See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/340631562

Research on Yaw Stability Control of Unmanned Vehicle Based on Integrated

Electromechanical Brake Booster

Conference Paper · April 2020

DOI: 10.4271/2020-01-0212

CITATION READS

1 21

6 authors, including:

Jian Wu
Jilin University
104 PUBLICATIONS   373 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Studies on Integrated Modeling and Control Methods for Intelligent and Electric Vehicles View project

All content following this page was uploaded by Jian Wu on 23 March 2022.

The user has requested enhancement of the downloaded file.

 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

2020-01-0212 Published 14 Apr 2020

Research on Yaw Stability Control of Unmanned

Vehicle Based on Integrated Electromechanical
Brake Booster
Dexing Lao, Jian Wu, Rui He, Bing Zhu, Jian Zhao, and Zhicheng Chen Jilin University

Citation: Lao, D., Wu, J., He, R., Zhu, B. et al., “Research on Yaw Stability Control of Unmanned Vehicle Based on Integrated
Electromechanical Brake Booster,” SAE Technical Paper 2020-01-0212, 2020, doi:10.4271/2020-01-0212.

Abstract
determined by the feedback PI control algorithm. Secondly,

T
he Electromechanical Brake Booster system (EMBB) the distribution layer optimally selects the wheel with the
integrates active braking and energy recovery and highest differential braking efficiency and assigns the brake
becomes a novel brake-by-wire solution that substitutes pressure to the corresponding wheel based on the additional
the vacuum booster. While the intelligent unmanned vehicle yaw moment calculated by the decision layer. In the execution
is in unstable state, the EMBB can improve the vehicle yaw layer, the servo motor unit employs the position-current
stability more quickly and safely. In this paper, a new type of double closed-loop PI algorithm to achieve rapid building-up
integrated EMBB has been designed, which mainly includes pressure and then precise control of the single-wheel pressure
two parts: servo motor unit and hydraulic control unit. is achieved by controlling the electromagnetic valve of the
Aiming at the dynamic instability problem of intelligent hydraulic unit. Finally, with the dSPACE products, we built
unmanned vehicle, a three-layer vehicle yaw stability control the hardware-in-the-loop (HIL) test bench based on integrated
structure including decision layer, distribution layer and EMBB. The proposed control strategy is verified by using
execution layer is proposed based on integrated EMBB. Firstly, Similator to run CarSim in real-time and MicroAutoBox to
the decision layer calculates the ideal yaw rate and the side run the algorithm. The HIL experiment results show that the
slip angle of the vehicle with the classic 2DOF vehicle dynamics yaw stability control algorithm based on integrated EMBB
model. The boundary of the stable region is determined by can effectively restore the unstable vehicle to the stable driving
the phase plane method and the additional yaw moment is zone and enhance the driving safety of the complete vehicle.

Introduction
Therefore, a new kind of brake system named

W
ith the revolution of automotive electrification and Electromechanical Brake Booster system (EMBB) utilizing a
intelligence, the traditional brake system using motor to drive the transmission mechanism to establish the
vacuum booster can no longer meet the require- master cylinder pressure-building, and then realizing brake
ments of active braking, compound braking and other force boost through semi-decoupling or decoupling method
advanced functions [1, 2, 3]. Therefore, more and more vehicles has received extensive attention. In [6,7,8], components
are beginning to equip with the brake-by-wire system (BBW). suppliers Bosch and Continental Teves have developed several
Among the various new brake system schemes emerged, they types of EMBB system such as iBooster and MKC-1. And
can be generally divided into two categories. The first one is Tongji University also design the integrated electro-hydraulic
the BBW of electro-hydraulic brake (EHB) system, which braking system (I-EHB) [9,10].
mainly aims to improve the hydraulic brake system based on As a promising trend in the development of brake systems
the traditional automobile. The second one is the BBW of on smart vehicles in the future, the advantages of vehicles
electro-mechanical brake (EMB) system. This type of brake equipped with the EMBB are more obvious in terms of active
system is equipped with a set of motor and reduction mecha- safety and stability control. Firstly, compared with the tradi-
nism systems as braking actuators on each wheel. However, tional electronic stability program (ESP) on active braking,
the EHB equipped with high-pressure accumulators and elec- the EMBB system has distinct advantages in terms of pressure-
tromagnetic linear valves has the disadvantages of high cost building speed and maximum pressure-building capacity.
and risk of brake fluid leakage. Some EMB systems worked Secondly, in terms of the vehicle stability control, the ESP
with 42V power pose a challenge to the current vehicle power using DC motor is not feasible for working continuously. In
supply system. At the same time, EMB cannot use actuation contrast, the EMBB could obtain fast response and continuous
energy from the driver for backup function and can require pressure building-up with superiority of high-speed and
more complex backup systems to comply with relevant regula- locked-rotor of high-performance permanent magnet
tions [4,5]. synchronous motor (PMSM). As developing trend of new
 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

2 RESEARCH ON YAW STABILITY CONTROL OF UNMANNED VEHICLE BASED ON INTEGRATED ELECTROMECHANICAL

energy vehicles, there will be more and more vehicles equipped  FIGURE 1   The schematic graph of the integrated EMBB
with the integrated EMBB system. Since it can independently
adjust the pressure of four-wheel cylinders by its PMSM and
hydraulic control unit, the function of stability control can
be accomplished by the integrated EMBB. Compared with the
ESP using DC motors, the integrated EMBB will have faster
response time and higher control precision in terms of vehicle
stability control.
Currently, there are many researches focusing on the
EMBB pressure control. In [11], the feedforward plus feedback
pressure control strategy is adopted for the electro-hydraulic
servo booster system matched on the plug-in hybrid electric
vehicle. In [12], adaptive position-pressure control is applied
for the BBW. And about vehicle yaw stability control, in [13]
the vehicle stability is analyzed according to the response of
vehicle yaw rate. In [14], the phase plane of yaw rate and side
slip angle is employed for vehicle yaw stability analysis.
However, there are few researches about vehicle stability
control with EMBB. Based on the EMBB designed by our team
[15, 16, 17, 18], this paper proposes a type of integrated EMBB
system and designs a hierarchical yaw stability controller

© SAE International.
including decision layer, distribution layer and execution layer
to improve the yaw stability of the vehicle with integrated
EMBB system. Firstly, the decision layer utilizes the 2DOF
vehicle dynamics model to calculate the ideal yaw rate and
side slip angle of the vehicle. The boundary of the stable region
is determined by the phase plane method and the additional
yaw moment is determined by the feedback PI control algo- On the whole, the integrated EMBB can be divided into
rithm. Secondly, the distribution layer optimally selects the the pressure building unit and the pressure control unit. The
wheel with the highest differential braking efficiency and function of the pressure building unit is to precisely establish
assigns the brake pressure to the corresponding wheel based the master cylinder pressure through identifying the driver's
on the additional yaw moment calculated by the decision layer. braking intention for the working mode of basic brake force
In the execution layer, the servo motor unit employs the posi- boost. The pressure building unit is mainly composed of a
tion-current double closed-loop PI algorithm to achieve rapid PMSM and its transmission mechanism, a master cylinder,
pressure building-up and then precise pressure control of the an input rod, an output rod, a displacement sensor, a encoder,
single-wheel is achieved through the electromagnetic valve of a ECU, a brake fluid reservoir, etc. The function of the pressure
the hydraulic control unit. control unit is to realize the pressure distribution of each
The rest of the paper is organized as follows. Section 2 wheel cylinder through the cooperation of the relevant elec-
introduces the main components and working processes of tromagnetic valves, which mainly includes DC motor, piston
the integrated EMBB. Section 3, the hierarchical yaw stability pump, inlet valve, outlet valve and low pressure accumulator.
controller is designed, which includes decision layer, distribu- Under normal working conditions, the brake fluid in the brake
tion layer and execution layer. Section 4 the HIL experimental master cylinder passes through the brake pipe, flows through
bench is built and the effectiveness of the control algorithm the inlet valve into each wheel cylinder to establish the wheel
is verified through rapid control prototype (RCP). Section 5 cylinder pressure, thereby forming the braking torque and
contains the summary of the full text. completing the braking function.
Among them, as the core component of the integrated
EMBB scheme, the structural diagram of EMBB is shown
below in Figure 2 and Figure 3.
The Structure of As shown in the Figure 2 and Figure 3, the EMBB is
mainly composed of PMSM, two-stage reduction mechanism
Integrated EMBB (a pair of gear pairs and a pair of screw-nut), reaction disk,
servo body, input rod, output rod return spring and
From the functional point of view, the currently proposed master cylinder.
brake system scheme has three options: 3-Box, 2-Box, 1-Box. When the EMBB is in the basic brake force booster mode,
The integrated EMBB proposed in this paper includes the the driver pushes down the brake pedal and the input rod
function of establishing the master cylinder pressure and moves forward and squeezes the main surface of the reaction
adjusting the wheel cylinder pressure. Moreover, the pressures disk, at the same time ECU judges the braking intention
of four wheels cylinder can be independently regulated. So it (apply, hold, and release) of the driver via a pedal rod displace-
belongs to the 1-Box scheme. The configuration of integrated ment sensor and then the PMSM is controlled to rotate. Then
EMBB is shown in Figure 1. the PMSM rotational torque is converted into the horizontal
 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

RESEARCH ON YAW STABILITY CONTROL OF UNMANNED VEHICLE BASED ON INTEGRATED ELECTROMECHANICAL 3

 FIGURE 2   The physical drawing of the EMBB The three-layer yaw stability control structure includes
decision layer, distribution layer and execution layer. The
upper controller decision layer inputs real-time vehicle state
information and road surface attachment, etc. and outputs
the target yaw moment to the distribution layer. In the distri-
bution layer, according to the current vehicle state and the
target yaw moment determined by the upper controller,
utilizing the differential braking, selecting one wheel at the
same time to apply the braking torque then converting the
© SAE International.

target yaw moment as the target wheel cylinder pressure to

the execution layer. The execution layer, as the bottom level
controller, receives the target wheel cylinder pressure, then
follows the target wheel cylinder pressure through the master
cylinder and the wheel cylinder pressure control strategy,
finally completes the yaw stability control of the whole vehicle.
 FIGURE 3   Internal structure of the EMBB

Decision Layer
In the process of vehicle stability analysis, especially yaw
stability analysis, the classic 2DOF vehicle dynamics models
are applied, as shown in Figure 5.
Where: L is the wheelbase, a is the front wheelbase, b is
the rear wheelbase, Fy2 is the lateral force of rear axle, αr is the
side slip angle of rear tire, ω is yaw rate, Vy is vehicle lateral
© SAE International.

speed, β is side slip angle, Vx is vehicle longitudinal speed, Fy1

is the lateral force of front axle, αf is the side slip angle of front
tire, δ is the front wheel angle.
Among them, the lateral forces Fy1 and Fy2 are calculated
by the linear 2 DOF vehicle tire models,

force on the servo body through the two-stage reduction ìFy 1 = K f a f

í (1)
mechanism, then the servo force of the servo body and the îFy 2 = K r a r
input force imposed by the driver are coupled to the reaction
The ideal value of the vehicle motion parameter repre-
disk to push the output rod to realize the master cylinder
sents the ideal motion state of the vehicle, and that ideal values
pressure. When the EMBB system is in the failure mode, the
are generally considered to be the desired response to the
force comes from the driver directly acts on the main surface
driver's steering input. When the vehicle is not in the extreme
of the reaction disk, pushing the output rod to complete pres-
state, the linear 2DOF vehicle dynamics model can be used
sure-building, and implementing the failure backup function.
to calculate the ideal yaw rate and side slip angle of the vehicle.
This gives the ideal vehicle yaw rate value at the steering
wheel angle input by the driver,
Three-Layer Vehicle Yaw Vx
wdes = d (2)
Stability Controller L(1 + KVx 2 )
a
To improve the dynamics safety of the vehicle under extreme b- mVx 2
conditions, based on the vehicle equipped with the integrated bdes = K r
× d (3)
EMBB system, this paper proposes a hierarchical control L × (1 + KVx 2 )
strategy, as shown in Figure 4.
 FIGURE 5   Vehicle dynamic model with 2DOF

 FIGURE 4   The structure drawing of three-layer controller

© SAE International.

© SAE International.
 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

4 RESEARCH ON YAW STABILITY CONTROL OF UNMANNED VEHICLE BASED ON INTEGRATED ELECTROMECHANICAL

Where: ωdes is the desired yaw rate value, K is the stability TABLE 1  Weight coefficient K value
mæ a b ö Side slip angle β(deg) 0 1 2 3 4
factor of the vehicle, = 2 ç - ÷ , Kf and Kr are the front
L è Kr K f ø Weight coefficient K 0 0 0 0.1 0.3
and rear tire cornering stiffness, m is the sprung mass of © SAE International.

the vehicle.
Considering the influence of adhesion, the upper limit of The yaw moment decision algorithm based on the side
the desired yaw rate value and side slip angle is, slip angle is expressed as,
mg t
wmax = 0.85 ×
Vx (4)
ò
DM b = K pb × e b ( t ) + K i b e b ( t ) dt (9)
b max = tan ( 0.02 m g )
-1 0

Where, eβ is the distance from the actual state point of

Therefore, the desired values for both can be obtained,
the vehicle to the stable boundary, which can be expressed by
ì ü the following formula,
ï Vx Vg ï
wdes = min í d , 0.85 × x ý × sign (d )
(
ïî l 1 + KVx
2
) Vx ïþ ì g
g
ïï B1 × b0 + b0 - m × B2 B1 × b0 + b0 > m × B2
eb = í (10)
ì a ü 1 + B12 g
ï
ïï b - K mVx
2
ïï B1 × b0 + b0 £ m × B2
ïî 0
bdes = min í r
× d , tan -1
( 0 .02 m g ) ý .sign (d ) (5)
(
ï L × 1 + KVx
2
) ï In summary, we can get the desired yaw moment as,
ïî ïþ
DM = (1 - K ) DM w + K DM b (11)
The yaw rate characterizes the vehicle's ability to follow
the driver's steering intention, and the side slip angle reflects In the formula, K is the weight coefficient, and the values
the degree of vehicle instability. When the vehicle is in the are as shown in Table 1. Considering the limitation of the
region with high linearity (at moment, lateral acceleration of bicycle model, the side slip angle should not be larger than 4
the vehicle is generally smaller than 0.4g), yaw rate is regarded degrees. Therefore, when the side slip angle is relatively small
as the control variable. When the vehicle is at a lower linearity (smaller than 4 degrees), the decision layer can calculate the
state, side slip angle become the control variable. Therefore, yaw moment to accomplish the vehicle stability control. That
we  should determine the control principle of the vehicle is, when the vehicle's side slip angle is relatively small, the
stability by an effective method. vehicle stability is mainly controlled by the vehicle yaw rate.
From the method mentioned in [19], we can determine When the vehicle's side slip angle is comparatively large
the stability boundary of the vehicle through the side slip angle (smaller than 4 degrees), the vehicle stability is controlled by
phase plane method. As shown in the following example, the yaw rate and side slip angle.
B1b + b = m × B2 (6)

In the formula, B1 and B2 are the constant of stability Distribution Layer

boundary, the values are 0.5 and 0.03 respectively. Therefore, The desired additional yaw moment is obtained in the decision
we can determine the additional yaw moment needed to bring layer, and then the distribution layer is responsible for
the current vehicle back to the steady state. receiving the target yaw moment sent by the upper controller
Firstly, we calculate the expected yaw rate and side slip and judging the current state of the vehicle, and selecting the
angle of the vehicle and then determine the stability boundary most effective braking wheel at one moment by means of
by the phase plane method. Secondly, we can check the weight differential braking.
coefficient through the current side slip angle. Finally, the From Figure 6, when the braking force is imposed to
additional yaw moment is obtained according to the current different wheels of the vehicle, a different yaw moment effect
weight coefficient. is generated at the whole vehicle. The wheel on the same side
The yaw moment decision algorithm based on yaw rate is, will produce the same yaw moment within a certain range.
t And for the front-inside wheel and the rear-outside wheel,
when the braking force is too large, the yaw moment will

ò
DM w = K pw × ew ( t ) + K iw ew ( t ) dt (7)
be opposite to that of the same side, that could induce difficul-
0
ties on yaw moment control. At the same time, we can also
In order to prevent the brake system from responding see that applying the braking force to the rear-inside wheel is
frequently to the decision from upper controller, the error most effective in generating inward yaw moment, and braking
range of the yaw rate is controlled to, force to the front-outside wheel is most effective in generating
outward yaw moment. Therefore, the intention of the decision
ìDw - Dwth + Dw > Dwth + layer is to determine whether the vehicle is in understeer or
ï - oversteer through the current steering wheel angle of the
ew = í Dw - Dwth Dw < Dwth - (8)
ï - + vehicle and the desired yaw moment, and then select the most
î 0 Dwth < Dw < Dwth effective wheel to apply the braking force.
 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

RESEARCH ON YAW STABILITY CONTROL OF UNMANNED VEHICLE BASED ON INTEGRATED ELECTROMECHANICAL 5

 FIGURE 6   Comparison of yaw moment generated by  FIGURE 7   Master cylinder pressure closed loop
braking each wheel control structure

© SAE International.
a quantitative relationship between the volume and the
displacement of master cylinder piston. So we can know the
EMBB system proposed in this paper also has a specific rela-
tionship between position of the output rod (PMSM rotor
position) and the master cylinder push rod, that is, there is a
© SAE International.

specific relationship between the PMSM rotor position and

the master cylinder pressure. Therefore, the master cylinder
pressure control strategy applied in this paper is a closed-loop
control strategy from the desired master cylinder pressure to
the desired rotor position to the desired motor current, it is
shown in Figure 7.
As shown in Figure 7, the input of the master cylinder
TABLE 2  Wheel-selection pressure is the desired master cylinder pressure value comes
from upper controller, and the master cylinder pressure
controller employs PID algorithm. Taking into account the
hydraulic pipe with practical car-size used in the HIL bench,
and using the feed-forward PV plus the PID algorithm, we can
obtain a better control effect on this system which has a large
© SAE International.

hydraulic inertia.
In order to verify the control algorithm, typical pressure
test experiment such as step, slope and sine are designed. The
experiment results are shown in Figure 8
Figure 8(A) (B) (C), show the displacement of servo body
under different test. In Figure 8(A), the raising time is
Thus, based on the current vehicle steering wheel angle
and the desired yaw rate value, we can determine the wheel
to brake. As shown in the Table 2 below, δf is the steering  FIGURE 8   The experiment results of position tracking
wheel angle, the counterclockwise direction is the
positive direct1718ion.

Execution Layer
The integrated EMBB system proposed in this paper realizes
its pressure-building by PMSM plus two-stage reduction
mechanism pushing the master cylinder push rod. The wheel
pressure regulation is realized by hydraulic control unit. The
basis for realizing vehicle yaw stability control is that the
hydraulic pressure of each wheel cylinder can follow the target
value issued by the upper controller. Therefore, it is the first
step to discuss the pressure control strategy of the master
cylinder and wheel cylinder.
Before conducting the master cylinder pressure control,
we  firstly explain the PV characteristics of the hydraulic
system. The PV characteristics of the system refer to the rela-
tionship between the pressure P generated by the brake fluid
entering the master cylinder and the volume V of the brake
© SAE International.

fluid of the system. It reflects the static characteristics of the

hydraulic system. For a particular brake system, the master
cylinder pressure has a specific relationship with the volume
of brake fluid entering the master cylinder, and there is also
 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

6 RESEARCH ON YAW STABILITY CONTROL OF UNMANNED VEHICLE BASED ON INTEGRATED ELECTROMECHANICAL

about 200ms, and in figure 8(B), (C), under slope and sine opening degree of the valve port, α is the direction angle, L is
test,  the maximum tracking error is around 0.2mm. In the damping length and ρ is the brake fluid density.
Figure 8(a) (b) (c), current of motor Q axle can closely track Meanwhile, when the valve is in open state, the relation-
the target value. Although the actual current signals have ship between the valve core opening degree and the flow rate
some jitters, the servo body position tracking performs well. flowing through the valve port can be expressed as,
After the master cylinder pressure is built up, the function
of the pressure control unit is to control the actual pressure Q = C d A ( x ) 2 Dp / r × sign ( Dp ) = Q ( x , Dp ) (15)
value of each wheel cylinder to follow the target value sent by Where, A(x) is the cross-sectional area of the valve an it
the upper controller. The hydraulic control unit of the inte- can be seen as a function of the displacement of the push rod.
grated EMBB proposed in this paper uses high-speed on-off Therefore, it can be considered that the flow rate of the valve
valve. For normally open high-speed on-off valves (such as port liquid is also related to the pressure difference between
the inlet valve), there are three working states: fully open, two ends and the displacement of the push rod.
partially open and fully closed, as shown in the Figure 9. For the control of high-speed on-off valves, pulse width
When the valve is fully opened, under the preload force modulation (PWM) is widely used at present. By selecting the
of the return spring the valve core push rod is in the initial appropriate PWM value, accurate pressure control can
position, which is the state of maximum flow. When the valve be achieved for this type of valve. After the practical calibra-
coil is conducted with a certain value of current, the moving tion and considering the actual vehicle communication cycle,
iron pushes the valve core downward under the action of the the PWM frequency is set to 50Hz.
electromagnetic force and the flow rate is reduced at this The wheel cylinder pressure control mode includes single
moment. Eventually, when the valve core moves down to the wheel boost mode, two-wheel differential boost mode, and
extreme position, the fluid channel is completely closed and four-wheel pressure regulation mode. The pressure source
the normally open valve is fully closed. during boosting process is provided by EMBB, and the
When the high-speed on-off valve works normally, its pressure reduction process and the pressure regulation process
force relationship is, are all performed by the hydraulic control unit. The control
logic diagram is shown below in Figure 10.
mx = Fm + Fh - Fs - Fv - Ff (12)
As shown in Figure 10, in the “valve initial state”, FL/RR/
Where, m is the mass of the valve core push rod, x is the FR/RL respectively stands for the PWM value of inlet valve
displacement of the valve core push rod, Fm is the electromag- in front-left/rear-right/front-right/rear-left wheel cylinder.
netic force, Fh is the hydraulic pressure, FS is the spring force, WCP stands for the wheel cylinder pressure, while MCP
Fv is the viscous resistance and Ff is the friction. Usually, stands for the master cylinder pressure. ∆P stands for the
compared to other forces, Fv  and Ff can be  ignored in difference between ideal pressure and actual pressure.
the calculation. “WCP1_IV=1” means the PWM value of inlet valve in wheel
The electromagnetic force and hydraulic pressure in cylinder1 is 1, and “WCP2_IV=PWM” means that the PWM
equation (12) can be expressed as, value of inlet valve in wheel cylinder1 is decided by
the controller.
Fm = ( NI ) / 2d 2 m0S0 = m0S0 N 2 I 2 / 2 ( l + x ) = Fm ( I ,x ) (13)
2 2
In single-wheel boost mode, only one wheel cylinder will
have a boost request at the same time. At this moment, for the
2Dp dx wheel cylinder with the demand for pressurization, the inlet
Fh = 2p C dC v R Dp sin(2q )cosa + r LC d p sin(2q ) (14)
r dt valve is opened, and the remaining inlet valves must be closed.
Where, N is the number of coil turns, I is the coil current,
δ is the air gap length, μ0 is the air gap permeability, S0 is the
air gap cross-sectional area, Cd is the liquid flow coefficient, l  FIGURE 10   The flowchart of wheel pressure
is the minimum working air gap, Cv is the liquid flow coeffi- regulation modes
cient, R is the radius of the push rod, Δp is the pressure differ-
ence between the two ends of the valve port, 2θ is the plane

 FIGURE 9   Three working stage of normally open valve

© SAE International.

© SAE International.
 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

RESEARCH ON YAW STABILITY CONTROL OF UNMANNED VEHICLE BASED ON INTEGRATED ELECTROMECHANICAL 7

The target pressure value of the master cylinder is obtained  FIGURE 11   The full view of the HIL test bench
by subtracting current wheel cylinder pressure value from the
target pressure value.
In the two-wheel differential boost mode, the wheel
cylinder with a larger target pressure demand is regulated by
the master cylinder and its inlet valve is fully opened. The
other valves utilize valves and DC motor to regulate the
pressure. The PID algorithm is applied based on the target
wheel cylinder pressure and the actual wheel cylinder pressure
at the same time taking into account the pressure difference
of the valve port.
In the four-wheel pressure regulation mode, the pressure
of the wheel cylinder with the maximum pressure demand is
controlled by the master cylinder, that is, the target master
cylinder pressure at this moment is the difference value
between the maximum target wheel cylinder pressure and the
actual pressure. The pressures of remaining wheel cylinders
are controlled by the pressure control unit.

HIL Test and Result

Analysis
© SAE International.

HIL Test Bench

In order to verify the effectiveness of the vehicle yaw stability
control algorithm based on the integrated EMBB system, the
Hardware-in- Loop (HIL) verification is carried out. Among
them, the rapid prototype controller and the actual braking in which the whole vehicle dynamic model is loaded through
system are tested as real hardware, which ensures the validity CarSim. The data transmission while HIL test is processing
and real-time of the control algorithm. The schematic diagram is required to be synchronous. Two phases are required for
and physical diagram of the entire hardware-in-the-loop test hardware-in-the-loop testing Synchronous, in this paper,
bench are shown in Figure A1 (in the Appendix A) and CAN communication is utilized to design a bus network
Figure 11. between the two real-time systems, in order to pass down the
As shown in Figure A1, the entire experimental platform pressure requirements of the upper controller and to transmit
includes hardware and software equipment, and the hardware upward the pressure following-up results of the lower
includes the dSPACE real-time system developmental tools controller. The physical diagram of the HIL bench is shown
MicroAutoBoxII, Mid-Size Simulator, RapidPro and practical in Figure 11.
hydraulic brake system. The software includes algorithm
developmental tool Matlab/Simulink, vehicle dynamics simu-
lation software CarSim and dSPACE experimental manage-
Verification of RCP Test
ment software ControlDeskNG. A Rapid Control Prototype experiment was conducted on the
In terms of analog signal acquisition, it includes 4-wheel HIL bench. The standard experimental scene of a double lane
cylinder pressure signals, 2 master cylinder pressure signals, change maneuver was set in CarSim, with an initial speed of
8 valve coil current signals, 1 pedal travel signal, etc.; In the 60 km/h and a road adhesion coefficient of 0.8. The steering
digital acquisition aspect, one motor rotor position Encoder system was set to a closed-loop driver path following model.
signal is included. The control signal includes 6 permanent The parameters related to the whole vehicle dynamics param-
magnet synchronous motor PWM output signals, 8 valve eters are shown in Table 3.
digital control signals, 1 DC motor digital control signal. From Figure 12, (A) is the steering wheel angle input value
Therefore, the selected MicroAutoBoxII is DS1401/1512/1513, under DLC maneuver. It can be seen that the steering wheel
with the 1512 board having a three-phase motor control block. angle changes greatly at this condition. And it can be seen
In terms of analog input and output, the 1513 board has 16 from (B) that the desired yaw rate changes drastically. If not
ADC channels, 16 AIO channels and 16 PWM channels, controlled, the vehicle yaw stability will be poor and it is easy
which basically meet the control input and output require- to get into dangerous situation. Therefore, after designing the
ments of the test bench. yaw stability controller, as shown in (B), under the desired
Two real-time systems are employed in this scheme. One yaw rate determined by decision layer, we can see that the
is MicroAutoBoxII, which is running yaw stability control actual yaw rate follows the target values well, and also we have
algorithm in real time and the other one is Mid-Size Simulator, to see whether the vehicle is in stable condition. Thus, in (C)
 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

8 RESEARCH ON YAW STABILITY CONTROL OF UNMANNED VEHICLE BASED ON INTEGRATED ELECTROMECHANICAL

TABLE 3  Vehicle parameters

Conclusions
M 1536.7kg
L 2910mm This paper proposes an integrated EMBB scheme that can
H 540mm be applied to intelligent unmanned vehicles. It employs PMSM

© SAE International.
a 1015mm to accomplish the master cylinder pressure-building and
b 1895mm respectively control the four-wheel cylinder pressure through
B 1675mm the hydraulic control unit. In order to improve the yaw
stability of the vehicle under extreme conditions, this paper
Kr/Kf 40000N/rad
designs a hierarchical yaw stability controller, including
decision layer, distribution layer and execution layer. The
 FIGURE 12   Results of DLC test decision layer uses the typical 2DOF vehicle reference model
to calculate the ideal yaw rate and side slip angle, while the
vehicle stability region is determined by the phase plane
method. And the desired additional yaw moment is obtained
through weighted method. In the distribution layer, one wheel
cylinder is selected at the same time by comparing the effi-
ciency for generating yaw moment to make the vehicle return
to the steady state as quickly as possible. The execution layer
acts as the bottom level controller that tracks the upper target
pressure. By designing the master cylinder and wheel cylinder
pressure control strategy, each wheel cylinder can accurately
follow the target pressure value in real-time. Finally, the
control algorithm was verified by building HIL test bench
containing the rapid prototype controller MicroAutoBoxII,
real-time simulation tool Mid-Size Simulator and practical
car-size braking system. The experiment results show that the
proposed hierarchical control strategy can improve dynamic
stability under extreme conditions and improve the safety of
the vehicle.

References
1. Yu, Z., Xu, S., Xiong, L., and Han, W., “An Integrated-
Electro-Hydraulic Brake System for Active Safety,” SAE
Technical Paper 2016-01-1640, 2016, https://doi.
org/10.4271/2016-01-1640.
2. Riley, B., Kuo, G., Nethercutt, T., Shipp, K. et al.,
“Development of a Haptic Braking System as an ACC Vehicle
© SAE International.

FCW Measure,” SAE Technical Paper 2002-01-1601, 2002,

https://doi.org/10.4271/2002-01-1601.
3. Ko, J., Ko, S., Son, H., Yoo, B. et al., “Development of Brake
System and Regenerative Braking Cooperative Control
Algorithm for Automatic-Transmission-Based Hybrid
we can see the side slip angle follows the target closely, and Electric Vehicles,” IEEE Transactions on Vehicular
the vehicle is always in the stable region. And in (D), the PWM Technology 64(2):431-440, 2015, doi:10.1109/
value of each valve shows the working state of the pressure TVT.2014.2325056.
control unit. In terms of the execution layer, in Figure 12 (E), 4. He, X., Yang, K., Ji, X., Liu, Y. et al., “Research on Vehicle
each wheel cylinder pressure can track the target wheel, Stability Control Strategy Based on Integrated-Electro-
though there are some the pressure tracking error, but the Hydraulic Brake System,” SAE Technical Paper 2017-01-
vehicle yaw stability has good control effect. Compared with 1565, 2017, https://doi.org/10.4271/2017-01-1565.
the traditional ESP, the integrated EMBB system can obtain 5. Soga, M., Shimada, M., Sakamoto, J., and Otomo, A.,
faster response and higher control precision, it has stronger “Development of Vehicle Dynamics Management System for
ability on pressure keeping and pressure dithering. Therefore, Hybrid Vehicles - ECB System for Improved Environmental
it proves that the integrated EMBB proposed in this paper is and Vehicle Dynamic Performance,” SAE Technical Paper
effective in yaw stability control. 2002-01-1586, 2002, https://doi.org/10.4271/2002-01-1586.
 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

RESEARCH ON YAW STABILITY CONTROL OF UNMANNED VEHICLE BASED ON INTEGRATED ELECTROMECHANICAL 9

6. Liu, G., Fang, E., and Zhu, Z., “Development of electronic 15. Chen, P., Wu, J., Zhao, J., He, R. et al., “Design and Power-
boost brake system,” Shanghai Auto (01):11-15, 2016, Assisted Braking Control of a Novel Electromechanical
doi:10.3969/j.issn.1007-4554. 2016.01.03. Brake Booster,” SAE Int. J. Passeng. Cars - Electron. Electr.
7. Xiong, L., Qian, C., and Yu, Z., “Review on Composite Syst 11(3):171-181, 2018, https://doi.org/10.4271/2018-01-
Braking System of Electric Vehicle,” Automobile Technology 0762.
01:1-8, 2015. 16. Zhang, H., Wu, J., He, R., and Chen, Z., “Development and
8. Büchler, R., “Future Brake Systems and Technologies-MK Verification of Control Algorithm for Permanent Magnet
C1®-Continental’s Brake System for Future Vehicle Synchronous Motor of the Electro-Mechanical Brake
Concepts,” in 7th International Munich Chassis Symposium Booster,” SAE Technical Paper 2019-01-1105, 2019, https://
2016, Springer Vieweg, Wiesbaden, 2017, 717-723, doi.org/10.4271/2019-01-1105.
10.1007/978-3-658-14219-3_45. 17. Wang, J., Wu, J., He, R., and Chen, Z., “Pressure
9. Yu, Z., Shi, B., and Xiong, L., “Coordinated Control of Optimization Control of Electro-Mechanical Brake System
Hybrid Braking Based on Integrated-Electro-Hydraulic in the Process of ABS Working,” SAE Technical Paper 2019-
Brake System,” Journal of Tongji University (Natural Science) 01-1104, 2019, https://doi.org/10.4271/2019-01-1104.
47(06):851-856, 2019, doi:10.11908/j.issn. 0253-374 x. 18. Chen, P., Wu, J., Zhao, J., He, R. et al., “Design and Power-
2019.06.015. Assisted Braking Control of a Novel Electromechanical
10. Li, H., Yu, Z., Xiong, L., and Han, W., “Hydraulic Control of Brake Booster,” SAE Int. J. Passeng. Cars - Electron. Electr.
Integrated Electronic Hydraulic Brake System Based on Syst 11(3):171-181, 2018, https://doi.org/10.4271/2018-01-
LuGre Friction Model,” SAE Technical Paper 2017-01-2513, 0762.
2017, https://doi.org/10.4271/2017-01-2513. 19. Guo, J., “Research on Electronic Stability Coordinated
11. Ohkubo, N., Matsushita, S., Ueno, M., Akamine, K. et al., Control System for Double-Axle Vehicles,” Jilin
“Application of Electric Servo Brake System to Plug-In University, 2008.
Hybrid Vehicle,” SAE Int. J. Passeng. Cars - Electron. Electr.
Syst 6(1):255-260, 2013, https://doi.org/10.4271/2013-01-0697.
12. Todeschini, F., Corno, M., Panzani, G. et al., “Adaptive
Contact Information
Position-Pressure Control of a Brake by Wire Actuator for For questions or to contact the authors, please email to the
Sport Motorcycles,” European Journal of Control 20(2):79-86, corresponding author Rui He at: herui@jlu.edu.cn
2014, doi:10.1016/j.ejcon.2013.12.003.
13. Shibahata, Y., Shimada, K., and Tomari, T., “Improvement of
Vehicle Maneuverability by Direct Yawmoment Control,” Acknowledgments
Vehicle System Dynamics 22((5-6)):465-448, 1993,
The authors wish to acknowledge the support of the National
doi:10.1080/00423119308969044.
Key R & D Program of China (2018YFB0105103).
14. Hiemer, M., “Model Based Detection and Reconstruction
Ofroad Traffic Accidents,” Kalsruhe University,
Kalsruhe, 2005.
 Downloaded from SAE International by Dexing Lao, Tuesday, May 19, 2020

10 RESEARCH ON YAW STABILITY CONTROL OF UNMANNED VEHICLE BASED ON INTEGRATED ELECTROMECHANICAL

Appendix A

 FIGURE A1   The schematic diagram of the HIL test bench

© SAE International.

© 2020 SAE International. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE International.

Positions and opinions advanced in this work are those of the author(s) and not necessarily those of SAE International. Responsibility for the content of the work lies
solely with the author(s).

ISSN 0148-7191

View publication stats