpp is large, the medium μ road and the bad road judgment index βp are used.
In a region where p is small but the road surface μ determination average value αh is large, it is determined to be a low μ road, and in a region where the bad road determination index βpp is medium, a high μ road or a medium μ road if the road surface μ determination average value αh is small. , High μ if the average μh
Road or middle μ road or low μ road, if the road surface μ judgment average value αh is large, middle μ road or low μ road, further, bad road judgment index βpp
Is small and the road surface μ judgment average value αh is medium,
Road or low μ road.

(6) Fuzzy inference (membership function,
hig3, mid3, low3) A road μ determination index is created from the average value αh of the road μ determination value and the rough road determination index βpp, and the road μ determination index uses the road μ determination value αh and the rough road determination index βpp as input conditions. Calculated by fuzzy inference. Here, as shown in FIG. 29, the degree of conformity hig1, m to each road surface [that is, high μ road, middle μ road, low μ road] based on the average value αh of the road surface μ determination values.
id1 and low1 are determined, and the rough road determination index βp
The degree of conformity hig2, mid2, low2 to each road surface [that is, high μ road, middle μ road, low μ road] based on p is obtained, and each road surface [that is, high μ road, middle μ road] is determined by the minimum method. μ
Road, low μ road], select the smaller one of these fitness levels and select the road surface (ie, high μ road, middle μ road, low μ road)
(Fuzzy number) hig3, mid3, l
ow3.

That is, the membership functions shown in FIGS. 30A, 30B, and 30C are set for the average value αh of the road surface μ judgment value, and the membership functions shown in FIG. The degree of conformity hig1 to the high μ road corresponding to the road μ determination value αh is obtained from the function, and the degree of conformity mi to the medium μ road corresponding to the road μ determination value αh is determined from the membership function of FIG.
d1 is obtained, and the degree of conformity low1 to a low μ road corresponding to the road surface μ determination value αh is obtained from the membership function of FIG.

Further, regarding the rough road determination index βpp, FIG.
A membership function as shown in (D), (E) and (F) of FIG. 30 is set, and adaptation from the membership function of FIG. 30 (D) to a high μ road corresponding to the rough road determination index βpp. Degree h
ig2, and the degree of conformity mid2 to the medium μ road corresponding to the rough road determination index βpp is obtained from the membership function of FIG.
From the membership function of FIG. 30 (F), and the degree of conformity low2 to the low μ road corresponding to the rough road determination index βpp is calculated.

Then, the degrees of conformity hig1 and hi to the high μ road are determined.
g2, the smaller value is the high μ road fitness hig.
Select 3 Also, the degree of conformity mid1 and mi to the middle μ road
d2, the smaller value is the mid-μ road suitability mid
Select 3 Furthermore, the degree of conformity low1 and l
ow2, the smaller value is the lower μ road fit lo
Select w3.

(7) Counter function (hig, mid, l
ow) As described above, high μ road fitness hig3, medium μ road fitness m
When the id3 and the low μ road suitability low3 are obtained, the weights (hig, mid, low) of each road surface (that is, the high μ road, the middle μ road, and the low μ road) are obtained based on these. Here, the high μ road conformity hig3 and the medium μ road conformity mid
3, the weight (h) is calculated using a so-called learning function in which the low μ road fitness level low3 is continuously obtained and evaluated cumulatively.
ig, mid, low) respectively. That is, the counter values (experience values) Nh, Nm, and Nl of each road surface μ [these are collectively referred to as Ni. (I = h, m, l)] and the high μ road fitness hig obtained as described above.
3, the counter values (experience values) Nh, Nm, Nl are respectively increased or decreased according to the medium μ road fitness mid3 and the low μ road fitness low, and the weight μg of the high μ road, the weight mid of the medium μ road, The weight low of the low μ road is obtained.

That is, high μ road suitability hig3, medium μ road suitability
The judgment value regarding the strength mid3 and the low μ road suitability low3
And then h₁ , H_Two , H_Three , H_Four (H₁ <H_Two 
<H _Three <H_Four ), M₁ , M_Two , M_Three , M_Four (M₁ <M_Two 
<M_Three <M_Four ), L₁ , L_Two, L_Three , L_Four (L₁ <L_Two 
<L_Three <L_Four ) Is set, and each fitness level hig3, m
By comparing id3 and low3 with these judgment values, FIG.
And updating the counter value as shown below. Note that n
n is a natural number.

[0181] When the high-μ road hig3> h _4, when _{Nh = Nh + nn h 3 <} hig3 ≦ h 4, when _{Nh = Nh + 1 h 2 <} hig3 ≦ h 3, Nh = Nh h 1 < a hig3 ≦ h ₂ when, when Nh = Nh-1 hig3 ≦ h 1, Nh = Nh-nn However, the counter range, when 0 ≦ Nh ≦ Nhmax in μ road _{mid3> m 4, Nm = Nm} + nn m 3 <mid3 ≦ m 4 when, when _{Nm = Nm + 1 m 2 <} mid3 ≦ m 3, when _{Nm = Nm m 1 <mid3 ≦} m 2, when Nm = Nm-1 mid3 ≦ m 1, Nm = Nm-nn However, the counter range when the 0 ≦ Nm ≦ Nmmax low μ road _{low3> l 4, Nl = Nl} + when _{nn l 3 <low3 ≦ l 4} , when _{Nl = Nl + 1 l 2 <} low3 ≦ l 3, Nl = Nl l 1 <low3 when ≦ l _2, N = When _{Nl-1 low3 ≦ l 1,} Nl = Nl-nn However, the counter range, calculation of 0 ≦ Nl ≦ Nlmax · high μ road weight hig, middle μ road weight mid, low μ road weights low Thus, the counter value (experience value) N corresponding to each road surface μ
When h, Nm and Nl are obtained, these counter values N
weights hig, mi of each road surface μ according to h, Nm, Nl
d and low are obtained from a map.

That is, the weight hig of the high μ road is obtained from the counter value Nh of the high μ road by the map shown in FIG. 32A, and the counter value of the middle μ road is obtained by the map shown in FIG. The weight mid of the middle μ road is determined from Nm, and the low weight of the low μ road is determined from the counter value Nl of the low μ road using the map shown in FIG. (8) Calculation of road surface μ judgment coefficient γ Thus, the weight hi of the high μ road and the weight mi of the middle μ road
d, When the low weight of the low μ road is obtained, a weight average value γ of these weights hig, mid, and low is obtained from the following equation, and this weight average value γ is set as a road surface μ determination coefficient γ.

Γ = (w1 * hig + w2 * mid + low) / (hig + mid + low + α) (2.4.1.12) Note that w1 and w2 in the above equations are weights used for calculating the weight average value γ. W1 is the weight value of the weight hig, w2 is the weight value of the weight mid, and the weight value of the weight low is 1, and the weight value is w1 being the largest and w2 being the largest (w1>). w2> 1). Α is an adjustment value, for example, α = 1.

The weight of each road surface, that is, the weight hi of the high μ road
g, the weight mid of the medium μ road, and the weight low of the low μ road are hig ₁ , mid ₁ , and low ₁ , respectively, the value (area) obtained by multiplying the weight of each road surface by each weight value is shown in FIG. 33
As shown in (A), they are Sh, Sm, and Sl, respectively. Γ is the sum (Sh + Sm + Sl) of these area values Sh, Sm, and Sl as the value (hig + mid + low +
α), the weight average value (road surface μ determination coefficient) γ is obtained as a value on the horizontal axis in FIG. 33 (B).

2.4.2 Estimation of Road Surface μ at Non-Linear Turning Next, estimation of road surface μ at the time of non-linear turning will be described. The determination of the non-linear turning is based on the side slip coefficient dgy of the tire, and the road surface μ at the time of the non-linear turning is based on the magnitude of the actual lateral G (rgy) when the side slip coefficient dgy has a non-linear magnitude. It is estimated as follows. Here, FIG.
As shown in (2), when the forced low μ determination condition is satisfied during the nonlinear turning, the low μ road determination is performed, and when the forced high μ determination condition is satisfied, the high μ road determination is performed.

(1) Forced low μ determination condition The forced low μ determination condition is that all of the following conditions are satisfied. -The side slip coefficient dgy has a non-linear magnitude [dgy> dgy1]-The magnitude of the actual lateral G (rgy) is smaller than a set value (rgy1) [| rgy | <rgy1]. The steering wheel angular velocity dθh is less than the set value (dθh1) [dθh <dθh1]. The vibration component βpp of the slip ratio difference is the set value (βpp1)
Less than [βpp <βpp1]. The vehicle speed vb is less than the set value (vb1) [vb
<1vb]. A state in which all of the above conditions are satisfied continues for a predetermined duration ct1 (ct1 is, for example, 100 msce) or more.

When the above condition (AND condition) is satisfied,
The road is determined to be a complete low μ road, and the road surface μ determination coefficient γ (γ2),
The counter values Nh, Nm, Nl corresponding to each road surface μ are set as follows. γ2 = 0, Nh = 0, Nm = 0, and N
1 = Nlmax (2) Condition for judging forced high μ The condition for judging forced high μ is based on only the value of the actual lateral G (rgy) when the sideslip coefficient dgy becomes nonlinear. That is, the actual horizontal G (rgy) is set to a preset value rgy2.
Rgy |> rgy2]. Such a condition is set because the actual lateral G (rgy) becomes equal to or more than the predetermined value rgy2 because it is impossible unless the road is a high μ road.

If the high μ determination condition is satisfied, abrupt transition to high μ road control is not preferable because a sudden change in control is caused. Therefore, when the high μ determination condition is satisfied, Then, the road surface μ determination coefficient γ (γ2) and the counter values Nh, Nm, Nl corresponding to each road surface μ are set as follows. Note that mm is a natural number larger than nn described above.

Γ2 = γ1 + 10 and Nh = Nh + m
m, Nm = Nm-mm, and Nl = N1-mm, where γ2 ≦ γmax, Nh ≦ Nhmax, Nm ≧ 0, N
Let l ≧ 0. In this way, the road is gradually approached to the high μ road. 2.4.3 Estimation of road surface μ at start (1) μ split road determination condition Here, as shown in FIG. 34, is the case where the road surface friction coefficients (road surface μ) of the left and right wheels are different at start (μ split road)? It is determined whether or not the road surface μ is to be estimated based on the determination. The determination of the μ split road is performed as follows mainly based on the torque movement direction taf and the left and right wheel speed difference dvrd of the rear wheels. However, the right turn and right moment shall be positive.

Condition 1 taf> taf1 and dvrd <−vd1 or taf <−taf1 and dvrd> vd1 (taf
1 is a positive set value, vd1 is a positive set value. That is, the torque movement direction taf is directed leftward (taf> ta).
f1) and the left and right wheel speed difference dvrd is negative (dvrd <−
vd1) [that is, the right wheel rotates faster than the left wheel]
Or, the torque movement direction taf is rightward (taf <−
taf1) and the left and right wheel speed difference dvrd is positive (dvrd
> Vd1) [that is, the left wheel is rotating at a higher speed than the right wheel]. In other words, this indicates that the wheel to which the torque has been moved is slipping.

The vehicle speed vb is less than a predetermined value vb2 (vb <
vb2). This is a condition for starting. The steering angle θh is smaller than a predetermined value θh1 (θh <θh1)
Low steering angle. This is a condition that the vehicle is traveling straight. The throttle opening tp is larger than the predetermined value tps1 (tps> tps1) (that is, there is a start operation).

A state where all of the above conditions are satisfied is maintained for a predetermined duration ct2 (ct2 is, for example, 100 msec).
c) Continue above. When the above condition (AND condition) is satisfied, the road is determined to be a μ-split road and a completely low μ road, and the coefficient myu, the road surface μ determination coefficient γ (γ2), and the counter values Nh and Nm corresponding to each road surface μ. , Nl are set as follows.

Myu = 1, γ2 = 0, and Nh
= 0, Nm = 0, and Nl = Nlmax where myu is a coefficient for preventing hunting between the forced high μ determination condition and the forced medium μ determination condition, and the vehicle speed vb is equal to or less than a predetermined value vb3 (vb ≦ vb3) and the torque transfer amount ta
f is equal to or smaller than a set value taf2 (| taf | ≦ taf
In the case of 2), myu = 0.

When the vehicle is not determined to be a μ-split road, the following forced low μ determination condition, forced medium μ determination condition, and forced high μ determination condition are set at the time of starting. , Forcibly, low μ road, medium μ
Road and high μ road. (2) Forced low μ determination condition Here, if even one wheel slips when starting straight ahead, it is forcibly determined to be a low μ road (that is, a μ split road).

Therefore, the condition for forced low μ determination is as follows. The steering angle θh is smaller than the predetermined value θh1 (θh <
θh1) is a low steering angle (that is, the vehicle is traveling straight). The bad road determination index βpp is less than a predetermined value (βpp2) (βpp <βpp2) (that is, the vibration component βp
p is not large).

The throttle opening tp is set to a predetermined value (tps).
1) (tps> tps1) (that is, there is a start operation). The vehicle speed vb is lower than the predetermined value vb2 (vb <vb2) (that is, at the time of starting). One of the wheels is slipping. That is, one of the wheel speeds vfl, vfr, vrl, vrr is higher than the vehicle speed vb by a predetermined speed (v1) or more (vfl> v1, or vfr> v1, or v
rl> v1, or vrr> v1).

A state where all of the above conditions are satisfied is a predetermined duration ct3 (ct3 is, for example, 100 msc).
e) Continue above. When the above condition (AND condition) is satisfied, it is determined that the road is a completely low μ road, and the coefficient myu, the road surface μ determination coefficient γ (γ2), and the counter value N corresponding to each road surface μ
h, Nm, and Nl are set as follows.

Myu = 1, γ2 = 0, and Nh
= 0, Nm = 0, and Nl = Nlmax (3) Forced μ determination condition When the vibration component of the wheel is large at the time of starting, the intermediate value between the medium μ and the low μ is forcibly taken. I do. However, when the forced low μ determination and the forced high μ determination are made, my
From u = 1 to myu = 0 (that is, vb ≦ vb
3, and | taf | ≦ taf2), this determination is not performed.

Therefore, the condition for judging μ during forced operation is as follows. Bad road determination index βpp is a predetermined value (βpp2)
(Βpp> βpp2) (that is, the vibration component βpp is large). The throttle opening tps is larger than a predetermined value (tps1) (tps> tps1)
(That is, there is a start operation).

The vehicle speed vb is less than a predetermined value vb2 (vb <
vb2) (that is, when starting). myu = 0. A state where all of the above conditions are satisfied must be continued for a predetermined duration ct4 (ct4 is, for example, 200 msce) or more.

When the above condition (AND condition) is satisfied,
The road surface μ determination coefficient γ (γ2) and the counter values Nh, Nm, Nl corresponding to each road surface μ are set as follows. γ
2 = γ ₁ , Nh = 0, Nm = Nmmax, and Nl = Nlmax, and γ1 is a value of about _１ ／ of γmax.

(4) Forced High μ Judgment Condition When the wheel does not slip at a certain acceleration or more at the time of starting, a high μ judgment is forcibly made. However, when the forcible low μ determination is made, from myu = 1 to myu = 0 (that is, vb ≦ vb3 and | taf | ≦ ta)
Until f2), this determination is not performed.

Accordingly, the conditions for judging the forced high μ are as follows. Bad road determination index βpp is a predetermined value (βpp2)
(Βpp> βpp2) (that is, the vibration component βpp is not large). The throttle opening tps is larger than a predetermined value (tps2) (tps> tps2)
(That is, there is a certain level of acceleration operation).

The vehicle speed vb is less than a predetermined value vb2 (vb <
vb2) (that is, when starting). myu = 0. G (gb) before and after calculation
Is greater than or equal to a predetermined value gb1 (gb ≧ gb1) (that is, at the time of starting).

The average speed vf of the front wheels is sufficiently close to the vehicle speed vb (| vf | <vb), and the average speed vr of the rear wheels
Is sufficiently close to the vehicle speed vb (| vr | <vb). These indicate that the wheels do not slip. When the above condition (AND condition) is satisfied, it is determined to be high μ, and the road surface μ determination coefficient γ and counter values Nh, Nm, Nl corresponding to each road surface μ are set as follows.

Myu = 1, γ2 = 0, and Nh
= Nhmax and Nm = 0 and Nl = 0 2.4.4 Output value setting (1) Output value setting of each control amount (γtb, γtc, γt
d, γte, tb, tc, td, te) As described above, the control amounts include the target ΔN follow-up control amounts tbh, tbl, the acceleration turning control amounts teh, tel, the tack-in corresponding control amounts tdh, tdl, the steering transient. The response control amounts tch and tcl, a high μ road control amount (high road surface friction resistance control amount) and a low μ road control amount (low road surface friction resistance control amount) are set, respectively. The output control amount tad is calculated while interpolating the two control amounts in accordance with the road friction coefficient γ as the road friction coefficient calculated by the road friction coefficient calculation means.

In other words, as shown in FIG. 35, each of the control amounts is steplessly switched between those for the high μ road and those for the low μ road in accordance with the value of the road surface μ determination coefficient γ. The gain-adjusted value is used as the output value (output gain).

For example, assuming that the control amount for high μ road (control gain for high μ road) is txh and the control amount for low μ road (control gain for low μ road) is txl, the output value (output gain) tx is
It is calculated from the road surface μ judgment coefficient γ by the following equation. The road surface μ determination coefficient γ is a value of 0 to γmax. Here, the case where the road surface μ determination coefficient γ is 0 is a low μ road, and the road surface μ determination coefficient γ
Is γmax, a high μ road, and between the low μ road and the high μ road,
That is, a case where the road surface μ determination coefficient γ has an intermediate value between 0 and γmax is referred to as a medium μ road.

Tx = {γ · txh + (γmax−γ) · txl} / γmax = {(txh−txl) · γ + γmax · txl} / γmax = {(txl−txh) · (γmax−γ) + γmax · txh} / Γmax (2.4.1.1) Here, the control gain (control amount) tx is set so as to shift to the high μ road side or to the low μ road side. Then, the output value is finely adjusted.

The control gain tx is set to the high μ side (the target Δ
N follow-up control: tb) The output value fine adjustment formula for the high μ side uses the corrected output value as tx
a, if the output value fine adjustment coefficient is a (a> 1), the following equation is obtained. txa = {a (txh−txl) · γ + γmax · txl} / γmax = {a · γ · txh + (γmax−a · γ) · txl} / γmax where 0 ≦ a · γ ≦ γmax (2.4.1.2) Note that txa is clipped at the upper limit by txh due to 0 ≦ a · γ ≦ γmax.

As described above, the degree of reflection of the control amount for the high μ road is set to be larger than the control amount for the low μ road when the control amounts of both the high μ and the low μ are interpolated. Such setting on the high μ side is performed with respect to the control gain tb of the target ΔN tracking control. Set the control gain tx to the high μ side and low μ
[Steering angular velocity proportional control (transient response control): tc, tack-in response control: td] In this case, fine adjustment of the output value is not substantially performed, and the above equation (2.4.1.1) is used. To calculate the control gain tx. Such a calculation is performed on the control gain (control amount) tc of the steering angular velocity proportional control (transient response control) and the control gain (control amount) td of the tack-in response control.

The control gain tx is set to the low μ side (acceleration turning control: te). The output value fine adjustment formula for the low μ side is as follows, where txb is the corrected output value and b (b> 1) is the output value fine adjustment coefficient. txb = {b (txl−txh) · (γmax−γ) + γmax · txh} / γmax = [b · (γmax−γ) · txl + {γmax−b · (γmax−γ)} · txh] / γmax (2.4.1.3) Note that txb is clipped at the lower limit by txl, where 0 ≦ b · γ ≦ γmax.

As described above, the degree of reflection of the control amount for the low μ road is set to be greater than the control amount for the high μ road when interpolating the control amounts of the high μ and the low μ. Such setting to the low μ side is performed by controlling the control gain te of the acceleration turning control.
About. Output values (output gain) tx, txa, txb obtained by appropriately performing such fine adjustment of the output value
Can be shown as in FIG. 36 when the road surface μ is illustrated. In FIG. 36, an alternate long and short dash line indicates an output value txa obtained by finely adjusting the control gain tx to the high μ side.
(I.e., the target ΔN tracking control amount tb), and the solid line indicates the output value tx (ie, the tack-in corresponding control amount td, the steering transient response control amount tc) when the control gain tx is not finely adjusted. The dashed line indicates the control gain t.
It shows the characteristics of the output value txb (that is, the acceleration turning control amount te) obtained by finely adjusting the output value of x to the low μ side.

As shown in FIG. 36, the lower the road surface μ (the smaller the road surface μ determination coefficient γ), the smaller the control amount (output value) tx. However, the lower the road surface μ, the higher the control effect. Therefore, in order to obtain the same control effect, it is necessary to reduce the control amount (output value) tx as the road surface μ decreases. Further, the reason that the target ΔN following control amount tb is increased on the medium μ road is that the target ΔN following control is a control capable of maintaining the behavior stability of the vehicle even when the road surface μ is relatively low. This is because the target ΔN tracking control is emphasized and it is desired to actively stabilize the behavior of the vehicle. The reason why the acceleration turning control amount te is reduced on the medium μ road is that the acceleration turning control amount te has a property that it is difficult to secure the behavior stability of the vehicle when the road surface μ is low.

In the characteristic of the output value tx in FIG. 36, the inclination can be changed by a parameter such as a constant unique to the vehicle. Thus, control matching, that is, fine adjustment of the output value can be performed according to the vehicle, and more stable control can be performed.
There is also the advantage that the same basic logic can be used depending on the vehicle.

(2) High-pass processing & final output value tad Here, in order to solve the response delay, FIGS.
As shown in FIG. 8, a high-pass process is performed for the target ΔN follow-up control amount tb, the tack-in corresponding control amount td, and the acceleration turning control amount te. This process is performed, for example, to correct a control delay with respect to a high-frequency input due to fast steering by a high-pass process and to advance a phase of each of these control items.

That is, in driving the actuator (rotational propulsion force distribution adjusting mechanism), it is not possible to avoid a response delay of the actuator with respect to the output of the control signal.
Therefore, it is necessary to perform processing so that the response delay of the actuator does not deteriorate control performance. The control signal includes a control amount (transient control amount) calculated based on a driving operation state such as a steering wheel angle or a steering angle (including a steering angular velocity) θh and a throttle opening tp.
For example, a steering transient response control amount (steering angular velocity proportional control amount) t
c, a control amount calculated based on a vehicle behavior such as a difference between left and right wheel rotational speeds and a lateral acceleration generated in the vehicle (vehicle behavior corresponding control amount), for example, a target ΔN follow-up control amount tb, a tack-in corresponding control amount td, acceleration There is a turning control amount te. Driving operation is essentially the main element of the control command, and the delay of the command does not particularly matter for the control amount according to the driving operation, but since the behavior of the vehicle occurs as a result of the control command, the The control amount set based on the delay has already occurred at the time when the control signal is issued, which may cause a problem.

For example, when the behavior of the vehicle changes suddenly, such a delay in the output of the control amount greatly reduces the control performance. Therefore, in the present apparatus, for example, in order to correct the delay of each control response to the steering input, the control amount corresponding to the vehicle behavior, that is, the target ΔN following control amount tb, the tack-in corresponding control amount td, and the acceleration turning control amount te The high-pass processing is performed to speed up the output of the control signal. As described above, since the steering transient response control amount (steering angular velocity proportional control amount) tc is a control for advancing the phase, there is no need for correction, and the high-pass processing is not performed.

In this control, each control amount tb, td,
The final output value tad is determined by adding te and tc. That is, the ECU 42 functions as an output control amount calculating unit that individually calculates each of the control amounts tb, te, td, and tc based on various parameters, and integrates them to obtain an output value tad. . Therefore, here, the control amounts tb and t necessary for the high-pass processing are described.
For d and te, these are added in advance, and a high-pass process is performed on the added value tfd (= tb + td + te).

High-pass processing The high-pass processing is processing for extracting only high-frequency components from each control output using a high-pass filter. Here, the high-pass processing is performed on the added value tfd of the control amounts tb, td, and te for performing the high-pass processing. To obtain a high-pass processing value tff.

Tfd = tb + td + te (2.4.1.4) tff = HPF [tfd] (2.4.1.5) By this high-pass processing, a control output as shown in FIG. A high-pass processing signal tff as shown in FIG. 38B is output from the signal tfd. That is, in the high-pass processing, only a portion having a large magnitude among the differential values of the control output signal tfd is output as a signal (high-pass processing value calculating means).

Further, the processing value tff subjected to the high-pass processing is added to the control output signal tfd (= tb + td + te) subjected to the high-pass processing [FIG.
(C)] to obtain an output control amount (total value) tf (output control amount calculating means). tf = tfd + tff (2.4.1.6) As shown in FIG. 38, the processing value tff is corrected by the gain (high-pass coefficient) kf (that is, tff * kf).
), The balance with other control amounts may be adjusted.

The final output value (tad) The output control amount calculating means calculates the output control amount tf as shown in the following equation.
Is added to the steering transient response control amount (steering angular velocity proportional control amount) tc that does not perform the high-pass processing to calculate the final output control amount (final output value) tad. tad = tf + tc (2.4.1.7) Limiter In the torque transfer control between the left and right wheels, if the amount of torque transfer is too large, the stability of the behavior of the vehicle may be reduced. The magnitude of the amount of torque movement between the left and right wheels is limited to a maximum value (which is referred to as limit) according to the friction coefficient state of the road surface (road surface μ state).

This limit value, that is, the maximum value limit is determined according to FIG.
As shown in the drawing in block B83 of FIG. 7, the relationship is set to a straight line LIM corresponding to the road surface μ determination coefficient γ. That is, the limit value limit is calculated by the following equation. limit = mg · γ + tal1 (2.4.1.8) where mg is the slope of the straight line LIM, and tal1 is l
This is the minimum value of the limit. As shown in a block B83 in FIG. 37, this minimum value tal1 is a limit value limit corresponding to a road surface μ determination coefficient 1 of a low μ road, and
tal2 is a limit value limit corresponding to the road μ determination coefficient γmid of the medium μ road, and tal3 is a limit value limit corresponding to the road μ determination coefficient γmax of the high μ road. In addition,
The road μ determination coefficient γmid of the medium μ road is set to １ ／ of the road μ determination coefficient γmax of the high μ road (γmid = γmax /
2).

With such a limit value limit, the final output value tad is limited as follows. The following equation also takes into account the case where the final output value tad becomes negative depending on the torque movement direction. −limit ≦ tad ≦ limit ・ ・ ・ ・ ・ ・ (2.4.1.9)

2.5 Actuator Driving In the driving processing (actuator driving processing or proportional valve / directional valve switching control processing), the output value (torque moving amount) t
In response to the torque value, the output value tad is converted into an actuator drive signal for performing a torque movement in a direction and an amount corresponding to the output value tad, and a proportional valve control signal is output to the proportional valve 106 according to the torque movement amount. Then, a directional valve control signal is output to the directional valve (directional switching valve) 107 according to the torque movement direction, and the proportional valve 106 and the directional valve 107 are driven. At the same time, a display command signal is output to the indicator lamp 110 (reference numerals 106, 107, 110 refer to FIG. 3).

The control of the proportional valve 106 and the direction valve 107 is performed by a method disclosed in, for example, Japanese Patent Application Laid-Open No. 7-156681. For example, the proportional valve 106
, The torque transfer-current map (see FIG. 39) and the current correction map (see FIG. 40) from the final output value ta.
Is used to perform control by converting to a target current baseh.

[0228] 3. Operation of this device and effects of this device 3.1 Operation of this device Since this device is configured as described above, for example, FIG.
The control is performed as shown in FIG. That is, first, control is started based on various initial setting inputs, and first, in step S10, an input calculation process is executed (see item 2.1, input calculation process). Next, in step S20, a drift determination logic as shown in FIG. 7 is executed based on the result of the input calculation processing (see item 2.2, drift determination logic). Further, the process proceeds to step S30, where the vehicle motion control logic is executed based on the results of the input calculation process and the drift determination (see item 2.3, vehicle motion control logic).

In this vehicle motion control logic, the target ΔN
Tracking control (see item 2.3.1 Target ΔN tracking control),
Acceleration turning control (refer to item 2.3.2 acceleration turning control),
Control amounts tb, td, te, tc of tack-in response control (see item 2.3.3 Tack-in response control) and steering transient response control (see item 2.3.4 steering transient response control)
Are calculated, and these control amounts tb, td, te, t
c is the high μ road control logic as shown in FIG.
5, the control amounts tbh, tdh, teh, tch on the high μ road and the control amounts tbl, tdl, tel, tcl on the low μ road.
Is calculated as

Then, the process proceeds to a step S40, wherein a μ judgment logic is executed (refer to item 2.4 road surface μ estimation). In this μ determination logic, a road surface μ determination coefficient γ is set (step S50), and a road surface μ determination is performed (step S50).
60), various output values are set (step S7).
0). Then, the process proceeds to step S80 to execute the actuator driving logic (see item 2.5, Actuator driving). That is, in response to the output value (torque moving amount) tad, a proportional valve control signal is output to the proportional valve 106 according to the torque moving amount corresponding to the output value tad, and the output value ta is output.
A directional valve control signal is output to a directional valve (directional switching valve) 107 according to the torque movement direction corresponding to d, and the proportional valve 106 and the directional valve 107 are driven. At the same time, a display command signal is output to the indicator lamp 110.

[0231] Such process, through decision step S90, is performed every predetermined period T _1. 3.2 Effects of the present device 3.2.1 Effects of the vehicle road surface friction coefficient estimating device The vehicle road surface friction coefficient estimating device has a first road surface unevenness state represented by a bad road determination index (road surface unevenness coefficient) βpp. And the road surface μ judgment value α (here, the average value αh)
Since the road surface μ determination is performed based on the second parameter indicating the slipperiness of the road surface, the road surface μ determination can be accurately performed.

Further, rough road determination index (first parameter)
βpp can also be obtained from the vibration component of the wheel speed difference. In this case, the influence of the driving force change (noise) due to sudden braking or shift change is caused by the bad road determination index βpp.
It is difficult to calculate the rough road determination index βpp with high accuracy because it has a large effect on the rough road determination index βpp. Based on the lateral acceleration of the above, it is subjected to turning correction, further high-pass processed, converted to an absolute value, and then calculated by low-pass processing, so that it is hardly affected by a change in driving force (noise). The rough road determination index can be obtained stably with high accuracy, and the road surface friction coefficient can be estimated with high accuracy.

The road surface μ determination value α is determined based on the difference between the left and right rotational speeds of the wheels (here, the rear wheels) provided with the torque transfer control device, and the difference between the inner and outer wheel speeds during turning and the load movement. Since the calculation is performed after the correction, the degree of freedom in the configuration of the drive system is greatly improved, for example, by installing an LSD between the left and right front wheels. Further, since the road surface μ determination is performed using the average value αh of the road surface μ determination value α without directly using the road surface μ determination value α, the accuracy and reliability of the road surface μ determination can be improved.

Then, based on the rough road determination index βpp and the average value αh of the road μ determination value α, an index (fitness) corresponding to each state (road surface state) of high μ road, middle μ road, and low μ road ), And furthermore, the index (fitness) is continuously obtained cumulatively and the cumulative evaluation is performed to calculate the road surface friction coefficient (the road surface μ determination coefficient according to the road surface friction coefficient) γ. In addition, it is possible to obtain an appropriate road surface μ determination coefficient γ with a small estimation error, and there is an advantage that the control performance of the power transmission control device between the left and right wheels for a vehicle is greatly improved.

In particular, in the present embodiment, fuzzy inference is used, and a membership function for obtaining an index (fitness) corresponding to each state (road surface state) of a high μ road, a medium μ road, and a low μ road is used as a vehicle. , It is possible to easily obtain an extremely accurate road surface μ determination coefficient γ. Also,
The calculation of the road surface μ determination coefficient γ (that is, γ1) based on the rough road determination index βpp and the average value αh of the road surface μ determination value α as described above is performed during steady running of the vehicle (particularly, during steady turning). Therefore, a reliable road surface μ determination coefficient γ can be obtained.

In addition, the road surface μ determination coefficient γ1 obtained during the steady running is compared with the specific running other than the steady running (specifically, at the time of starting, at the time of the nonlinear running, or at the time of the μ split running). The road surface μ determination coefficient γ2 is estimated by a method different from that during the steady driving, and the road surface μ determination coefficient (first parameter) γ1 obtained during the steady driving and the road surface μ determination coefficient (the second parameter ) Γ2 is integrated (selected) to obtain the final road surface μ determination coefficient γ, so that the road surface friction coefficient can be estimated in a wider running state other than during steady driving, and the vehicle can be estimated. There is an advantage that contributes greatly to the improvement of the control performance of the power transmission control device between the left and right wheels.

In the μ split state, the torque is moved from the low μ wheel side to the high μ road wheel side by the torque transfer control according to the road surface μ, so that, as shown in FIG.
The driving force transmitted from the wheels on the high μ road side to the road surface is increased, and the vehicle can be started and accelerated more quickly,
It can be performed efficiently. 3.2.2 Effect of Control Corresponding to Vehicle Road Surface Friction Coefficient (Road Surface μ) In this embodiment, during normal control (other than during drift control), the control amount is reduced when the road surface friction coefficient is low. When the road surface friction coefficient is high, the control amount is set to be large, and at the time of drift control, the control amount is set to an intermediate value such that the control amount change according to the road surface friction coefficient becomes small. By performing the correction (gain adjustment), the correction according to the road surface friction coefficient is suppressed. However, the present invention is not limited to this. At the time of drift control, the control amount correction according to the road surface friction coefficient is performed. An intermediate control amount, which is prohibited and does not change according to the road surface friction coefficient, may be provided.

For example, a basic control amount when the road surface friction coefficient is high is set, and during normal control other than the drift control, the basic control amount is corrected by a correction coefficient α such that the basic control amount becomes small when the road surface friction coefficient is low. , During drift control, a value unrelated to the road surface friction coefficient (correction coefficient β)
The basic control amount is corrected so as to be an intermediate value. In this case, the correction coefficient α continuously changes from 1 to a minimum value αmax (αmax <1) according to the road surface friction coefficient, and the correction coefficient β is a fixed value smaller than 1 and larger than the minimum value αmax. .

In this apparatus, a control amount for a high μ road (a control amount corresponding to a high road surface friction resistance) and a control amount for a low μ road (a control amount for a low road surface friction resistance) are set, and these control amounts are set. The output control amount tad is calculated while reflecting the output control amount tad interpolatively according to the road surface μ determination coefficient γ as the road surface friction coefficient calculated by the road surface friction coefficient calculation means. In particular, for the target ΔN following control amount tb, For the same difference ddvr, the map for the high road friction resistance gives a larger control amount than the map for the low road friction resistance, and the degree of reflection of the high μ road control amount is set in proportion to the road friction coefficient. Since the control amount for the high μ road and the control amount for the low μ road are set to a higher μ side than the proportional distribution value for the road surface μ determination coefficient γ, the target ΔN The control gain tb of the follow-up control becomes relatively high.

The target ΔN follow-up control is a control in which the difference between the wheel speeds is interposed between the wheel speeds. Therefore, even when the road surface friction coefficient is low, the influence of the control hardly deviates from the assumed area. Therefore, as described above, in the target ΔN tracking control, a relatively large torque movement is performed while the control amount for the road surface friction coefficient is set relatively large, so that the vehicle behavior can be quickly changed to the target one. The advantage of being able to do it is obtained.

Further, regarding the acceleration turning control amount te, in the region where the lateral G (ggy) is small, the similar
(Ggy), the map for the low road frictional resistance gives a larger control amount than the map for the high road frictional resistance, and
The control amount for the high μ road and the control amount for the low μ road are compared with the road μ determination coefficient γ so that the degree of reflection of the low μ road control amount is greater than the degree of reflection set in proportion to the road surface friction coefficient. Since the value is set to the μ side which is lower than the proportional distribution value, the control gain te of the acceleration turning control amount becomes relatively low.

In the case where the lateral acceleration, such as acceleration turning control, is used as a parameter, especially when the road surface friction coefficient is low, the relationship between the calculated lateral G and the actual lateral G immediately falls into a nonlinear region as shown in FIG. The control influence easily deviates from the assumed area.
Therefore, as described above, in the acceleration turning control, in the region where the lateral G is small, the low road surface friction resistance map gives a larger control amount than the high road surface friction resistance map, and also has a low μ.
The degree of reflection of the road control amount (control amount corresponding to low road surface friction resistance) is set to be greater than the degree of reflection set in proportion to the road surface friction coefficient, so that even when the road surface friction coefficient is low, Stable control can be performed quickly.

When the road surface friction coefficient is high, the high μ
By making the degree of reflection of the road control amount (control amount corresponding to high road surface friction resistance) greater than the degree of reflection set in proportion to the road surface friction coefficient, control in the region where the lateral G is small can be reduced as much as possible. Therefore, energy loss can be suppressed accordingly. Also, in the tack-in correspondence control, it is possible to set the same as the acceleration turning control,
In this case, the same effect as in the case of the acceleration turning control can be obtained.

The fine adjustment of the control amount with respect to the road surface friction coefficient is not limited to the target ΔN follow-up control and the acceleration turning control, but may be applied to those which are hardly affected by the road surface friction coefficient and those which are likely to come out. it can. In addition, since the control amount setting map is provided with the dead zone, the control becomes stable.

The fine adjustment of the control amount with respect to the road surface friction coefficient does not need to be limited to the above embodiment. For example, as a control map for setting a control amount based on parameters such as a rotational speed difference and a lateral acceleration between left and right wheels, a high road surface friction resistance map that provides a high road surface friction resistance control amount, and a low road surface friction resistance A map for low road frictional resistance that provides a corresponding control amount is provided, and an output control amount is calculated while interpolating the high road surface frictional resistance control amount and the low road surface frictional resistance control amount in an interpolation manner according to the road surface friction coefficient. At the same time, when the two control amounts are interpolatively reflected, the degree of reflection of the high road surface friction resistance corresponding control amount and the low road surface friction resistance corresponding control amount in the middle road surface friction resistance is changed in accordance with parameters such as vehicle-specific constants. The adjustment may be made as described above.
In this case, a more appropriate control amount according to the vehicle can be given.

In the above embodiment, the target ΔN follow-up control amount tb and the acceleration turning control amount te are added to secure the turning performance of the rapid acceleration turning. Immediately after the start, the control may be temporarily performed based on the acceleration turning control amount, and thereafter, the control may be switched to the target ΔN tracking control amount for the steady control. In short, it is important to control so that the turning force of the turning outer wheel is increased immediately after the start of the sharp turning.

Although the present embodiment has been described for a four-wheel drive vehicle, the power transmission control device for left and right wheels for the vehicle is as follows.
It can be provided between the left and right driven wheels and left and right driven wheels of two-wheel drive vehicles such as front wheel drive vehicles and rear wheel drive vehicles.
It is conceivable that the present invention is applied between the front and rear wheels of a wheel drive vehicle, and in this case, it is configured as a vehicle power transmission control device. In the present embodiment, fuzzy inference is used in the vehicle road surface friction coefficient estimating apparatus. However, the present invention is not limited to this, and a parameter (βpp
Corresponds to) and a parameter (αh) indicating the slipperiness of the road surface.
(Corresponds to). Indices (hig3, mid3, low3) corresponding to each state (high μ road, middle μ road, low μ road) of other parameters that unify the road surface state from the value ) May be calculated, and another method may be used.

Further, in this embodiment, the vehicle road surface friction coefficient estimating device is used as the vehicle left and right wheel power transmission control device, but it goes without saying that the application of the vehicle road surface friction coefficient estimating device is not limited to this. No. For example, the parameter (rough road determination index) βpp indicating the uneven state of the road surface may be applied for estimating the road surface friction coefficient of a vehicle that does not include the power transmission control device for left and right wheels for a vehicle.

[0249]

As described above in detail, according to the road friction coefficient estimating apparatus for a vehicle according to the first aspect of the present invention, the calculation of the parameters for estimating the road friction coefficient includes the torque movement control apparatus. This can be performed based on the difference between the left and right wheel rotation speeds of the left and right wheels, so that the degree of freedom in the configuration of the drive system is greatly improved, for example, by installing an LSD between the left and right wheels on the side that does not have a torque transfer control device There is.

According to the vehicle road surface friction coefficient estimating apparatus of the second aspect of the present invention, the parameters for estimating the road surface friction coefficient are calculated based on the actual lateral acceleration generated in the vehicle. (Noise),
There is an advantage that the rough road determination index can be obtained stably, and the road surface friction coefficient can be estimated with high accuracy.

[Brief description of the drawings]

FIG. 1 is a schematic overall configuration diagram of a drive system of a vehicle including a vehicle left and right wheel power transmission control device according to an embodiment of the present invention.

FIG. 2 is a schematic configuration diagram showing a rotation propulsion force distribution adjusting mechanism (torque moving mechanism) of the vehicle left-right wheel power transmission control device according to one embodiment of the present invention.

FIG. 3 is a schematic diagram illustrating a configuration of a hydraulic unit and a control system of a rotational propulsion force distribution adjusting mechanism (torque moving mechanism) of the vehicle left-right wheel power transmission control device according to one embodiment of the present invention.

FIG. 4 is a control block diagram of a power transmission control device for left and right wheels for a vehicle according to an embodiment of the present invention.

FIG. 5 is a control block diagram relating to input calculation processing of the power transmission control device for left and right wheels for a vehicle according to one embodiment of the present invention.

FIG. 6 is a diagram for explaining an input calculation process of the power transmission control device between the left and right wheels for a vehicle according to the embodiment of the present invention.

FIG. 7 is a control block diagram relating to a drift determination process of the power transmission control device for left and right wheels for a vehicle according to one embodiment of the present invention.

FIG. 8 is a diagram illustrating a drift determination process of the vehicle left-right wheel power transmission control device according to the embodiment of the present invention.

FIG. 9 is a diagram illustrating a drift determination process of the power transmission control device for left and right wheels for a vehicle according to the embodiment of the present invention.

FIG. 10 is a diagram illustrating a drift determination process of the power transmission control device for the left and right wheels for a vehicle according to the embodiment of the present invention.

FIG. 11 is a diagram illustrating a drift determination process of the power transmission control device for the left and right wheels for a vehicle according to the embodiment of the present invention.

FIG. 12 is a diagram illustrating a drift determination process of the power transmission control device for left and right wheels for a vehicle according to one embodiment of the present invention.

FIG. 13 is a diagram for explaining drift control performed by the power transmission control device for left and right wheels for a vehicle according to the embodiment of the present invention.

FIG. 14 is a control block diagram relating to a control amount calculation process (process for a high μ road) of the power transmission control device for the left and right wheels for a vehicle according to an embodiment of the present invention.

FIG. 15 is a control block diagram relating to a control amount calculation process (process for a low μ road) of the power transmission control apparatus for a vehicle between left and right wheels according to an embodiment of the present invention.

FIG. 16 is a diagram illustrating acceleration turning control of the power transmission control device for left and right wheels for a vehicle according to one embodiment of the present invention.

FIG. 17 is a map (high μ) related to acceleration turning control of the power transmission control device for left and right wheels for a vehicle according to the embodiment of the present invention.
It is a figure which shows a road map.

FIG. 18 is a map (low μ) relating to acceleration turning control of the power transmission control device for left and right wheels for a vehicle according to one embodiment of the present invention.
It is a figure which shows a road map.

FIG. 19 is a diagram illustrating road surface μ determination of the vehicle left and right wheel power transmission control device according to one embodiment of the present invention.

FIG. 20 is a control block diagram illustrating a vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention.

FIG. 21 is a diagram showing a map for correcting a first parameter of the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention.

FIG. 22 is a diagram illustrating an apparatus for estimating a road friction coefficient for a vehicle as one embodiment of the present invention, and is a diagram illustrating an example of a change in the magnitude of a driving force with respect to a wheel slip ratio.

FIG. 23 is a schematic diagram illustrating setting of a second parameter used in the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention.

FIG. 24 is a schematic diagram illustrating setting of a second parameter used in the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention.

FIG. 25 is a schematic diagram illustrating setting of a second parameter used in the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention.

FIG. 26 is a diagram illustrating a relationship between a driving force ratio characteristic (Fx / Fz and a slip ratio S) of a wheel at the time of turning, which explains setting of a second parameter used in the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention. FIG.

FIGS. 27A and 27B are graphs showing actually measured data of a second parameter used in the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention by gentle acceleration circular turning, and FIG.
(B) shows data on a low μ road.

FIG. 28 is a diagram illustrating a surface frictional resistance estimation area (road surface μ determination area) in estimation by the vehicle road surface friction coefficient estimation apparatus as one embodiment of the present invention.

FIG. 29 is a diagram illustrating fuzzy inference by the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention.

FIG. 30 is a diagram showing a membership function for explaining fuzzy inference by the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention, wherein (A), (B) and (C) are road surface μ judgment values; (D), (E), and (F) show the degree of conformity to the high μ road, the middle μ road, and the turning time for the vibration component βpp of the slip ratio difference with respect to αh. The conformity is indicated respectively.

FIG. 31 is a diagram illustrating fuzzy inference by the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention.

FIG. 32 is a diagram illustrating fuzzy inference by the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention;
(A) relates to the weight hi of the high μ road, (B) relates to the weight mid of the medium μ road, and (C) relates to the weight low of the low μ road.

FIG. 33 is a diagram illustrating road surface frictional resistance estimation (operation of road surface μ determination coefficient γ) by the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention, where (A) and (B) are estimations, respectively; The process will be described.

FIG. 34 is a control block diagram illustrating road surface frictional resistance estimation (road surface μ determination) by the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention.

FIG. 35 is a control block diagram illustrating output value setting based on road surface frictional resistance estimation (road surface μ determination) performed by the vehicle road surface friction coefficient estimating apparatus as one embodiment of the present invention.

FIG. 36 is a diagram illustrating fine adjustment of an output value of the power transmission control device for left and right wheels for a vehicle according to one embodiment of the present invention.

FIG. 37 is a main part control block diagram relating to an output value by road surface μ determination (road surface friction coefficient determination) of the vehicle left and right wheel power transmission control device according to one embodiment of the present invention.

FIG. 38 is a diagram illustrating a high-pass process of the vehicle left-right wheel power transmission control device according to one embodiment of the present invention.

FIG. 39 is a diagram showing a map for describing a drive process of the vehicle left and right wheel power transmission control device according to the embodiment of the present invention.

FIG. 40 is a diagram showing a map for explaining a drive process of the power transmission control device for left and right wheels for a vehicle according to one embodiment of the present invention.

FIG. 41 is a flowchart showing an outline of the operation of the power transmission control device for left and right wheels for a vehicle according to one embodiment of the present invention.

FIG. 42 is a view for explaining the control contents intended by the vehicle left and right wheel power transmission control device according to one embodiment of the present invention.

[Explanation of symbols]

2 Engine 50 Rotational propulsion force distribution control mechanism (rotational force adjusting means, torque moving mechanism) 200 Parameter calculating means 202 Road surface index calculating means 204 Road surface friction coefficient calculating means 210 First parameter calculating means (Road surface unevenness parameter calculating unit) 211 Turning correction Unit 212 Turning correction gain calculation unit 213 High-pass processing unit 214 Absolute value processing unit 215 Low-pass processing unit 220 Second parameter calculation means (slippery parameter calculation unit) 221 Rear wheel left / right actual speed difference calculation unit 221 Inner / outer wheel speed difference calculation Unit 223 inner / outer wheel difference correction amount calculation unit 224 load movement correction amount calculation unit 225 slip ratio difference calculation unit 226 road surface μ judgment value α calculation unit 227 average value calculation unit 230 road surface μ judgment condition check unit

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁶ Identification code FI F16H 48/30 G01N 19/02 B G01N 19/02 F16H 1/445

Claims (2)

[Claims] 1. A vehicle state detecting means for detecting a running state of a vehicle, a first parameter indicating a road surface unevenness state based on a running state of the vehicle during steady running detected by the vehicle state detecting means, and a road surface A parameter calculating means for calculating a second parameter indicating the ease of slip, and calculating an index corresponding to each state of a third parameter which represents the road surface state from the values of the first and second parameters. A road surface friction coefficient calculating means for calculating a road surface friction coefficient by continuously and cumulatively obtaining and evaluating the index calculated by the road surface index calculating means; A torque movement control device for controlling the movement of torque between the wheels, a left and right wheel rotation speed difference detecting means for detecting a rotation speed difference between the left and right wheels provided with the torque movement control device, Vehicle speed detecting means for detecting or estimating both vehicle speeds; longitudinal acceleration detecting means for detecting actual longitudinal acceleration occurring in the vehicle; and lateral acceleration detecting means for detecting actual lateral acceleration occurring in the vehicle. The parameter calculation means includes: a vehicle speed detected or estimated by the vehicle speed detection means; a left and right wheel rotation speed difference detected by the left and right wheel rotation speed difference detection means; and an actual longitudinal acceleration detected by the longitudinal acceleration detection means. And calculating the second parameter on the basis of the actual lateral acceleration detected by the lateral acceleration detecting means. 2. A vehicle state detecting means for detecting a running state of a vehicle, a first parameter indicating an uneven state of a road surface from a running state of the vehicle during steady running detected by the vehicle state detecting means, and a slipperiness. Parameter calculating means for calculating a second parameter indicating the road surface, and a road surface for calculating an index corresponding to each state of a third parameter that represents the road surface state from the values of the first and second parameters. An index calculating means, and a road surface friction coefficient calculating means for calculating a road surface friction coefficient by continuously accumulatively calculating and evaluating the index calculated by the road surface index calculating means; Lateral acceleration detecting means for detecting a lateral acceleration of the vehicle, and lateral acceleration calculating means for calculating a theoretical lateral acceleration generated in the vehicle from a steering wheel angle and a vehicle speed of the vehicle. The data calculation means performs gain correction, high-pass processing and low-pass processing on the actual lateral acceleration detected by the lateral acceleration detection means according to the theoretical lateral acceleration calculated by the lateral acceleration calculation means. Thus, the vehicle road surface friction coefficient estimating device, wherein the first parameter is calculated.