JP2007223388A - Device for state estimation and control of vehicle - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device for state estimation and control of a vehicle capable of estimating a vehicle state before the influence of a road surface friction coefficient appears as a vehicle behavior such as a yaw rate. <P>SOLUTION: This state estimation device is provided with: a first lateral force sum calculation means for calculating the first lateral force sum of front and rear wheels; a vehicle state estimation means for estimating a vehicle body slip angle and a road surface friction coefficient; a first wheel slip angle calculation means for calculating the first wheel slip angle based on the estimated vehicle body slip angle; and an each wheel lateral force calculation means for calculating the lateral force of each wheel based on the first each wheel slip angle and the estimated road surface friction coefficient. The vehicle state estimation means is provided with a second lateral force sum calculation means for calculating the second each wheel slip angle, and for setting the initial value of the vehicle body slip angle and the initial value of the road surface friction coefficient, and for calculating the second lateral force sum of the front and rear wheels based on the initial value of the road surface friction coefficient and the second each wheel slip angle. The estimated body slip angle and the estimated road surface friction coefficient are set so that a difference between the first lateral force sum and the second lateral force sum can be minimized. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a road surface friction coefficient estimating device and a control device for stabilizing vehicle behavior.

Conventionally, a technique for estimating a road surface friction coefficient and a slip angle of a vehicle and using the estimated value for vehicle behavior control has been disclosed. In this technique, a vehicle state such as a road surface friction coefficient and a vehicle slip angle is estimated based on a lateral acceleration and a yaw rate of the vehicle, a vehicle motion equation, and the like (see, for example, Patent Documents 1 and 2).
JP 2003-118554 A JP 2003-118612 A

  However, in the above prior art, since the vehicle state is estimated only after the influence of the road surface friction coefficient appears as the vehicle behavior such as the yaw rate, there is a problem that it takes time to estimate.

  The present invention has been made paying attention to the above-mentioned problem, and its purpose is to estimate the vehicle state before the influence of the road surface friction coefficient appears as a vehicle behavior such as a yaw rate, and to estimate the vehicle state. It is to provide a control device.

  In order to achieve the above object, in the present invention, a first tire sum lateral force calculation means for calculating each of the lateral force left and right lateral force sums of the front wheels and rear wheels based on the vehicle state detection value, and each wheel load are estimated. Wheel load estimating means, braking / driving reaction force estimating means for estimating the braking / driving reaction force of each wheel, front wheel and rear wheel against the road surface friction coefficient and vehicle slip angle based on the estimated wheel load and estimated braking / driving reaction force of each wheel A second tire lateral force sum calculating means for calculating each of the lateral force sum of the lateral force of the wheel and a search range for the road surface friction coefficient and a vehicle slip angle calculated by the tire lateral force characteristic calculating means are set. Calculation based on the first tire sum lateral force calculation means from the search range setting means and the second tire lateral force sum that is calculated based on the second tire lateral force characteristic calculation means within the search range. The condition that the difference from the first tire lateral force sum is reduced In addition, a search means for searching for the second tire lateral force sum, an estimated vehicle slip angle and an estimated road surface for determining a combination of the estimated vehicle slip angle and the estimated road surface friction coefficient from the searched second tire lateral force sum. A friction coefficient estimating means.

  Therefore, by estimating the vehicle state based on the relationship between the braking / driving road surface reaction force and lateral force generated on each wheel, the lateral force change due to the change in the road surface friction coefficient and the slip angle of the vehicle body can change the vehicle behavior such as the yaw rate change. It becomes possible to estimate the vehicle state immediately before appearing as a change, and it is possible to realize high vehicle motion performance.

  DESCRIPTION OF THE PREFERRED EMBODIMENTS The best mode for realizing a vehicle state estimation and control apparatus according to the present invention will be described below based on an embodiment shown in the drawings.

[System configuration]
The first embodiment will be described with reference to FIGS. FIG. 1 is a system configuration diagram of an electric vehicle in which four wheels are driven by independent electric motors. The electric vehicle includes three-phase synchronous motors 3fl, 3fr, 3rl, and 3rr provided with permanent magnets on the rotor. The motors 3fl to 3rr are connected to the wheels 2fl and 2fr via the speed reducers 4fl, 4fr, 4rl, and 4rr, respectively. 2rl and 2rr. The output characteristics, the reduction ratio, and the wheel radius of each motor 3fl to 3rr, each reduction gear 4fl to 4rr, and each wheel 2fl to 2rr are the same.

  The steering wheel 11 is connected to the steering rack 14 via a shaft, and the front wheels fl and fr are steered. Further, the turning angle δ of each of the wheels 3fl to 3rr is detected by the turning angle sensor 21 based on the steering angle STR with respect to the steering wheel 11, and is output to the integrated controller 30.

  The integrated controller 30 is an arithmetic unit that calculates torque command values tTFL (left front wheel), tTFR (right front wheel), tTRL (left rear wheel), and tTRR (right rear wheel). In addition to the detected values from the yaw rate sensor 8, the turning angle sensor 21, the brake pedal sensor 22, the accelerator pedal sensor 23, the lateral acceleration sensors 24 and 25, and the longitudinal acceleration sensor 26, the current rotational speeds of the current motors 3fl to 3rr Output torque is input. Based on these input values, torque command values tTFL (left front wheel), tTFR (right front wheel), tTRL (left rear wheel) and tTRR (right rear wheel) for each motor 3fl-3rr are calculated.

  The drive circuits 5fl to 5rr control the power running and regenerative torque of the motors 3fl to 3rr by controlling the power of the battery 6 based on the torque command values tTFL to tTRR from the integrated controller 30.

[Torque command value calculation control]
The integrated controller 30 calculates the torque command values tTFL to tTRR after estimating the vehicle slip angle β and the road surface friction coefficient μ in advance before the vehicle behavior such as the yaw rate is detected as the actual measurement value. More precise motor control is performed by calculating torque command values tTFL to tTRR based on the estimated vehicle slip angle β and road surface friction coefficient μ.

[Torque command value calculation control processing]
FIG. 2 is a main flowchart of a torque command value calculation control process executed by the integrated controller 30. These are executed at regular intervals, for example, every 5 ms. Hereinafter, each step will be described.

  In step S201, each sensor signal and data is captured. That is, the detected values of accelerator opening APS, brake pedal force BRK, steering angle STR, lateral acceleration YG1, YG2, longitudinal acceleration XG, and yaw rate γ are taken in, and the rotational speeds Nfl to Nrr and torque command values of the motors 3fl to 3rr are taken. The current value data of TFL to TRR is fetched, and the process proceeds to step S202. Lateral accelerations YG1 and YG2 are two detection values of the two lateral acceleration sensors 24 and 25.

In step S202, the vehicle speed V is calculated by the following equation, and the process proceeds to step S203.
(Formula 1)
V = (Nfl / GG * R + Nfr / GG * R + Nrl / GG * R + Nrr / GG * R) / 4
Here, R is the radius of each wheel fl to rr, and GG is the reduction ratio of the reducers 4fl to 4rr.

  In step S203, the vehicle body slip angle β and the road surface friction coefficient μ are estimated using a vehicle state estimation routine described later, and the process proceeds to step S204.

  In step S204, based on the estimated vehicle slip angle β and road surface friction coefficient μ, motor torque command values TFL to TRR for each wheel 2fl to 2rr are calculated using a motor torque command value calculation control routine described later, and then driven. Output to the circuits 5fl to 5rr to finish the control.

[Vehicle condition estimation routine: estimation of vehicle slip angle and road friction coefficient]
FIG. 3 is a diagram showing the relationship between the vehicle slip angle β and each vehicle behavior parameter in the vehicle state estimation routine in step S203 in the flow of FIG. Based on the steering angle STR, the brake pedal force BRK, and the lateral accelerations YG1 and YG2 detected by the turning angle sensor 21, the brake pedal sensor 22, and the lateral acceleration sensors 24 and 25, the turning angle δ, the braking / driving road surface reaction force Fx, The yaw moment MM and the lateral force YG are calculated, and the vehicle slip angle β and the road surface friction coefficient μ are determined based on the yaw moment MM and the lateral force YG.

(Calculation of turning angle)
Since the vehicle in the present application is a front wheel steering type, the rear wheel turning angle is zero. Accordingly, the values of the front wheel turning angles δfl and δfr corresponding to the steering angle STR are read from the steering angle-steering angle maps MAP_STRfl and MAP_STRfr of the front wheels and the rear wheels, and are calculated as follows.
(Formula 2)
δfl = MAP_STRfl (STR)
δfr = MAP_STRfr (STR)
δrl = 0
δrr = 0

(Calculation of braking / driving road surface reaction force)
The braking / driving road surface reaction forces Fxfl to Fxrr are estimated from the motor torque command values TFL to TRR and the rotation speed variation of each wheel 3fl to 3rr. As a specific estimation method, for example, a method disclosed in JP-A-6-98418 (Equation 11) is used. In particular, when using a mechanical brake, the front and rear wheel braking force maps are read from the brake pedal force BRK value, and the braking torque TBL_FBR (BRK) and TBL_RBR (BRK) are combined with the motor torque TFL to TRR. The driving road surface reaction forces Fxfl to Fxrr are calculated. Further, when the anti-skid braking system is operated, a braking torque decrease due to the system operation is detected, and the braking torque is corrected.

(Calculation of lateral force and yaw moment)
FIG. 4 is a diagram showing the relationship between the lateral acceleration and each vehicle behavior parameter in the vehicle state estimation routine in step S203 in the flow of FIG. The lateral force Fy is positive when the vehicle travels to the left. When the installation position of the lateral acceleration sensor is the position shown in FIG. 4, the yaw angular acceleration γ ′ (unit is rad / s ² , counterclockwise is positive) and the lateral acceleration YG at the center of gravity (unit is m / s ² , where the left direction of the vehicle is positive) is obtained as follows.
(Formula 3)
γ '= (YG2-YG1) / L2
YG = YG1 + (YG2-YG1) * Lc / L2

From the yaw angular acceleration γ 'and lateral acceleration YG value, the lateral force Fy and yaw moment MM acting on the vehicle are calculated using the vehicle yaw inertia design value I _γ [kgm ² ] and vehicle mass design value M [kg]. Calculate as follows.
(Formula 4)
Fy = M * YG
MM = I _γ * γ '

ここでLf は前輪軸重心点距離[m]、Lr は後輪軸重心点距離[m]、Lt はトレッドベース距離(前後輪同一)[m]であり、それぞれ設計定数である。
上記数式５の第３項の影響は小さいものとして無視し、(Fyfl+Fyfr)および(Fyrl+Fyrr)を次のように求める。
（数式６）

ここでδfとして(δfl+δfr)/2を用い、δrとして(δrl+δrr)/2を用いることとし、(Fyfl+Fyfr)を第１前輪横力和Fy_f、(Fyrl+Fyrr)を第１後輪横力和Fy_rに代入する。 (Calculation of left and right lateral force sum of front and rear wheels)
As shown in FIG. 3, the lateral force generated at the left front wheel is Fyfl, the lateral force generated at the right front wheel is Fyfr, the lateral force generated at the left rear wheel is Fyrl, and the lateral force generated at the right rear wheel is Fyrr, If the left and right turning angles of the front wheels are equal (δfl = δfr = δf) and the left and right turning angles of the rear wheels are equal (δrl = δrr = δr), the following relational expression is given for the lateral force Fy and yaw moment MM acting on the vehicle: Holds.
(Formula 5)

Here, Lf is the front wheel center of gravity distance [m], Lr is the rear wheel center of gravity center distance [m], and Lt is the tread base distance (same front and rear wheels) [m], which are design constants.
The influence of the third term of Equation 5 is ignored as being small, and (Fyfl + Fyfr) and (Fyrl + Fyrr) are obtained as follows.
(Formula 6)

Here, (δfl + δfr) / 2 is used as δf, (δrl + δrr) / 2 is used as δr, (Fyfl + Fyfr) is the first front wheel lateral force sum Fy_f, and (Fyrl + Fyrr) is the first. Substitute in the rear wheel lateral force sum Fy_r.

ここで、hは重心高[m]、Ltはトレッド幅[m]である。そして、これらの値に基づき、以下の式で各輪の輪荷重を演算する。
（数式９）

この方法以外に、センサを用いて検出した輪荷重値を用いても良い。 (Calculation of each wheel load)
The front wheel load Wf when the vehicle is stationary and the rear wheel load Wr (unit is N) when the vehicle is stationary are calculated by the following equations.
(Formula 7)
Wf = M * Lr / (Lf + Lr) /2*9.8
Wr = M * Lf / (Lf + Lr) /2*9.8
Next, the front-rear wheel load movement amount ΔWd and the left-right wheel load movement amount ΔWc are calculated as follows from the vehicle longitudinal acceleration XG and the lateral acceleration YG at the center of gravity obtained by Expression 5.
(Formula 8)

Here, h is the center of gravity height [m], and Lt is the tread width [m]. Based on these values, the wheel load of each wheel is calculated by the following equation.
(Formula 9)

In addition to this method, a wheel load value detected using a sensor may be used.

(Setting of vehicle slip angle search range)
In estimating the vehicle slip angle β, a method of searching for and estimating a value that meets a condition from a certain fixed range is used. Therefore, it is necessary to determine a fixed range to be searched in advance. Therefore, the vehicle reference position is the vehicle body reference position P in FIG. 3 (the center point on the left and right sides of the rear wheel shaft in FIG. 3), and the search range of the slip angle β based on the value of the first rear wheel lateral force sum Fy_r as follows: To decide. The minimum and maximum values of the search range are indicated by b_min and b_max, respectively.
・ When the first rear wheel lateral force sum Fy_r is positive: Minimum value b_min = -π / 4 Maximum value b_max = 0
-π / 4 ≦ slip angle β search range ≦ 0
・ When the first rear wheel lateral force sum Fy_r is negative: Minimum value b_min = 0 Maximum value b_max = π / 4
0 ≦ Slip angle β Search range ≦ π / 4
・ When the first rear wheel lateral force sum Fy_r is 0: Minimum value b_min = Maximum value b_max = 0
No slip angle β search range. Further, the search step width Δb of the search range is set by the following equation, for example.
Δb = (b_max- b_min) / 10
It is assumed that the vehicle body slip angle has a unit of rad, and the forward direction of the point P with respect to the vehicle forward direction is a positive direction.

(Setting the road friction coefficient search range)
The road surface friction coefficient μ is also estimated by searching for a value that meets the condition from a certain range, similar to the vehicle body slip angle β. Therefore, 0.1 is set as the minimum value of the search range as m_min, and 1.0 is set as the maximum value as m_max. Further, the search step width Δm of the search range is set to 0.1, for example.

(Search for body slip angle and road friction coefficient)
In searching for the vehicle body slip angle β and the road surface friction coefficient μ, the difference between the estimated values of the first front-rear wheel lateral force sums Fy_f and Fy_r and the actual measurement values is preferably the minimum, and the minimum is considered in consideration of the required detection accuracy ( A combination of (β, μ) is selected and used as an estimated value (b_opt, m_opt) of (β, μ). An optimal combination selection method will be described later in the search routine for the road surface μ and the slip angle β.
(Calculation of road surface friction coefficient based on each wheel slip ratio)
The road surface friction coefficient m0 based on the slip ratio S is estimated from the relationship between the slip ratios Sfl to Srr of each wheel and the braking / driving road surface reaction force. Here, it is determined from the following conditions that the vehicle is “almost straight ahead”.
Judgment conditions: | Fy_f | and | Fy_r | are both not more than a predetermined value a, and the absolute value of the steering amount STR may be within a predetermined value.
If this condition is met and it is determined that it is not “almost straight running”, m0 = 1 is set, and if it is determined that it is “almost straight running”, the following formula is used for each wheel. The road surface friction coefficient m0 is estimated from the slip ratios Sfl to Srr and the road surface reaction force. The predetermined value a is, for example, a vehicle mass (0.5 or the like).
(Formula 10)
m0_fl = MAP_FM (| Sfl |, | Fxfl / Wfl |)
m0_fr = MAP_FM (| Sfr |, | Fxfr / Wfr |)
m0_rl = MAP_rlM (| Srl |, | Fxrl / Wrl |)
m0_rr = MAP_FM (| Srr |, | Fxrr / Wrr |)
m0 = min (m0_fl, m0_fr, m0_rl, m0_rr)
Here, the slip ratios Sfl to Srr of each wheel are obtained by the following equations.
(Formula 11)
Left front wheel: When V- (Nfl / GG * R) ≥ 0, Sfl = {V- (Nfl / GG * R)} / V
When V- (Nfl / GG * R) <0, Sfl = {V- (Nfl / GG * R)} / (Nfl / GG * R)
Right front wheel: When V- (Nfr / GG * R) ≥ 0, Sfr = {V- (Nfr / GG * R)} / V
When V- (Nfr / GG * R) <0, Sfl = {V- (Nfr / GG * R)} / (Nfr / GG * R)
Left rear wheel: Srl = {V- (Nrl / GG * R)} / V when V- (Nrl / GG * R) ≥ 0
When V- (Nrl / GG * R) <0, Srl = {V- (Nrl / GG * R)} / (Nrl / GG * R)
Right rear wheel: When V- (Nrr / GG * R) ≥ 0, Srr = {V- (Nrr / GG * R)} / V
When V- (Nrr / GG * R) <0, Srr = {V- (Nrr / GG * R)} / (Nrl / GG * R)
MAP_FM and MAP_RM are slip ratios Sfl to Srr and the ratio of road surface reaction force to wheel load at each wheel fl to rr, respectively. MAP_FM is a characteristic map for the front wheel and MAP_RM is a characteristic map for the rear wheel. FIG. 5 shows an example of MAP_FM.

(Calculation of each wheel slip angle)
The slip angle of each wheel is calculated from the slip angle b_opt among the estimated values (b_opt, m_opt) of (β, μ) obtained by searching the vehicle body slip angle and the road surface friction coefficient. The relationship between the vehicle slip angle β and the tire slip angles αfl to αrr of each wheel is approximately as follows. Chapter 3 of “Automobile Movement and Control (Sankaido, Author: Masato Abe)” 54). In addition, the traveling direction of the center of the tire is a clockwise direction with respect to the direction of the tire rotation surface (see FIG. 3).
αfl = (V * b_opt + (Lf + Lr) * γ) / (V-Lt * γ / 2)-δfl
αfr = (V * b_opt + (Lf + Lr) * γ) / (V + Lt * γ / 2)-δfr
αrl = V * b_opt / (V-Lt * γ / 2)-δrl
αrr = V * b_opt / (V + Lt * γ / 2)-δrr

(Calculation of lateral force for each wheel)
Of the estimated values (b_opt, m_opt) of (β, μ) obtained by searching for the first slip angle α1fl to α1rr of each wheel 3fl to 3rr and the vehicle body slip angle and the road surface friction coefficient, the estimated road friction coefficient Using m_opt, the lateral forces Fyfl to Fyrr (see FIG. 3) of the wheels 3fl to 3rr are calculated by the following equation. The calculation uses tire characteristic maps MAP_FT and MAP_RT for the front and rear wheels.
(Formula 12)
Fyfl = MAP_FT (Wfl, Fxfl, m_opt, α1fl)
Fyfr = MAP_FT (Wfr, Fxfr, m_opt, α1fr)
Fyrl = MAP_RT (Wrl, Fxrl, m_opt, α1rl)
Fyrr = MAP_RT (Wrr, Fxrr, m_opt, α1rr)

[Vehicle behavior estimation control processing]
FIG. 6 is a routine of the vehicle behavior control process executed in step S203 of FIG. Hereinafter, each step will be described.

  In step S401, each wheel turning angle δfl to δrr is calculated, and the process proceeds to step S402.

  In step S402, each wheel drive road surface reaction force Fxfl to Fxrr is calculated, and the process proceeds to step S403.

  In step S403, yaw moment MM and lateral force YG are calculated, and the process proceeds to step S404.

  In step S404, first front and rear wheel lateral force sums Fy_f and Fy_r are obtained, and the process proceeds to step S405.

  In step S405, the wheel loads Wfl to Wrr are calculated, and the process proceeds to step S406.

  In step S406, the search range minimum value b_min, maximum value b_max, and search step width Δb of the vehicle slip angle β are determined, and the process proceeds to step S407.

  In step S407, 0.1 is set as the minimum value m_min of the search range of the road surface friction coefficient μ, and 1.0 is set as the maximum value m_max. Further, the search step width Δm is set to 0.1, and the process proceeds to step S408.

  In step S408, a combination (m_opt, b_opt) of the estimated road surface friction coefficient and the estimated vehicle slip angle is selected and calculated, and the process proceeds to step S409.

  In step S409, a road surface friction coefficient m0 considering the slip ratio is estimated from each wheel slip ratio Sfl to Srr, and the process proceeds to step S411.

  In step S411, it is determined whether or not the estimated road surface friction coefficient m_opt is larger than the road surface friction coefficient m0 considering the slip ratio. If YES, the process proceeds to step S412. If NO, the process proceeds to step S414. As a result, when it is determined that the vehicle is substantially straight, the estimated value b_opt of the vehicle body slip angle β is recalculated based on the road surface friction coefficient m0 in consideration of the slip ratio.

  In step S412, m_opt = m0 is set, and the process proceeds to step S413.

  In step S413, the estimated value b_opt of the vehicle slip angle β is recalculated, and the process proceeds to step S414.

  In step S414, the first tire slip angles α1fl to α1rl of each wheel are calculated, and the process proceeds to step S415.

  In step S415, each wheel lateral force Fyfl to Fyrr is calculated, and this routine ends.

[Search routine for road surface μ and slip angle β]
(Search initial value setting)
In order to search for the estimated values m_opt and b_opt of the vehicle body slip angle β and the road surface friction coefficient μ, first, the road surface friction coefficient m_min and the vehicle body slip angle b_min at the start of the search are set.

  Next, m_min is set as the initial value of the road surface friction coefficient estimated value candidate m_opt, and b_min is set as the initial value of the estimated vehicle body slip angle β estimated value b_opt. Also, a sufficiently large value (for example, 1000000) is input as J_opt (described later).

  Further, when m_opt = m0 is set in S412 described above and the vehicle slip angle β is most calculated in S413, the lower limit m_min and the upper limit m_max of the search range of the road surface friction coefficient are both set to m_opt, and the increment Δm Is set to 1. (To avoid continuing from step S503 to step S512 infinitely)

(Each wheel tire slip angle calculation)
The second tire slip angles α2fl to α2rr of the wheels 3fl to 3rr with respect to the vehicle slip angle β are calculated by the following equation. The relationship between the vehicle slip angle β and the second tire slip angle α2fl to α2rr of each wheel is approximately as follows. Chapter 3 of “Motor Movement and Control (Sankaido, Author: Masato Abe)” p.54). In addition, the traveling direction of the center of the tire is a clockwise direction with respect to the direction of the tire rotation surface (see FIG. 3).
(Formula 13)
α2fl = (V * b + (Lf + Lr) * γ) / (V-Lt * γ / 2)-δfl
α2fr = (V * b + (Lf + Lr) * γ) / (V + Lt * γ / 2)-δfr
α2rl = V * b / (V-Lt * γ / 2)-δrl
α2rr = V * b / (V + Lt * γ / 2)-δrr

(Calculated estimated value of lateral force sum of front and rear wheels)
Using the tire characteristic map, the second front and rear wheel lateral force sums Fymf and Fymr corresponding to the estimated road surface friction coefficient m are calculated based on the following equations.
(Formula 14)
Second front wheel lateral force sum Fymf = MAP_FT (Wfl Fxfl, m, α2fl) + MAP_FT (Wfr, Fxfr, m, α2fr)
Second rear wheel lateral force sum Fymr = MAP_RT (Wrl Fxrl, m, α2rl) + MAP_RT (Wrr, Fxrr, m, α2rr)

  Here, MAP_FT inputs data so as to output the lateral force generated by the front wheel tires 3fl and 3rr with respect to the wheel load, braking / driving road surface reaction force, road surface friction coefficient, and second tire slip angles α2fl to α2rr. 4 is a map of four inputs and one output. The MAP_RT data is input to output the lateral force generated by the rear wheel tires rl and rr for the wheel load W, braking / driving road surface reaction force Fx, road surface friction coefficient μ, and second tire slip angle α2. 4 is a map of four inputs and one output.

  Both are data maps of measured tire characteristics, and the smaller the road surface friction coefficient, the smaller the absolute value. Further, when the braking / driving road surface reaction force is positively large or negatively large, the absolute value becomes small.

(Minimization of the difference between the first and second lateral force sums)
After estimating the second front-rear wheel lateral force sum Fymf, Fymr, the difference between the second front-rear wheel lateral force sum estimated value (Fymf, Fymr) and the first front-rear wheel lateral force sum (Fy_f, Fy_r) is minimized. In order to search for a combination of the vehicle body slip angle β and the road surface friction coefficient μ, first, the closer the first front wheel lateral force sum Fy_f and the second front wheel lateral force sum Fymf detected in step S404 in FIG. The value J of the function that outputs a smaller value is calculated as the rear wheel lateral force sum Fy_r and the second rear wheel lateral force sum Fymr are closer. Here, the following functions are used.
(Formula 15)
J = | 2nd front wheel lateral force sum Fymf-1st front wheel lateral force sum Fy_f |
+ | Second rear wheel lateral force sum Fymr-First rear wheel lateral force sum Fy_r |
Other
J = (2nd front wheel lateral force sum Fymf-1st front wheel lateral force sum Fy_f) ²
+ (Second rear wheel lateral force sum Fymr-first rear wheel lateral force sum Fy_r) ² or the like.

Here, if J is less than or equal to the stored J_opt, the second front wheel lateral force sum Fymf and the second rear wheel among the combinations of the road surface friction coefficient m and the vehicle body slip angle β. It is determined that the lateral force sum Fymr is closest to the first front wheel lateral force sum Fy_f and the first rear wheel lateral force sum Fy_r. If J is less than or equal to the stored J_opt, the evaluation value J_opt, the road surface friction coefficient candidate value m_opt, and the vehicle slip angle candidate value b_opt are updated as follows. If it is more than J_opt, it is not updated.
(Formula 16)
J_opt = J, m_opt = m, b_opt = b

  Next, a value obtained by adding Δb is set as the next search value b of the vehicle slip angle β (setting of the vehicle slip angle search range in the above-described vehicle behavior estimation control: see step S406 in FIG. 6). However, when Δb is 0, in order to avoid the infinite loop of steps S503 to S509, the b value is calculated as b = b + 1.

(Candidate values for estimated vehicle slip angle and estimated road friction coefficient)
It is determined whether or not the next search value b of the vehicle body slip angle β exceeds the search range upper limit value b_max. If not exceeded, the second wheel tire slip angles α2fl to α2rr and the second front and rear wheel lateral force sums Fymf and Fymr are calculated again with the search value b, and the difference between the first and second lateral force sums is calculated. Decrease and set the next search value b + Δb. When the upper limit value b_max is exceeded, b = b_min is set as a candidate value for the estimated vehicle slip angle β.

  Further, the next search candidate value m of the road surface friction coefficient μ is updated to a value obtained by adding Δm. After the update, it is determined whether or not the search value m of the road surface friction coefficient μ exceeds the upper limit value m_max of the search range. If not, start again from the second wheel tire slip angles α2fl to α2rr. If the upper limit m_max is exceeded, set this search value m as the estimated value of the road surface friction coefficient μ. To do.

  That is, in the search and estimation control of the vehicle slip angle and the road surface friction coefficient, the vehicle slip angle and the road surface friction coefficient are set to Δb between the vehicle slip angle range b_min to b_max and the road surface friction coefficient range m_min to m_max, respectively. Change the combination in increments of Δm. Among the combinations of m and b at the time of change, the combination of the second front and rear wheel lateral force sums Fymf and Fymr calculated using the tire characteristic map MAP_FT and the first front and rear wheel lateral force sums Fy_f and Fy_r are close ( b_opt, m_opt) are extracted.

[Search / estimation control process for vehicle slip angle and road friction coefficient]
FIG. 7 shows a search / estimation control processing routine for a vehicle slip angle and a road surface friction coefficient. Hereinafter, each step will be described.

  In step S501, m_min is set as the road surface friction coefficient μ at the start of the search, and b_min is set as the vehicle body slip angle β at the start of the search, and the process proceeds to step S502.

  In step S502, m_min is set as the initial value of m_opt, and b_min is set as the initial value of b_opt. Further, 1000000 is input to J_opt, and the process proceeds to step S503.

  In step S503, second tire slip angles α2fl to α2rr of the respective wheels 3fl to 3rr with respect to the vehicle slip angle β are calculated, and the process proceeds to step S504.

  In step S504, the second front wheel lateral force sum Fymf and the second rear wheel lateral force sum Fymr corresponding to the tire slip angle b and the road surface friction coefficient m at the start of the search are read from the map, and the process proceeds to step S505.

  In step S505, the value J of the function that outputs a smaller value as the first front and rear wheel lateral force sums Fy_f and Fy_r and the second front and rear wheel lateral force sums Fymf and Fymr in the front wheels are closer is calculated, and the process proceeds to step S506.

  In step S506, it is determined whether or not the function J is equal to or less than the stored J_opt. If YES, the process proceeds to step S507, and if NO, the process proceeds to step S508.

  In step S507, the evaluation value J_opt, the road surface friction coefficient candidate value m_opt, and the vehicle slip angle candidate value b_opt are updated as J_opt = J, m_opt = m, and b_opt = b, and the process proceeds to step S508.

  In step S508, Δb is added to the next search value b of the vehicle body slip angle β, and the process proceeds to step S509.

  In step S509, it is determined whether or not the next search value b of the vehicle slip angle β exceeds the search range upper limit value b_max. If YES, the process proceeds to step S510, and if NO, the process returns to step S503.

  In step S510, b = b_min is set as a candidate value for the estimated vehicle slip angle β, and the process proceeds to step S511.

  In step S511, Δm is added to the next search candidate value m of the road surface friction coefficient μ, and the process proceeds to step S512.

  In step S512, it is determined whether or not the search value m of the road surface friction coefficient μ exceeds the upper limit value m_max of the search range. If YES, the search value m is used as a candidate value of the estimated value of the road surface friction coefficient μ. If NO, the process returns to step S503.

[Motor torque command value calculation control]
In the motor torque command value calculation control in step S204 of FIG. 2, first, the target driving force tTD of the vehicle is read from the vehicle speed-accelerator opening degree map MAP_tTD (see FIG. 8).

Next, motor torque command values tTFL to tTRR for the respective wheels fl to rr are calculated based on the following equation. Note that tU is determined by reading the V-tU map of FIG. 9 based on the target driving force vehicle speed and the steering amount STR.
(Formula 17)
Left front, rear wheel tTFL, tTRL = tTD * R / GG / 4-tU * R / GG / 4
Right front, rear wheel tTFR, tTRR = tTD * R / GG / 4 + tU * R / GG / 4
Here, GG is a reduction ratio in the reduction gears 4fl to 4rr. The yaw rate γ and the vehicle lateral acceleration YG may be corrected so as to have a desired transient response, and are not particularly limited. For details on the correction method, see Chapter 1.8.

  With respect to the calculated motor torque command values tTFL to tTRR, the driving torque of each wheel fl to rr is redistributed so that the lateral force Fy, yaw moment MM, and longitudinal acceleration XG of the vehicle do not change. Specifically, according to the sensitivity k of the lateral force YG to the longitudinal acceleration (longitudinal force) XG based on the estimated wheel load W of each wheel, the first tire slip angle α1, the estimated turning angle δ, and the estimated road friction coefficient μ. Redistribute. Motor torque command values tTFL to tTRR of the respective wheels fl to rr after the redistribution are output to the motor drive circuit.

[Motor torque command value calculation control process]
FIG. 10 is a motor torque command value calculation control processing routine. Hereinafter, each step will be described.

  In step S601, the target driving force tTD is calculated, and the process proceeds to step S602.

  In step S602, motor torque command values tTFL to tTRR for the respective wheels 3fl to 3rr are calculated, and the process proceeds to step S603.

  In step S603, the drive torque of each wheel fl to rr is reallocated, and the process proceeds to step S604.

  In step S604, motor torque command values tTFL to tTRR for the respective wheels fl to rr are output, and this routine ends.

[Effect of the embodiment of the present application]
(1) In the present embodiment, the first lateral force sum calculating means for calculating the lateral force sums Fy_f and Fy_r of the first front wheels and the rear wheels based on the vehicle state, the vehicle body slip angle β and the road surface friction coefficient μ are calculated. Vehicle state estimation means to be estimated, first wheel slip angle calculating means for calculating first wheel slip angles α1fl to α1rr based on the estimated vehicle slip angle β, and first wheel slip angles α1fl ˜α1rr and each road lateral force calculation means for calculating the lateral force Fyfl˜Fyrr of each wheel based on the estimated road surface friction coefficient m_opt, and the vehicle state estimation means includes a second wheel slip angle α2fl based on the vehicle state. ~ Α2rr is calculated, the initial value b_min of the vehicle slip angle β and the initial value m_min of the road surface friction coefficient are set, and based on the vehicle condition, the initial value m_min of the road surface friction coefficient, and the second wheel slip angle α2fl to α2rr , Second lateral force sum calculating means for calculating the lateral force sums Fymf and Fymr of the second front wheel and the rear wheel, Lateral force sum Fy_f of, Fy_r a second lateral force sum Fymf, as the difference between the Fymr is minimized, it was decided to set the estimated vehicle body slip angle b_opt and the estimated road friction coefficient M_opt.

  As a result, the estimation uses the characteristic that the lateral force of the tire changes according to the road friction coefficient and the tire slip angle, and changes in the lateral force Fy due to changes in the road friction coefficient μ and vehicle slip angle β. It is possible to quickly estimate the vehicle state before appears as a change in vehicle behavior, so that a high vehicle movement performance can be realized.

  (2) Lateral acceleration sensors 24 and 25 for detecting the lateral acceleration of the vehicle, vehicle lateral force estimating means for estimating the vehicle lateral force Fx based on the detected lateral accelerations YG1 and YG2, and detected lateral acceleration YG1, Based on YG2, yaw moment estimating means for estimating yaw moment γ ', each wheel load calculating means for calculating each wheel load Wfl to Wrr, and control force Fxfl to Fxrr for estimating each wheel load 3fl to 3rr. Drive reaction force estimating means, and turning angle calculating means for calculating the turning angles δfl to δrr of each vehicle wheel, and the first lateral force sums Fy_f and Fy_r are the estimated vehicle lateral force Fx and the estimated yaw moment, respectively. The second wheel slip angles α2fl to α2rr are calculated based on the vehicle speed V, the yaw rate γ, and the wheel turning angles δfl to δrr. The second lateral force sums Fymf and Fymr are calculated on the road surface. Initial value m_min of friction coefficient μ, each wheel load Wfl to Wrr, each wheel braking / driving reaction force Fxfl to Fxrr, and each second wheel slip angle α2 The calculation is based on fl to α2rr.

  By using the characteristic that the lateral force Fy of the tire changes according to the road surface friction coefficient μ and the tire slip angle α, the estimation is performed based on the relationship between the estimated braking / driving road surface reaction force Fxm and the lateral force Fy. The vehicle slip angle β and the road surface friction coefficient μ can be estimated at high speed.

  (3) A straight travel determination means for determining that the vehicle is traveling substantially straight, a slip ratio detection means for detecting a slip ratio of each wheel, and a braking / driving road surface of each wheel when it is determined that the vehicle is traveling substantially straight. The road surface friction coefficient estimating means for calculating the road surface friction coefficient from the relationship between the reaction force estimated value and the slip ratio is provided.

  When the vehicle is in a straight-ahead state where the difference in the lateral force Fy due to the difference in the road surface friction coefficient μ is small, the above method (2) tends to deteriorate the estimation accuracy of the road surface friction coefficient μ (as an extreme example, When the vehicle is traveling straight ahead, both the front wheel lateral force sum Fyf and the rear wheel lateral force sum Fyr become 0 regardless of the road surface friction coefficient μ, and the road surface friction coefficient μ cannot be estimated in principle). Therefore, to compensate for this disadvantage, the road surface friction coefficient μ is estimated from the relationship between the slip ratio S of the wheel and the braking / driving road surface reaction force Fx when the vehicle is running straight. As a result, the chances of estimating the road surface friction coefficient μ can be increased, and the estimation accuracy of the road surface friction coefficient μ in a substantially straight traveling state can be improved.

  (4) The straight traveling determination means determines whether or not the vehicle is traveling substantially straight based on the magnitudes of the first front wheel lateral force sum Fy_f of the front wheels 3fl and 3fr and the first rear wheel lateral force sum Fy_r of the rear wheels 3rl and 3rr. Was decided. As a result, when the road surface friction coefficient μ is accurately estimated from the relationship between the wheel slip ratio S and the braking / driving road surface reaction force Fx, the vehicle body slip angle β is re-searched and based on the accurate road surface friction coefficient μ. Thus, the slip angle β of the vehicle body can be obtained with higher accuracy.

  (5) When it is determined that the vehicle is traveling substantially straight, the searching means determines the vehicle slip angle that minimizes the difference between the second lateral force sum Fym calculated based on the lateral force estimating means and the first lateral force sum Fy. The combination of β and road friction coefficient μ was re-searched.

  Thus, the first tire slip angles α1fl to α1rr and the lateral forces Fyfl to Fyrr of each wheel are estimated based on the estimated road surface friction coefficient μ, the vehicle slip angle β, the tire characteristic map, and the like. Since the quantities representing these tire conditions can be estimated, the driving force distribution as shown in step S603 of FIG. 10 can be performed, and the vehicle behavior can be applied to the road surface friction coefficient μ and the tire condition. Highly controllable.

  (6) The search range of the vehicle slip angle β in the search means is a range including the previous search value of the vehicle slip angle β. As a result, the fact that the vehicle slip angle β does not change discontinuously in terms of vehicle motion characteristics can be utilized to limit the search range for the vehicle slip β and reduce the computation load.

  (7) The search range of the vehicle slip angle β is set to a range including the value of the vehicle slip angle β when the change speed of the vehicle slip angle β is assumed to be maintained at the previous search value. Thereby, the change information of the vehicle body slip angle β can be used, and the search range can be further optimized.

  (8) The integrated controller 30 is mounted on a vehicle that can independently distribute braking force or driving force to a plurality of wheels, and when one of the wheels 3fl to 3rr cannot sufficiently obtain the braking / driving road surface reaction force Fx. The braking / driving force is redistributed based on the estimated value m_opt of the estimated first tire slip angle α1 and the road surface friction coefficient μ. Thus, the drive torque of each wheel 3fl to 3rr is redistributed so that the lateral force Fy, yaw moment MM, and longitudinal acceleration XG of the vehicle do not change with respect to the calculated motor torque command values tTFL to tTRR. This makes it possible to further stabilize the vehicle behavior.

(Other examples)
As mentioned above, although the steering device of the present invention has been described based on the embodiments, the specific configuration is not limited to these, and the design of the steering device is not limited unless it deviates from the gist of the invention according to each claim of the claims. Changes and additions are allowed.

  For example, a vehicle that steers rear wheels may be used. In this case, the rear wheel steering angles δrl and δrr calculated in step S401 in FIG. 2 are not always set to 0, but are determined according to the detected rear wheel steering angle.

  Further, the drive mode to which the present invention can be applied is not limited to that in which the four wheels are independently motor driven as shown in FIG. The drive source may be an engine, or any of front wheel drive, rear wheel drive, and four wheel drive. A hybrid vehicle using both an engine and a motor may be used, and similar effects can be obtained by appropriately changing the embodiment according to the drive mode of the drive source.

It is a system block diagram of the electric vehicle which drives 4 wheels with an independent electric motor. It is a main flowchart of a torque command value calculation control process. It is a figure which shows the relationship between a vehicle body slip angle and each vehicle behavior parameter. It is a figure which shows the relationship between a lateral acceleration and each vehicle behavior parameter. It is a map of the ratio of slip ratio and road surface reaction force / wheel load. It is a routine of a vehicle behavior control process. It is a search / estimation control processing routine for a vehicle slip angle and a road surface friction coefficient. It is a vehicle speed-accelerator opening degree map. It is a vehicle speed-steering angle map. It is a motor torque command value calculation control processing routine.

Explanation of symbols

3 fl to 3 rr Wheel 4 fl to 4 rr Reducer 5 fl to 5 rr Drive circuit 6 Battery 8 Yaw rate sensor 11 Steering wheel 14 Steering rack 21 Steering angle sensor 22 Brake pedal sensor 23 Accelerator pedal sensor 24, 25 Lateral acceleration sensor 26 Longitudinal acceleration sensor 30 Integration controller

Claims (8)

First tire sum lateral force calculating means for calculating each of the lateral force sum of the lateral force of the front wheels and the rear wheels based on the vehicle state detection value;
Wheel load estimating means for estimating each wheel load, braking / driving reaction force estimating means for estimating braking / driving reaction force of each wheel, road surface friction coefficient and vehicle body based on estimated wheel load and estimated braking / driving reaction force of each wheel Second tire lateral force sum calculating means for calculating each of the lateral and lateral lateral force sums of the front and rear wheels with respect to the slip angle;
A search range setting means for setting a search range for a road surface friction coefficient calculated by the tire lateral force characteristic calculation means and a search range for a vehicle slip angle;
The first tire that is calculated based on the first tire sum lateral force calculating means from the second tire lateral force sum calculated based on the second tire lateral force characteristic calculating means within the search range. Search means for searching for a second tire lateral force sum that satisfies the condition that the difference from the lateral force sum is reduced;
Estimated vehicle slip angle and estimated road friction coefficient estimating means for determining a combination of an estimated vehicle slip angle and an estimated road friction coefficient from the searched second tire lateral force sum;
A vehicle state estimation device. The vehicle state estimation device according to claim 1,
The second tire lateral force characteristic calculating means includes a vehicle speed detection means, a yaw rate detection means, a steering angle detection means for detecting a steering wheel steering angle, and a vehicle body at a vehicle body reference position according to the steering angle, the vehicle speed, and the yaw rate. A vehicle state estimation device comprising: a tire slip angle calculating means for calculating each wheel tire slip angle corresponding to the slip angle. The vehicle state estimation device according to claim 2,
Straight-running determining means for determining that the vehicle is traveling substantially straight; slip rate detecting means for detecting the slip ratio of each wheel;
Road friction coefficient estimation means for calculating a road surface friction coefficient from the relationship between the braking / driving road surface reaction force estimated value of each wheel and the slip ratio when it is determined that the vehicle is traveling substantially straight. A vehicle state estimation device. In the vehicle state estimation device according to claim 3,
The straight traveling determining means determines whether or not the vehicle is traveling substantially straight based on the detected magnitudes of the left and right lateral force sum of the front wheels and the left and right lateral force sum of the rear wheels. Estimating device. In the vehicle state estimation device according to claim 3 or 4,
The search means satisfies a condition that a difference between the calculated lateral force sum calculated based on the lateral force estimating means and the detected value of the left and right lateral force sum is reduced when the vehicle is determined to be substantially straight. A vehicle state estimation device characterized by re-searching a combination of a vehicle slip angle and a road surface friction coefficient. In the vehicle state estimation device according to any one of claims 1 to 5,
The vehicle state estimation device, wherein the search means uses a range including a previous search value for the vehicle slip angle as the search range for the vehicle slip angle. The vehicle state estimation device according to claim 6,
The vehicle state estimation device according to claim 1, wherein the search means sets a range including the value of the vehicle slip angle when it is assumed that the change speed of the vehicle slip angle continues with the previous search value. In a vehicle control device comprising braking / driving force distribution control means for distributing braking / driving force to each wheel based on an estimated value of the vehicle state,
The vehicle state estimation device according to any one of claims 1 to 7, as means for estimating the state estimated value,
The braking / driving force distribution control means is configured to control braking / driving of each wheel based on the estimated value of the tire slip angle and the estimated value of the road friction coefficient when a certain wheel cannot obtain a target braking / driving road surface reaction force. A vehicle control device that redistributes power.

Images