TWI672235B - Method for tire force reserve estimation - Google Patents

Abstract

A wheel grip margin estimation method for estimating a wheel grip margin of a vehicle. The wheel grip margin estimation method continuously detects vehicle dynamic information, including longitudinal acceleration, lateral acceleration, wheel rotation angle change with time, yaw angle change with time, steering wheel steering angle, etc. A current positive force of each wheel, a current longitudinal force, and a current lateral force. Finally, using the current positive force, the road friction coefficient of the wheel relative to the road, the current longitudinal force, and the current lateral force, a longitudinal grip margin and a lateral grip margin of each wheel are obtained.

Description

Wheel grip margin estimation method

The invention relates to monitoring the driving dynamics of a vehicle, in particular to a method for estimating the grip margin of a wheel.

During the turning of the vehicle, the lateral force of the wheel is required to maintain the vehicle in the lane so that the vehicle does not deviate from the lane due to centrifugal force. At this time, part of the wheel force is occupied by the lateral force, so that the remaining maximum available wheel longitudinal force is lowered, and the wheel may be slipped and idling. In addition, in the turning lane, more lateral forces are needed during acceleration and deceleration to cope with different road conditions. At this time, the longitudinal force must be moderately reduced to meet the lateral demand, otherwise the lateral slippage of the wheel may occur. Most of the existing Advanced Driver Assistance Systems (ADAS) only issue warnings and intervene to adjust the wheel force when the slip phenomenon is detected, including the intervention of the brake system to decelerate individual wheels, or redistribute the wheels. Output).

If the remaining maximum available wheel force (including longitudinal and lateral forces) can be estimated, the ADAS system will be able to issue early warnings and even directly intervene in the control of individual wheels to avoid slippage. The above estimation can also be applied to an automatic driving system, allowing the automatic driving to perform a more intense operation (fast steering or rapid lane change) in the case where the remaining maximum available wheel force is sufficient.

In view of the above problems, the present invention proposes a wheel grip margin estimation method for estimating the grip margin of each wheel through the detection of vehicle dynamic information.

In order to achieve the above object, the present invention provides a wheel grip margin estimation method for estimating a wheel grip margin of a vehicle; wherein the vehicle has a body mass, a center of gravity and a plurality of wheels, and the wheels are at least A pair of steering wheels is included; the center of gravity has a height from the ground, the rolling direction of the tire is defined as a longitudinal direction, and the direction perpendicular to the direction of the rolling on a horizontal plane is defined as a lateral direction; the method for estimating the wheel grip margin includes: obtaining An initial positive force of each wheel on a ground; according to a longitudinal acceleration of the vehicle, lateral acceleration, mass of the vehicle, mass of one wheel of each wheel, and relative position of each wheel and center of gravity, longitudinal load on each wheel is obtained Transfer and lateral load transfer; correct initial force by longitudinal load transfer and lateral load transfer to obtain a current positive force of each wheel; according to one wheel speed of one wheel, one wheel torque, one wheel rolling effective radius a wheel moment of inertia and a wheel rotation angle change with time to obtain a current longitudinal force of each wheel; To the acceleration, a yaw angular acceleration of the vehicle, a steering angle of the steering wheel with respect to the longitudinal direction, the current longitudinal force of each wheel, and the current positive force of each wheel, the sum of the lateral forces of the wheels is obtained; a current of each wheel is obtained Lateral force; and according to the current positive force of each wheel, a road friction coefficient of each wheel with respect to the road, the current longitudinal force and the current lateral force, a longitudinal grip margin and a lateral grip force of each wheel are obtained. degree.

The invention can effectively estimate the longitudinal grip margin of the individual wheels and the lateral grip margin. When the margin is insufficient, the warning is issued early for the driver to change the operation of the vehicle, and even the advanced driver assistance system (Advanced Driver Assistance) Systems;ADAS) can intervene early on the premise of wheel slip, which can effectively improve driving safety. Wheel grip margin estimate The measurement method can be further applied to the self-driving system to estimate the driving strategy, and the self-driving system is not allowed to make an exercise decision that may cause slippage.

110‧‧‧Microprocessor

120‧‧‧Storage device

130‧‧‧Longitudinal acceleration gauge

140‧‧‧ Lateral acceleration gauge

150‧‧‧Wheel angle sensor

160‧‧‧ brake pressure sensor

170‧‧‧ Torque sensor

180‧‧‧ yaw angle sensor

190‧‧‧Steering angle sensor

S140~S180‧‧‧Steps

1 is a system block diagram of a vehicle control system for performing a wheel grip margin estimation method.

2 and 3 are flowcharts of a method according to an embodiment of the present invention.

Figure 4 is a vehicle parameter for estimation.

Figure 5 shows the speed change for the linear acceleration and deceleration process.

Figure 6 is a graph showing the longitudinal margin of each wheel as a function of time.

Figure 7 is a graph showing the lateral margin of each wheel as a function of time.

Figure 8 is a schematic illustration of a vehicle changing lane.

Figure 9 is a graph showing the longitudinal margin of each wheel as a function of time.

Figure 10 is a graph showing the lateral margin of each wheel as a function of time.

Please refer to FIG. 1 , which is a vehicle control system, which is suitable for performing a wheel grip margin estimation method disclosed in an embodiment of the present invention to perform a steering operation of a vehicle within a safe range. The vehicle control system includes a microprocessor 110, a storage device 120, a longitudinal acceleration gauge 130, a lateral acceleration gauge 140, a wheel angle sensor 150, a brake pressure sensor 160, a torque sensor 170, and a A yaw angle sensor 180 and a steering angle sensor 190.

Referring to FIG. 1 and FIG. 2, based on the above vehicle control system, an embodiment of the present invention provides a wheel grip margin estimation method for estimating a grip margin of a vehicle to prevent slippage of the wheel when steering. . The vehicle has a body mass m _s , a center of gravity, and a plurality of wheels. The body mass m _s and the position of the center of gravity can be obtained in advance by actual measurement or by vehicle specifications. The wheels are generally four, which are the right front wheel fr , the left front wheel fl , the right rear wheel rr, and the left rear wheel rl . Among them, the right front wheel fr and the left front wheel fl are usually a pair of steering wheels of the vehicle. The wheel itself also has its wheel mass m _u and the wheel rolling effective radius r _w , from which the wheel moment of inertia I _w can be calculated. The wheel grip margin estimation method can determine whether the grip of each wheel is sufficient by using four wheels as the respective wheels.

The aforementioned grip margin refers to the difference between the horizontal force of the wheel and the maximum grip of the wheel on the ground; the maximum grip of the wheel on the ground (that is, the maximum wheel force that the wheel can reach) needs to be greater than that currently provided by the wheel. Wheel force (ie the horizontal force that the wheel is subjected to, including longitudinal forces) Lateral force ), in order to avoid wheel slip.

The center of gravity of the aforementioned vehicle has a ground height h _s , the rolling direction of the vehicle tire is defined as a longitudinal direction, and the direction perpendicular to the tire rolling direction on a horizontal plane is defined as a lateral direction.

First, the microprocessor 110 obtains the initial positive force F _{zs,i of} each wheel to the ground according to the mass m _{s of the} vehicle body and the relative positions of the wheel and the center of gravity, as shown in step S110.

In the aforementioned initial positive force F _zs,i , i represents four wheels of fl , fr , rl and rr , respectively, and the so-called positive force is the vertical load of the wheel bearing the weight of the vehicle body. The relative mass position of the vehicle mass m _s , the wheel mass m _u and the center of gravity is usually a fixed value, so the microprocessor 110 can obtain the initial positive force F _{zs,i of} each wheel after one calculation, and will initially The force F _{zs,i is} stored in the storage device 120. When the microprocessor 110 re-executes the method, the initial forward force F _zs,i can be directly loaded by the storage device 120 without recalculating each time. In addition, the body mass m _s , the relative position of the wheel mass m _u and the center of gravity, and the initial positive force F _{zs, i} can be directly obtained from the outside, for example, for the vehicle type to be directly stored in the storage device 120 after being downloaded from the database.

The longitudinal acceleration gauge 130 and the lateral acceleration gauge 140 continuously detect the vehicle longitudinal acceleration a _x and the lateral acceleration a _y according to the moving state of the vehicle. In the state in which the vehicle is moving, the grounding forward force of each wheel is no longer the initial positive force F _zs,i , but the load transfer phenomenon is affected by the acceleration of the mass m _{s of} the vehicle body. Therefore, the microprocessor 110 then obtains the longitudinal load transfer W _x and the side of each wheel according to the longitudinal acceleration a _x , the lateral acceleration a _y , the vehicle mass m _s , the wheel mass m _u , and the relative positions of the respective wheels and the center of gravity. The W _yf , W _{yr is} transferred to the load as shown in step S120. The longitudinal load transfer W _x and the lateral load transfer W _yf , W _yr are estimated as follows:

Where L is the distance from the front axle to the rear axle, m _u is the mass of each wheel, l _f is the distance from the center of gravity of the vehicle to the front axle, l _r is the distance from the center of gravity of the vehicle to the rear axle, and d is the distance between the axles of each axle.

With the longitudinal load transfer W _x and the lateral load transfer W _yf , W _yr , the microprocessor 110 can correct the initial positive force F _zs,i according to the longitudinal load transfer W _x and the lateral load transfer W _yf , W _yr , Get a current positive force for each wheel , as shown in step S130.

Among them, i represents four wheels of fl , fr , rl and rr respectively.

Longitudinal force of the wheel Mainly related to the rotational dynamics of the wheel itself, including wheel speed ω _w , wheel angular acceleration And wheel torque T _w . The microprocessor 110 obtains the wheel rotation angle θ _w and the wheel speed ω _w through the wheel angle sensor 150, and obtains the wheel angular acceleration according to the change of the wheel rotation angle θ _w with time. .

Wheel angular acceleration Combined with the wheel rolling effective radius r _w and the wheel moment of inertia I _w , the current longitudinal force can be listed Related torque balance formula:

The aforementioned wheel torque T _w includes the braking torque T _b and the driving torque T _t . After the microprocessor 110 obtains the brake pressure value through the brake pressure sensor 160, the brake torque T _b can be obtained according to the brake specification. The driving torque T _t can be directly measured by the torque sensor 170 to the torque output of the vehicle power system to the wheel.

T _w = T _t + T _b (10)

Wheel torque T _w nor excluding torque sensing means with a single, acquired by directly detecting the wheel. With the estimation of the above data, the current longitudinal force can be estimated . The estimated method is to define the state vector as , input signal , the output signal is Can be rewritten as shown in equations (11), (12), and (13):

among them For the system matrix, As an input vector, Is the output vector.

Then, the state space motion model under continuous time is converted into a state space equation under discrete time, and the feedback gain value is estimated using a Kalman filter. Converted to the discrete equation (14) as follows, and then calculate the current longitudinal force of the four wheels by the Kalman estimator , , with :

As shown in FIG. 3, the wheel rolling effective radius r _w and the wheel moment of inertia I _w can be measured in advance, or downloaded and stored in the storage device 120 directly from the database according to the vehicle model specifications, and are loaded by the microprocessor 110 when needed. In. Therefore, after obtaining the wheel wheel speed ω _w , the wheel torque T _w and the wheel rotation angle θ _w with time, the microprocessor 110 can be effective according to the wheel speed ω _w , the wheel torque T _w , and the wheel rolling of each wheel. The radius r _w , the wheel moment of inertia I _w and the wheel rotation angle θ _w change with time to obtain the current longitudinal force of each wheel , as shown in step S140.

To get the current lateral force of each wheel The lateral forces of the wheels (the total force acting on all the wheels from the side in the lateral direction) are first obtained from the lateral dynamics of the body. The yaw angle sensor 180 continuously detects the yaw angle r of the vehicle, and the steering angle sensor 190 continuously detects a steering angle δ of the steering wheel with respect to the longitudinal direction. The microprocessor 110 is based on the lateral acceleration a _y of the vehicle, a change in the yaw angle r of the vehicle over time, a steering angle δ of the steering wheel relative to the longitudinal direction, and the current longitudinal force of each wheel. And the current positive force of each wheel The sum of the lateral forces of the wheels is obtained as shown in step S150. Among them, the yaw angle r changes with time, mainly used to calculate the yaw rate To further estimate the yaw rate acceleration .

The sum of the lateral forces can be established by the force balance as follows:

Then, the equations (18) and (19) can be obtained by simplifying the simultaneous equations (16) and (17), and then the values can be directly brought in:

Yaw angular acceleration Yaw angular velocity Estimated with the Kalman estimator, yaw angle r , yaw rate Yaw angular acceleration The relationship is as shown in (20):

From the above equation, the state space representation at continuous time can be obtained, as shown in equation (21):

For the state vector, Input.

Then the state space equation for continuous time is converted to the state space equation at discrete time using the forward rectangular rule, as shown in (23) and (24):

y _r,k = H _r x _r,k (24)

among them , Γ _r = G _r , w _r,k is the system's interference noise, T is the sampling time, y _r,k is the system output, H _r =[1 0] is the output matrix, and Φ _r and Γ _{r are} expressed as Equation (25):

Using the motion model described above, a Kalman filter is used in conjunction with the feedback gain matrix, since u _r,k cannot be measured, and is assumed to be 0 for closed loop state estimation, as shown in equations (26) and (27). :

among them For the estimated system state, L _r is the feedback gain matrix.

The microprocessor 110 then uses the sum of the current longitudinal forces of the wheels and the current longitudinal force of each wheel. , assigning the sum of the current lateral forces to obtain the current lateral force of each wheel , as shown in step S160. The microprocessor 110 obtains the maximum grip force of each wheel based on the current positive force and the road friction coefficient μ of each wheel with respect to the road, as shown in step S170.

Finally, the microprocessor 110 has maximum grip , Current longitudinal force Current lateral force Obtain the longitudinal grip margin of each wheel △ And lateral grip margin △ , as shown in step S180.

Since the maximum grip is limited by the road friction coefficient μ , the current longitudinal force of each wheel Current lateral force with each wheel The frictional circle relationship is used to estimate the wheel grip margin of each wheel. As shown in equations (28) and (29), the maximum longitudinal grip of each wheel is obtained first. And maximum lateral grip .

Where i represents fl , fr , rl , rr , respectively . Maximum longitudinal grip for each wheel. Maximum lateral grip for each wheel.

Next, with maximum longitudinal grip Subtract current longitudinal force Is the longitudinal wheel grip margin △ Maximum lateral grip Subtract current longitudinal force , that is, the lateral wheel grip margin △ .

△ The grip margin for the longitudinal wheel of each wheel. △ The lateral wheel grip margin for each wheel. u _p is the maximum road friction coefficient. After estimating the tire side slip angle and the steering wheel positive moment, the tire side slip angle and the estimated backwind moment are calculated. Using the stateflow judgment formula, the maximum road friction coefficient is first determined. The possible range, then find the maximum road friction coefficient and find the maximum road friction coefficient range. Or, the maximum road friction coefficient can be recorded in the database. According to the location of the road where the vehicle is located and the weather, the corresponding maximum road friction coefficient is obtained from the database.

Have a vertical grip margin △ And lateral grip margin △ The ADAS system can be used to intervene in the vehicle's travel, avoiding the wheels slipping (idling) or sideways (laterally drifting) in the longitudinal direction. For example, the microprocessor 110 can load a threshold value, the longitudinal grip margin △ Or lateral grip margin △ Less than the threshold value, a wheel slip warning is generated. The aforementioned wheel slip warning may be provided to the driver simply by means of information display, so that the driver decelerates or reduces the steering angle of the vehicle operation. Alternatively, the microprocessor 110 is involved in braking or wheel power distribution in accordance with the wheel slip warning. In the case of automatic driving, the automatic driving system corrects the vehicle speed and the planned route according to the wheel slip warning (especially the speed at which the vehicle moves laterally between the two lanes when the lane is changed).

Please refer to Figure 4, Figure 5, Figure 6, and Figure 7. For the comparison of the wheel grip margin estimation method and the CarSim (vehicle simulation analysis software) issued by Mechanical Simulation, the vehicle parameters used are shown in Figure 4. The comparison situation is the acceleration/deceleration of the vehicle in a straight line. The speed changes as shown in Fig. 5. The vehicle starts to accelerate at the initial speed of about 50km/hr and then decelerates to 40km/hr. In the figure, Vx is the second speed of the vehicle (meter/second).

As shown in Fig. 6, the front wheel is the driving wheel and the longitudinal grip force is used during the acceleration process, so the longitudinal grip margin of the front wheel decreases during the acceleration of 10 seconds, and in the deceleration after 10 seconds, There is an upward response when decelerating, and increasing the tire force margin for a moment causes the longitudinal grip margin of the front wheel to rise, followed by the longitudinal grip force margin due to the deceleration of each wheel brake, and the longitudinal grip margin is consumed. Reduce the vertical grip margin. As shown in Fig. 7, since the vehicle has no lateral dynamics, the lateral grip margin is mainly affected by the vertical load transfer. The acceleration process is mainly transferred to the rear wheel due to the vertical load, and the vertical load is mainly transferred to the rear gear during deceleration. The front wheel has a significant lateral grip margin change around 10 seconds.

Referring to FIG. 8, FIG. 9 and FIG. 10, the comparison scenario 2 is that the vehicle overtakes in the lane change. As shown in Fig. 9 and Fig. 10, the linear motion is 0 to 5 seconds, so the longitudinal grip margin is △ And lateral grip margin △ stay the same. Later, after starting the steering, the lateral grip margin of the front and rear wheels is △ There are obvious changes, and the longitudinal grip margin is changed due to the proportional relationship of the friction circle. . Finally, when the lane is changed, the longitudinal grip margin is △ And lateral grip margin △ Return to the initial state.

The invention can effectively estimate the longitudinal grip margin of each wheel △ And lateral grip margin △ When the margin is insufficient, an early warning will be given to the driver to change the operation of the vehicle, and even the Advanced Driver Assistance Systems (ADAS) can intervene early in the premise of the wheel slip, which can effectively improve driving safety. The wheel grip margin estimation method can be further applied to the self-driving system to estimate the driving strategy, and the self-driving system is not allowed to make an exercise decision that may cause slippage.

Claims (10)

A wheel grip margin estimation method for estimating a wheel grip margin of a vehicle; wherein the vehicle has a body mass, a center of gravity and a plurality of wheels, and the wheels include at least one pair of steering wheels; The center of gravity has a ground height, the rolling direction of the tire is defined as a longitudinal direction, and the direction perpendicular to the rolling direction on a horizontal plane is defined as a lateral direction; the wheel gripping margin estimation method comprises: obtaining each of the wheels An initial positive force to a ground; according to a longitudinal acceleration of the vehicle, a lateral acceleration, the mass of the vehicle body, a wheel mass of each of the wheels, and a relative position of each of the wheels and the center of gravity, Longitudinal load transfer of the wheel and lateral load transfer; correcting the initial forward force by the longitudinal load transfer and the lateral load transfer to obtain a current positive force of each wheel; according to a wheel speed of each wheel, a wheel Torque, a wheel rolling effective radius, a wheel moment of inertia, and a wheel rotation angle change with time to obtain a current longitudinal force of each of the wheels; The lateral acceleration of the vehicle, an yaw angular acceleration of the vehicle, a steering angle of the steering wheel relative to the longitudinal direction, a current longitudinal force of each of the wheels, and a current positive force of the respective wheels, obtaining a side of the wheel a sum of the force forces; obtaining a current lateral force of the respective wheels; and obtaining, according to the current positive force of each of the wheels, a road friction coefficient of each of the wheels with respect to the road, the current longitudinal force, and the current lateral force A longitudinal grip margin and a lateral grip margin for each of the wheels. The wheel grip margin estimation method according to claim 1, wherein the step of obtaining the initial positive force of each of the wheels on the ground comprises: according to the mass of the vehicle body, the relative positions of the wheels and the center of gravity, respectively The initial positive force of each of the wheels on the ground is obtained. The wheel grip margin estimation method of claim 1 or claim 2, wherein the initial positive force is stored in a storage device. The wheel grip margin estimation method according to claim 1, wherein a relative position of each of the wheels and the center of gravity includes a distance from a front axle to a rear axle of the vehicle, a distance from the center of gravity to the front axle, and The distance from the center of gravity to the rear axle. The wheel grip margin estimation method of claim 1, wherein the wheel rotation angle changes with time includes a wheel angular acceleration. The wheel grip margin estimation method of claim 5, wherein the wheel torque comprises a brake torque and a drive torque. The method for estimating a wheel grip margin according to claim 1, wherein the step of obtaining the longitudinal grip margin and the lateral grip margin of each of the wheels comprises: according to a current positive force of each of the wheels Obtaining a maximum grip force of each of the wheels with respect to a road friction coefficient of the wheel; and obtaining the longitudinal grip margin of each of the wheels with the maximum grip force, the current longitudinal force, and the current lateral force This lateral grip margin. The wheel grip margin estimation method of claim 7, wherein the maximum grip force comprises a maximum longitudinal grip force and a maximum lateral grip force. The wheel grip margin estimation method of claim 8, wherein the step of obtaining the longitudinal grip margin and the lateral grip margin comprises subtracting the current longitudinal force from the maximum longitudinal grip to obtain The longitudinal wheel grip margin; and subtracting the current longitudinal force from the maximum lateral grip to obtain the lateral wheel grip margin. The wheel grip margin estimation method according to claim 1, further comprising: generating a wheel slip warning when the longitudinal grip margin or the lateral grip margin is less than a threshold value.