WO2012023162A1

WO2012023162A1 - A vehicle

Info

Publication number: WO2012023162A1
Application number: PCT/JP2010/005152
Authority: WO
Inventors: Nobuyoshi Mutoh; Tadahiko Kato; Kazutoshi Murakami
Original assignee: Univance Corporation
Priority date: 2010-08-20
Filing date: 2010-08-20
Publication date: 2012-02-23
Also published as: WO2012023305A1

Abstract

This invention is concerning a driving torque distribution method for front-and-rear-wheel-independent-drive-type electric vehicles (FRID EVs) in which it is possible to get stable steering on a low friction coefficient road surface. This method is characterized by distributing driving torque to the left and right wheels of the front and rear wheels, considering not only load movement of the longitudinal direction but also load movement of the lateral direction which is generated at cornering. The load movements are estimated by detecting components of the 3-axis directions, i.e., longitudinal and lateral accelerations and yaw rate, and the steering angle and friction coefficient of the road surface. The effectiveness of the proposed driving torque distribution method was verified using simulators equivalent to the prototype FRID EV simulated with Matlab/Simulink and CarSim software. This method is expected to be indispensable to improving running performance of FRID EVs.

Description

A VEHICLE

This invention is concerning a vehicle.

Electric vehicles (EVs) are becoming important not only as an environmental measure against global warming, but also as an industrial policy. In order for EVs to be used widely, the development of the next generation EVs with compatible in safety and running performance is indispensable. To meet such social requirements, the conventional propulsion force generation mechanism, i. e. motor-drive structure, which strongly influences the safety and running performance, has been investigated. Many studies on front or rear one-motor-drive-type EVs (Fig. 1A) have been done from the viewpoint of economical efficiency and these EVs are already being marketed commercially. Moreover, studies on two or four in-wheel motor-drive-type EVs (Figs. 1B and 1C) have also been done from the viewpoints of the control technique and packaging. First, by focusing on the running performance, EVs of Figs. 1A and 1B cannot cope with such dangerous vehicle problems as wheel spin and wheel lock which are caused by the load movement always generated when accelerating or decelerating. When attention is paid to the safety at the time of failure, EVs of Fig. 1C have difficulties with steering ability. Since EVs of Fig.1C have more drive structures as compared with other EVs, economical efficiency and maintenance is not good, and a reliability issue may be aggravated.

JP-A-2005-184911

Fig. 1A is a plain view of a front or rear wheel drive type EV. Fig. 1B is a plain view of a front or rear two in-wheel drive type EV. Fig. 1C is a plain view of four in-wheel drive type EV. Fig. 2 is a plain view of a front-and-rear-wheel-independent-drive-type EV (FRID EV). Fig. 3 is a plain view of a situations caused by steering operations at the time of cornering. Fig. 4A is a friction circle which should be taken into consideration when distributing driving torque. Fig. 4B is a side and plain view showing difference of the torque distribution when going straight and cornering, which should be taken into consideration when distributing driving torque. Fig. 5 is a graph showing a stability operation domain of slip ratio. Fig. 6 is a flow chart explaining the basic principal of the driving torque distribution method. Fig. 7A is a side view of a moment diagram of forces acting on a vehicle in a standstill state. Fig. 7B is a side view of a moment diagram of forces acting on a vehicle in an accelerating state. Fig. 8 is a plain view of two-wheel vehicle model equivalent to a four-wheel vehicle (left turn). Fig. 9 is a rear view of roll moment which acts at the time of a right turn. Fig. 10 is a control block diagram of the torque controller when the proposed driving torque distribution method is applied to the FRID EV. Fig. 11 is a procedures for performing the driving torque control by the torque controller shown in Fig. 10. Fig. 12A is a graph showing effects of the proposed driving torque distribution method on a very low m-road when cornering on a low m-load (m = 0.2) at steering angle=3 deg ,and lateral force is not distributed to the front and rear wheels properly. Fig. 12B is a graph showing effects of the proposed driving torque distribution method on a very low m-road when cornering on a low m-load (m = 0.2) at steering angle=3 deg ,and lateral force is distributed to the front and rear wheels properly. Fig. 13A is a graph showing vehicle trajectories of effects of the proposed driving torque distribution method on vehicle trajectories under the same simulation conditions as Fig.12. Fig. 13B is a perspective view showing road condition used for simulations of effects of the proposed driving torque distribution method on vehicle trajectories under the same simulation conditions as Fig.12. Fig. 14A is a graph showing effects of the proposed driving torque distribution method when starting at a corner on an ultra low mu-road while accelerating (mu= 0.1, steering angle=3 deg) without proposed torque distribution method. Fig. 14B is a graph showing effects of the proposed driving torque distribution method when starting at a corner on an ultra low mu-road while accelerating (mu= 0.1, steering angle=3 deg) with the slip ratio control. Fig. 14C is a graph showing effects of the proposed driving torque distribution method when starting at a corner on an ultra low mu-road while accelerating (mu= 0.1, steering angle=3 deg) with proposed torque distribution method. Fig. 14D is a graph showing effects of the proposed driving torque distribution method when starting at a corner on an ultra low mu-road while accelerating (mu= 0.1, steering angle=3 deg) on vehicle trajectories when cornering while accelerating. Fig. 15A is a graph showing effects of the proposed driving torque distribution method when cornering on a high mu- road while accelerating (steering angle=3 deg, mu=0.75) without the proposed torque distribution method. Fig. 15B is a graph showing effects of the proposed driving torque distribution method when cornering on a high mu- road while accelerating (steering angle=3 deg, mu=0.75) with the proposed torque distribution method. Fig. 15C is a graph showing effects of the proposed driving torque distribution method when cornering on a high mu- road while accelerating (steering angle=3 deg, mu=0.75) on vehicle trajectories.

After proposing front and rear wheel independent drive type electric vehicles FRID EVs (Fig. 2) compatible in safety and running performance, their research is being done from various angles by positioning them as a next-generation EV. As shown in Fig. 2, a electric vehicle 1 includes a front right wheel 2fr, a front left wheel 2fl, a rear right wheel 2rr, a rear left wheel 2rl, a front motor 3f, a rear motor 3r,

differential gears

4f, 4r, a battery 7, a front inverter 8f, a rear inverter 8r, a front driving circuit 9f, a rear driving circuit 9r, a torque controller 10,

encoders

16f,16r, a steering wheel 19, cameras 20fr,20fl, an accelerator pedal sensor 22, a brake pedal sensor 23, a shifter sensor 24, a vehicle body 25 , 3-axis acceleration sensor 26, angular speed sensors 28fr, 28fl, 28rr, 28rl, a steering angle sensor 29. It has been clarified through vehicle dynamics analysis experiments that outstanding running performance is obtained using a structural feature which can freely control longitudinal force of the front and rear wheels 2fr, 2fl, 2rr, 2rl according to the running and road surface conditions. In FRID EVs, distribution of the lateral force to the right and left tires of a front wheel 2fr, 2fl and a rear wheel 2rr, 2rl is performed through a

differential gear

4f, 4r like in ordinary gas-powered vehicles. Accordingly, in currently available vehicles, since sufficient lateral force required for revolution cannot be secured, under-steering (Fig. 3) is apt to be caused when cornering on a low mu-road or when cornering at high speed on a dry road. It is only the four in-wheel-motor-drive-type EV (Fig. 1C) that can directly handle the longitudinal and lateral forces of four wheels. However, in-wheel-motor-drive-type EVs have the serious problems cited above as a next generation EV.

Thus, a driving torque distribution method suitable for FRID EVs which can secure sufficient lateral forces of the front and rear wheels is proposed here. The lateral forces required for revolution are estimated based on the conditions regarding the friction circle in consideration of the longitudinal and lateral load movements. Here, the effectiveness of the proposed driving torque distribution method is verified using a simulator equivalent to an actual prototype FRID EV.

<Principal of Driving Torque Distribution at the Time of Cornering>
When vehicle 1a speed is in a steady state (which can include the standstill state), front and rear tire loads (normal forces) F_{zd_f} and F_{zd_r} act on each tire of the front and rear wheels 2 ,i.e. 2fr, 2fl, 2rr, 2rl. In this case, the vehicle 1a is driven by the front and rear driving forces F_{xd_f}and F_{xd_r}which are supplied from front and

rear motors

3f, 3r, respectively. Shifting to an acceleration mode, the longitudinal load movement Z_xto the rear wheels 2rr, 2rl from the front wheels 2fr, 2fl is generated. As a result, front and rear tire loads F_{zd_f} and F_{zd_r} are changed to F_{z_f}(=F_{zd_f} - z_x) and F_{z_r}(=F_{zd_r}+ z_x) by z_x (Fig.4B). Therefore, in order for the vehicle 1b to maintain ideal running based on vehicle dynamics, driving torque distribution to the front and rear wheels 2 should be done so that the front and

rear motors

3f, 3r can generate the proper driving forces F_{x_f} and F_{x_r} corresponding to the front and rear tire loads F_{z_f} and F_{z_r}which are changed by the load movement.

Next, when starting cornering, the new lateral load movement z_y takes place between the left wheels 2fl,2rl and right wheels 2fr,2rr. Front and rear tire loads F_{z_f} and F_{z_r} are further changed by the produced load movement z_y . For example, when cornering to the right, the left and right tire loads F_{z_fl} and F_{z_fr} , F_{z_fl} and F_{z_fr} which act on the left and right tires of the front and rear wheels 2 are changed to (F_{z_fl} + z_y ) and (F_{z_fr} - z_y), (F_{z_rl} + z_y ) and (F_{z_rr} - z_y), respectively. Here, subscripts l and r indicate left and right tires, respectively. Thus, in order for vehicles to corner stably, it is necessary to always secure the front and rear lateral forces corresponding to lateral load movement z_y. To do so, driving torque distribution should be performed according to the fact that each of the driving and braking forces and the cornering force cannot exceed the friction force muW(mu: friction coefficient, W: tire load) (Fig. 4A). That is, specifically, after securing lateral force required for revolution, in a friction circle, the maximum F_{x_j_max} of the longitudinal force needed to propel the vehicle is secured by

Hereafter, in this study, the driving torque distribution uses the assumption that if slip angle is small, the lateral force almost agrees with the cornering force. The maximum longitudinal force F_{x_j_max} (j = l: front, j = r : rear) obtained for each of the front and rear wheels 2 is converted by

Here k_j (j = l and r) is a torque conversion gain for front and rear wheels 2 that get a match with the torque reference tau_R for the whole vehicle as generated from an accelerator which is split between the front and rear wheels 2 by

Here, tau_Rf and tau_Rr are the front and rear torque references split from tau_R based on the load movement. The front and rear driving torque references tau_Rf ^* and tau_Rr ^* are determined through comparisons between tau_Rf and tau_{x_f} , and between tau_Rr and tau_{x_r} by the next procedures:

Where tau_{x_f} and tau_{x_r} are the front and rear driving references determined from the front and rear lateral forces F_{y_f} and F_{y_r}.

Finally, in order to secure the stability of the front and rear wheels 2 the above driving torque references tau_Rf ^* and tau_Rr ^* are compensated for according to the following slip ratio conditions (Fig. 5). That is,

Here r_eff is effective radius of a tire; V is vehicle speed; omega_{f_l} and omega_{f_r} are angular speeds of the left and right tires of the front wheels 2fr,2fl; and omega_{r_l} and omega_{r_r} are angular speeds of the left and right tires of the rear wheels 2rr,2rl. Under the slip ratio conditions of other than (6) and (7), the torque references tau_Rf ^* and tau_Rr ^* to satisfy the procedures of (4) and (5) become the actual torque references for the front and rear torque controllers.

As shown in the flow chart of Fig. 6 (S1 - S13), according to the steering angle delta, the driving torque references are distributed to the front and rear wheels 2, dividing the two driving torque distribution procedures into a procedure to determine the distribution from lateral forces and a procedure to determine the distribution from longitudinal forces based on the torque reference generated from accelerator. Hereafter, each procedure is described in detail.

<A Procedure to Determine Driving Torque Distributed to the Front and Rear Wheels Based on Torque Reference Obtained from the Accelerator>
The main point of this procedure is to distribute longitudinal forces to the front and rear wheels 2 in consideration of load movement caused by acceleration and deceleration. Then, first, longitudinal load movement z_x is derived using a moment diagram of forces acting on a vehicle which is in a standstill state (Fig. 7A) and an accelerating state (Fig. 7B). In a standstill state, the vehicle load which acts on the center of gravity is distributed to the normal forces F_{zd_f} and F_{zd_r} on the front and rear tires by

When vehicles are accelerated, the force moments acting on the contacting points (X and Y) of the front-and-rear wheels 2 are given by

and

where vehicle weight F_mg = mg (m: vehicle mass, g: acceleration due to gravity); alpha_car: longitudinal acceleration; H_car : height of the center of gravity; L_car : wheelbase of a vehicle; and L_f: length between the axles of the front wheels 2fr,2fl and the center of gravity. Therefore, the normal forces (F_{z_f} and F_{z_r} ) which act on the tire of the front and rear wheels 2 are expressed by

and

Using F_{z_f} , F_{z_r} and friction coefficient mu of a road surface, the longitudinal forces (F_{x_f} and F_{x_r} ) acting between road surfaces and the tires of the front and rear wheels 2 are given by

and

Then, so that the front and

rear motors

3f, 3r can generate the driving torques corresponding to these longitudinal forces, torque references tau_Rf and tau_Rr are distributed to the front and rear torque controllers by

and

Here the front distribution ratio R_f is given by

<A Procedure to Determine Driving Torque Distribution Based on Lateral Force Required for Cornering>
Here, using Fig. 8, movement of vehicles is described, by transposing to a two-wheel model equivalent to the four-wheel model of vehicles in general. Assuming that the steering angle delta and the slip angle beta of vehicle are small, the lateral motion of the vehicle and the yaw dynamics at the center of gravity of the vehicle are studied that can be handled using longitudinal and lateral accelerations alpha_x and alpha_y, and yaw rate gamma detected from the 3-axis acceleration sensor 26 installed at the center of gravity of the vehicle. The lateral motion for front and rear wheels 2 is expressed by

Taking the moment balance about the z-axis into consideration, the equation for yaw dynamics is given by

where I_z is yaw moment of inertia about the z-axis. Solving (18) and (19) about F_{y_f} and F_{y_r} , the front and rear lateral forces are derived as

and

Next, the lateral load movement z_y appearing during cornering is obtained by considering roll moment at the center of gravity of the vehicle shown in Fig. 9. Moment balance at the center of gravity of the vehicle yields the equation for the normal force F_{z_l} on the left tires of the front and rear wheels 2 as

Accordingly, using the obtained F_{z_l} and (9), the lateral load movement z_y is obtained by

By using the derived z_y, the left-and-right-normal forces F_{z_fl} and F_{z_fr} on tires for the front wheels 2fr,2fl, and the left-and-right-normal forces F_{z_rl} and F_{z_rr} on tires for the rear wheels 2rr,2rl are given as follows:

and

Considering that the front and rear lateral forces 2F_{y_f} and 2F_{y_r} of (20) and (21) are divided with ratios of F_{z_fl} : F_{z_fr} , and F_{z_rl} : F_{z_rr} ,, respectively, the left and right lateral forces F_{y_fl} and F_{y_fr} of the front wheels 2fr,2fl and the left and right lateral forces F_{y_rl} and F_{y_rr} of the rear wheels 2rr,2rl are expressed by

and

Finally, applying these lateral forces to (1), the maximum F_{x_j_max} of the longitudinal force needed to propel the vehicle is obtained.

The above procedures are applied to the torque controller 10 shown in Fig. 10. As shown in Fig. 10, the torque controller 10 includes, a part for longitudinal and lateral force splitter using friction circle 110, a part for torque reference optimization 120,

wheel speed detectors

131f,131r, a front torque controller 132f, a rear torque controller 132r, and a vehicle speed calculator 133.

The part for longitudinal and lateral force splitter using friction circle 110 includes, a longitudinal torque splitter for front and rear wheels calculation part 111, a longitudinal load movement calculation part 112, a lateral load movement calculation part 113, a lateral force for front and rear wheels calculation part 114, a road surface friction coefficient estimator calculation part 115, a maximum driving torque estimator using friction circle calculation part 116, and a optimal driving torque reference discriminator calculation part 117. Cameras 20fr,20fl pick-up images of a road surface in front of the front wheels 2fr,2fl. The road surface friction coefficient estimator calculation part 115 decides a road surface condition such as a dry road, a wet road, or a frozen road based on the images picked-up by the cameras 20fr,20fl, and estimates a friction coefficient of the road surface in front of the front wheels 2fr,2f based on the road surface condition.

The part for torque reference optimization 120 includes, a front torque reference generator 121, a rear torque reference generator 122, a front slip ratio control part 123, and a rear slip ratio control part 124. The front slip ratio control part 123 includes a front slip ratio calculator 123a, a front stability judgment part 123b, and a front torque reference compensator 123c. The rear slip ratio control part 124 includes a rear slip ratio calculator 124a, a rear stability judgment part 124b, and a rear torque reference compensator 124c.

In this controller, the driving torque references are determined based on the signal generated from the accelerator, longitudinal and lateral accelerations, yaw rate and steering angle according to the torque control flow of Fig. 11 (S21 - S36).

<Verification of the Proposed Driving Torque Distribution Method by Simulations>
Simulations are performed using a prototype FRID EV with the following specifications. m: 1900 kg; H_car: 670 mm; L_f:1500 mm, L_r:1125 mm; L_car: 2625 mm. Fig. 12 shows simulation results when turning toward the left on a low mu-road (mu= 0. 2) with a 3-deg steering angle at 3 s after accelerating. First, the proposed driving torque distribution method is evaluated through simulations under the severe driving condition of turning to the left on the low mu- road. When the proposed method is not applied (Fig. 12A), at around 2 s after stating cornering the slip angle of the front wheels 2fr,2fl is increasing gradually and the front and rear lateral forces are saturated. As a result, the vehicle shifts from the traveling lane, without the ability to corner (Fig.13). On the other hand, when the proposed method is applied, the saturation of lateral forces is suppressed by making driving forces of the front and rear wheels 2 decrease with cornering. Since the lateral forces required for cornering are secured (Fig. 12B), the vehicle can turn to the left along the traveling lane (Fig.13).

Next, the proposed method is evaluated under severer conditions that make the vehicle accelerate while cornering on an ultra low mu- road (mu= 0. 1). The difference in the cornering performance between the proposed method and the conventional slip control is also made clear. Fig. 14 shows the simulation results under the conditions that the vehicle turns to the left at a 3-deg steering angle while accelerating on the ultra low mu-road at time t₁ at about 10 s after stating. When the proposed method is not applied (Fig. 14A), the slip ratio immediately increases to 1. Then, the wheel spin occurs, and a skid is caused. Since sufficient lateral forces required for revolution cannot be secured, the vehicle strays from the traveling lane without being able to corner (Fig. 14D). Moving to the case of the slip ratio control, when making a vehicle accelerate, slip ratios are soon increased to 1, and although once they lead to the wheel spin state, they are suppressed to less than 0. 2 and then the wheel spin state is avoided. However, the slip angle increased gradually (Fig. 14A), and eventually, the vehicle cannot turn to the left (Fig. 14D)). On the other hand, when the proper driving torque is applied to front and rear wheels 2 using the proposed driving torque distribution method, the slip ratio is kept at a value near zero and skidding is also not seen. Then, lateral forces required for revolution are also secured, and the vehicle can turn to the left properly along the traveling lane (Figs. 14C and 14D).

Finally, the effects when turning to the left at high speeds on the high mu-road (mu= 0. 75) are investigated under the conditions of beginning to turn to the left at the time t₁ while accelerating. When the proposed driving torque distribution method is not applied, the vehicle can turn to the left properly up to corner B along the traveling lane, 10s after starting to accelerate (Figs.15A and 15C). However, it is impossible to secure the lateral forces required for revolution because they are gradually saturated after passing B point. As a result, the vehicle cannot run along the road, and it is accelerating too much. On the other hand, when the proposed driving torque distribution method is applied, lateral forces required for revolution are secured by decreasing the front and

rear motor

3f, 3r torques in accordance with increase in speeds, and the turn to the left can be effectively carried out to bring the vehicle to the final destination.

In the above simulations, steering angle is kept constant, and adjustment of the accelerator by the driver is not taken into consideration. Although in fact, since

motor

3f, 3r torque is also adjusted by steering angle through accelerator operations by the driver, leaving the traveling lane does not occur during running. However, these simulation results mean that the proposed driving torque method allows drivers to get more stable and safe revolution on low mu- roads or at high speeds.

<CONCLUSION>
When cornering on low mu-roads or at high speeds, it is very difficult for all conventional vehicles to perform steering for revolution. This paper described a driving torque distribution method to solve these problems using the FRID EV structural feature that it can freely distribute driving forces to the front and rear wheels 2 according to road surface and running conditions. The driving torque distribution method is characterized by distributing the front and rear torques to the front and

rear motors

3f, 3r so that lateral forces required for revolution can be secured based on the information about steering angle, friction coefficient, and the lateral and longitudinal accelerations. The effectiveness of the proposed driving torque distribution method was verified through simulations about the cornering performance on low mu-roads and at high speeds.

Furthermore, in addition to the method of controlling the FRID EVs which was previously developed, the method proposed here has added a still more powerful function to FRID EVs from the viewpoints of safety and running performance.

Although the case of controlling the driving force has been described in the above-mentioned embodiment, it is possible to apply the present invention to the case of controlling the braking force in the same manner. In this case, when the torque controller 10 controls a braking force, a sign of z_x becomes "minus".
Further, although the case of controlling when turning to the left has been described in the above-mentioned embodiment, it is possible to apply the present invention to the case of controlling when turning to the right in the same manner. In this case, when turning to the right, a sign of z_y becomes "minus".
Furthermore, a sign of turning acceleration is "plus" for indicating accelerating and the sign of turning acceleration is "minus" for indicating decelerating.

This invention can be applied to a vehicle such as a passenger car, a bus and a truck.

1 electric vehicle
2fr front right wheel
2fl front left wheel
2rr rear right wheel
2rl rear left wheel
3f front motor
3r rear motor
4f,4r differential gear
7 battery
8f front inverter
8r rear inverter
9f front driving circuit
9r rear driving circuit
10 torque controller
16f,16r encoder
19 steering wheel
22 accelerator pedal sensor
23 brake pedal sensor
24 is shifter sensor
25 vehicle body
26 3-axis acceleration sensor
28fr,28fl,28rr,28rl angular speed sensor
29 steering angle sensor
110 section for longitudinal and lateral force splitter using friction circle
120 section for torque reference optimization
131f,131r wheel speed detector
132f front torque controller
132r rear torque controller
133 vehicle speed calculator

Claims

A vehicle, comprising:
a first electric motor configured to transmit a driving or braking force to a right front wheel and a left front wheel through a first differential unit;
a second electric motor configured to transmit a driving or braking force to a right rear wheel and a left rear wheel through a second differential unit;
a friction coefficient estimator configured to estimate a friction coefficient of a driving road surface,
a tire load calculation unit configured to calculate a tire load of each wheel based on longitudinal load movement generated during acceleration or deceleration and lateral load movement generated during turning to left or right; and
a control unit configured to control the driving or braking forces of the first and second electric motors, thereby controlling a resulting combined force of a longitudinal force and a lateral force of each wheel to fall within a friction circle , wherein the friction circle is determined from the friction coefficient and the tire load of each wheel.
The vehicle according to claim 1, comprising:
a slip ratio calculation unit configured to calculate a slip ratio of each wheel,
wherein the control unit starts a control operation when the slip ratio exceeded a predetermined value,
wherein the control unit controls the driving or braking forces of the first and second electric motors, thereby controlling the resulting combined force of the longitudinal force and the lateral force of each wheel to fall within the friction circle and to reduce a slip ratio of the each wheel.
The vehicle according to claim 2, wherein the friction coefficient estimator estimates the friction coefficient of the road surface in front of the front wheels based on images picked-up,
wherein the control unit controls the driving or braking force of the second electric motor using an information including the friction coefficient.