Int. J. Vehicle Design, Vol. 58, Nos.

2/3/4, 2012 223

Integrated control of electromechanical braking and

regenerative braking in plug-in hybrid electric
vehicles

Zhiguang Zhou
College of Mechanical and Vehicle Engineering,
Hunan University,
Yuelu District, Changsha City,
Hunan Province, 410082, China
and
Department of Electrical and Computer Engineering,
University of Michigan-Dearborn,
4901 Evergreen Road,
Dearborn, MI 48128, USA
E-mail: [email protected]

Chris Mi*
Department of Electrical and Computer Engineering,
University of Michigan-Dearborn,
4901 Evergreen Road, Dearborn, MI 48128, USA
E-mail: [email protected]
*Corresponding author

Guixiang Zhang
College of Mechanical and Vehicle Engineering,
Hunan University,
Yuelu District, Changsha City,
Hunan Province, 410082, China
E-mail: [email protected]

Abstract: This paper proposes the use of Electromechanical Brakes (EMB) in

combination with regenerative braking in PHEV so braking force can be
distributed to front and rear axles according to an optimal curve instead of a
linear line. Therefore, more braking force will be distributed to front axle,
which will offer more kinetic energy for regenerative braking. Fuzzy logic
control is used to allocate braking force between regenerative braking and
electromechanical braking. The paper also studies antilock braking control of
EMB to maintain safety and stability of a vehicle using sliding-mode control
with switching gain fuzzy adjusting.

Keywords: plug-in; HEV; hybrid electric vehicle; EMB; electromechanical

brakes regenerative braking; frictional braking; sliding-mode control; fuzzy
logic.

Copyright © 2012 Inderscience Enterprises Ltd.

224 Z. Zhou et al.

Reference to this paper should be made as follows: Zhou, Z., Mi, C. and
Zhang, G. (2012) ‘Integrated control of Electromechanical Braking and
regenerative braking in plug-in hybrid electric vehicles’, Int. J. Vehicle Design,
Vol. 58, Nos. 2/3/4, pp.223–239.

Biographical notes: Zhiguang Zhou received the BS from Xiangtan

University, China, in 2004 and the MS from Hunan University, China, in 2007.
He is now a Doctoral Student of Hunan University, China. At present, he is
doing some research at the University of Michigan-Dearborn as a Visiting
Scholar. His research interests focus on the design, modelling and control of
hybrid electric vehicle.

Chris Mi is an Associate Professor of Electrical Engineering at the University

of Michigan, Dearborn. He received his BSEE and MSEE from Northwestern
Polytechnical University, China, and his PhD from the University of Toronto,
Canada. His research interests are in power electronics, hybrid electric vehicles,
electric machines and renewable energy systems. He worked for General
Electric Canada Inc. from 2000 to 2001. He is the recipient of numerous
awards including the Distinguished Teaching Award and Distinguished
Research Award from the University of Michigan, National Innovation Award
and Government Special Allowance Award from China, and IEEE and SAE
Awards. He is a Senior Member of IEEE and has more than 100 publications.

Guixiang Zhang received the BS from Hunan University, China, in 1979.

She is now a Professor at the College of Mechanical and Vehicle Engineering,
Hunan University. Her teaching and research interests are in the areas of
mechanical transmission control and computer control technology.
She has published more than 30 technical papers in mechanical transmission
control and computer control areas.

1 Introduction

Hybrid and Plug–In Hybrid Electric Vehicles (PHEVs) offer significant fuel savings
compared with conventional gasoline- and diesel-powered vehicles (Simpson, 2006).
In HEV and PHEV, the braking force is coordinated between regenerative braking and
frictional braking (Gao and Ehsani, 2001). Regenerative braking force is produced by the
electric motor and the mechanical braking force is supplied by hydraulic brakes or air
pressure brakes (Jonas, 2007). Regenerative braking is activated during coasting and
decelerations to recapture a portion of the kinetic energy to charge the on-board energy
storage system. This recaptured energy is then used to provide the electrical power
required by the powertrain, as well as the vehicle accessory power loads during vehicle
idle conditions (Gao et al., 1999). Because of the slow response of the traditional
frictional brakes, coordination of the two braking systems has been proven to be
extremely difficult, and the capability of recovering the maximum amount of energy
during braking has been a challenge. On the contrary, EMB can control the braking
torque precisely and quickly. To optimise the regenerative braking control so as to
maximise the regenerative braking energy, we propose to use EMB to achieve the
distribution of braking force to front and rear axes according to an optimal curve, also
known as ideal curve or I-curve. We then use fuzzy logic control strategy to distribute the
braking force between regenerative braking and frictional braking.
 Integrated control of electromechanical braking 225

EMB can also realise anti-lock function during braking to maintain the safety and
stability of the vehicle when it is on wet, snowy, or icy roads. The main difficulty arising
in the design of braking control system is the high non-linearity of the system and the
presence of disturbances and parameter uncertainties. A robust control methodology
needs to be adopted to solve the problem. At present time, the commonly used method is
logic threshold method based on experience, and plenty of experiments need to be carried
out to obtain the threshold values (Chen, 1995). Other control strategies, such as PID
control (Chen, 1995), fuzzy control (Chen and Liu et al., 2002; Li et al., 2001), neural
network (Huang et al., 2003) and sliding-mode control (Wang et al., 2004; Ahn et al.,
2009), have been studied for ABS control, but they are difficult to implement if real time
in the real system.
Sliding-mode methodology has been adopted to design the observer since it is
applicable to non-linear systems and has good robustness properties against disturbances,
modelling inaccuracy and parameter uncertainties. However, the conventional sliding-
mode control generates a discontinuous control action that has the drawback of producing
high-frequency chattering, with the consequent excessive mechanical wear and
passenger’s discomfort. To reduce the vibrations induced by the controller, fuzzy logic
control is used to adjust the switching gain of sliding-mode control. Simulation results
indicate that the chattering is greatly decreased.

2 Braking principle of PHEV with EMB

The architecture of the studied vehicle is a blended mode, front wheel drive, power-split
PHEV as shown in Figure 1. The engine and motor can supply power at the same time or
alone under different driving conditions, according to power management algorithms.
To convert kinetic energy to electrical energy, the system uses the motor as a generator.
The generator produces a torque to resist the rotation of the wheels, providing brake force
needed to slow the vehicle. The generator can also produce electricity with the excessive
power from the engine, so as to adjust the operation area of the engine to keep the engine
operating in the high-efficiency region. Four EMBs are installed in the vehicle, which can
supply braking torque to all four wheels independently. In addition, the EMB also
provides the anti-lock braking function. The EMBs communicate with the brake
controller by FlexRay or CAN bus.
In a regular HEV, braking torque is a combination of regenerative braking, frictional
braking and frictional torque of the engine. Since the engine frictional torque will
decrease the regenerative amount, to obtain more regenerative energy, in this paper,
engine will be disconnected at the time of braking. The main parameters of the PHEV are
shown in Table 1.

Table 1 Parameters of PHEV

Component Rating
Engine 5.7 L 257 kW
Motor 65 kW
Generator 65 kW
Battery pack Lithium-ion battery, 10 kWh, 300 V
Vehicle gross weight 3298 kg
226 Z. Zhou et al.

Figure 1 Architecture of PHEV (see online version for colours)

In this paper, a non-linear model of the vehicle is adopted (Yang et al., 2009).
The vehicle is modelled as a rigid body, and only longitudinal motion is considered.
The difference between the left and right tyres is neglected, making reference to a
so-called bicycle model. The lateral, yawing, pitch, and roll dynamics, as well as actuator
dynamics, are also neglected.

3 EMB System

The EMB system replaces the conventional hydraulic system with an electromechanical
system, a type of Brake By Wire (BBW) system, which employs an EMB module as the
braking module for each wheel. The EMB system is able to provide a large braking
torque, using only a small brake pedal reaction force and a short pedal stroke
(Ahn et al., 2009).
The EMB system consists of the EMB controller, motor, deceleration and motion
conversion device (Line et al., 2008). The EMB controller is used to control the output
torque of the motor. The requirement of EMB system to the motor is that the machine can
produce large torque at low speed, both Switched Reluctance Motor (SRM) and
permanent magnetic motor are candidates for EMB system. EMB system would produce
large heat during braking, which will cause the temperature to rise and the
demagnetisation phenomenon will occur when the permanent magnetic motor stays at
high temperature for a certain time. So, we select the SRM as our implementation device.
Another reason is the robustness and simplicity of SRM. The deceleration device has the
function of deceleration and increasing mechanical torque. The motion conversion device
is used to convert rotary motion into linear motion, so as to compress braking disc. The
architecture of the proposed EMB is shown in Figure 2.

Figure 2 The architecture of the proposed EMB (see online version for colours)
 Integrated control of electromechanical braking 227

3.1 A mathematical model of EMB

Among the three most common SRMs, three-phase 6/4-pole SRM, four-phase 8/6-pole
SRM and five-phase 10/8-pole SRM (Krishnamurthy et al., 2009), the three-phase
6/4-pole SRM produces large torque ripple and a five-phase 10/8 SRM is more complex
and has high cost. Therefore, the four-phase 8/6-pole SRM is used in this paper as shown
in Figure 3.

Figure 3 Structure of four 8/6 SRM (see online version for colours)

The rating of the four-phase 8/6-pole SRM used in this paper is 200W, 42V. Each SRM
can supply as high as 13kN braking force.
Set the kth (k = 1, 2, 6) phase voltage, current, resistance, inductance, flux linkage and
torque as uk, ik, Rk, Lk, ψk and Tk, respectively, rotor position angle is θ, rotor speed is ω,
winding co-energy is Wm, moment of inertia is J, viscous friction coefficient is B and load
torque is TL. To simplify the calculation, inter-phase mutual inductance is neglected and
interaction between phases produced by stator and rotor yoke saturation when more than
two phases conducted is also not considered. The kth phase flux linkage equation of SRM
can be described as
ψ k = ψ k ( ik , θ ) = Lk (ik , θ ) ik . (1)

The kth phase winding voltage equation of SRM is represented by

dψ k
uk = Rk ik + . (2)
dt
Substituting equation (1) in equation (2), we can obtain
 ∂L  di ∂L dθ
uk = Rk ik +  Lk + ik k  k + ik k . (3)
 ∂ik  dt ∂θ dt

According to electromechanical energy conversion principle, the kth phase

electromagnetic torque of SRM can be written as
∂Wm (ik , θ )
Tk = = Tk (ik , θ ). (4)
∂θ
SRM output electromagnetic torque stacked by each phase electromagnetic torque can be
described as
228 Z. Zhou et al.

4
Te = ∑ Tk ( ik , θ ) . (5)
k =1

Mechanical motion equation of SRM is given by

dω
J = Te − Bω − TL (6)
dt
dθ
= ω. (7)
dt
The mathematical model of the actuator is (Yang et al., 2009)

θ s = θ / id (8)

S n = θ m Ls / (2π ) (9)

Tbs = Fs Ls /(2πηt ). (10)

In equations (8)–(10), θs is the rotating angle of screw, id is the deceleration ratio of

decelerator, Sn is the shift distance of nut, Ls is the thread lead of screw, Tbs is the thread
torque of screw, Fs is the axial working load of screw and ηt is the transmission
efficiency.
The mathematical model of the brake is
T fb = 2 Fs µ f Rb . (11)

In equation (11), Tfb is the braking torque of brake and µf is the friction coefficient
between braking friction plate and braking disc.

3.2 Torque control strategy of EMB

The block diagram of the proposed braking torque control strategy is shown in Figure 4,
where Te∗ is the required braking torque. In the paper, it is assumed that there is only two
phases conducting at the same time, so after Torque Distribution Function (TDF), Te∗ is
divided into Ti ∗ and T j∗ (i, j = A, B, C, D), which describes the sequence phase rotor
torque. The TDF is shown in Figure 5, which can be realised easily (David, 2001). With
Ti ∗ T j∗ and rotor position angle θ, through a look-up table, which contain the relationship
among phase torque, rotor position angle and phase current, sequence phase current
ii∗ and ii* and i*j (i, j = A, B, C, D) can be obtained. The relationship among torque, rotor
position angle and current can be obtained by experiment, as shown in Figure 6. Then,
hysteresis current control method is used to produce gating signal to control the power
converter with the input parameters ii* , i*j , θ and rotor current im. The power converter
will generate input voltage to SRM according to the gating signal. So rotor torque will be
produced and transferred to the gear. Simulation result shown in Figure 7 illustrates that
the control method used in this paper can satisfy the torque precision of braking
systems.
 Integrated control of electromechanical braking 229

Figure 4 Block diagram of braking torque control strategy

Figure 5 Torque distribution function (see online version for colours)

Figure 6 Relationship among torque, rotor position angle and current (see online version
for colours)

Figure 7 Actual torque follows required torque (see online version for colours)
230 Z. Zhou et al.

4 Integrated control of EMB and regenerative braking

The target of braking control system is to maximise regenerative energy while

maintaining vehicle stability and safety. To achieve this goal, this paper uses an optimal
or ideal curve (I-curve) to distribute the braking torque between front and rear axle, and
then uses fuzzy logic to distribute braking torque between regenerative braking and
friction braking in the front axle. Fuzzy-sliding mode control is used to avoid wheels
being locked. The structure of braking control system is shown in Figure 8. In Figure 8
Txb_com is the braking torque command from the driver. Txb_req is the required braking
torque after the adjusting of anti-lock braking controller. ∆Txb is the adjusting quantity.
Txb_f and Txb_r are braking torque distributed to front and rear axles, respectively. Txb_freg
and Txb_ffric are braking torque distributed to regenerative braking and frictional braking,
respectively. Txb_fl, Txb_fr, Txb_rt and Txb_rr are braking torque distributed to front-left,
front-right, rear-left and rear-right wheels, respectively. ωfl, ωfr, ωrl and ωrr are angular
speed of front-left, front-right, rear-left and rear-right wheels, respectively. vx is the
vehicle longitudinal speed.

Figure 8 Structure of braking control system

4.1 Torque distribution between front and rear axle

In conventional vehicles, the braking torque is difficult to be distributed to front and rear
axle according to the I-curve because it is difficult to precisely control the hydraulic
system. Traditional method distributes the total braking torque to front and rear axes
according to some proportion. With EMB system, the braking torque can be precisely
controlled with fast response time. Therefore, braking torque can be distributed between
front and rear axle according to the I-curve as shown in Figure 9. To compare the
distribution results from I-curve and from the traditional method, the torque distribution
for the UDDS driving cycle is shown in Figures 10 and 11 for conventional and the
proposed method, respectively. It can be seen that by using I-curve, more braking torque
is distributed to the front axle, so it offers more possibility to capture more regenerative
energy.

4.2 Torque distribution between regenerative braking and friction braking

In this paper, both regenerative and friction braking can supply braking torque to the
front axle. It is important to properly distribute the front axle braking torque between
regenerative and friction braking to maximise energy capture while maintaining safety of
the vehicle and health operation of components (motor, inverter and battery). To achieve
this goal, this paper uses fuzzy logic control strategy to distribute braking torque to
 Integrated control of electromechanical braking 231

regenerative braking as much as possible under some constraints. The inputs of fuzzy
logic controller are battery State Of Charge (SOC) and torque change ratio δ, which is
defined as

δ (t ) = Txb req
(t ) (Txbreq ( t − 1) + ∆Txb (t ) ) . (12)

The output of fuzzy controller is regenerative braking factor β, so the regenerative

braking torque Txb_freg and friction braking torque in front axle can be obtained as
Txb _ freg = β ⋅ Txb _ f (13)

Txb _ ffric = (1 − β ) ⋅ Txb _ f . (14)

The membership function of input variables and output variable are shown in
Figures 12–14, respectively. In Figure 12, TL means too low, L means low, M means
middle, H means high and TH means too high. In both Figures 13 and 14, TS means too
small, S means small, M means middle, L means large and TL means too large.
According to engineering expertise and insight, the fuzzy control rules can be constructed
as shown in Figure 15.

Figure 9 I-curve (see online version for colours)

Figure 10 Braking torque distributed between front and rear axle using traditional method
(see online version for colours)

Figure 11 Braking torque distributed between front and rear axle using I-curve (see online version
for colours)
232 Z. Zhou et al.

Under the precondition of distributing braking torque to front and rear axle, the results of
distributing braking torque between regenerative braking and friction braking using
look-up table method and using fuzzy logic control algorithm are shown in
Figures 16 and 17, respectively. The parameters of the look-up table are obtained from
vehicle simulation software Advisor 2002. It can be seen that more braking torque is
distributed to regenerative braking with fuzzy logic control, which means more
regenerative energy can be recovered. The regenerative energy obtained using the two
methods are shown in Table 2.

Table 2 Regenerative energy using lookup table and using fuzzy logic control

Methods Regenerative energy

Look-up table 971 kJ
Fuzzy logic control 1136 kJ

Figure 12 Membership function of SOC. TL – too low; L – low; M – Medium; H – high;

TH – too high (see online version for colours)

Figure 13 Membership function of δ . TL – too low; L – low; M – Medium; H – high; TH – too

high (see online version for colours)

Figure 14 Membership function of β . TL – too low; L – low; M – Medium; H – high; TH – too

high (see online version for colours)
 Integrated control of electromechanical braking 233

Figure 15 Rules of fuzzy logic control for braking torque distribution (see online version
for colours)

Figure 16 Braking torque distributed between front friction braking and regenerative braking
using look-up table method (see online version for colours)

Figure 17 Braking torque distributed between front friction braking and regenerative braking
using fuzzy logic control (see online version for colours)

4.3 Anti-lock braking control

EMB system has the function of anti-lock braking. Vehicle braking system is a non-linear
and time-variant system, and the vehicle model is uncertain. So, it is hard to build the
precise mathematical model for the vehicle braking process. Sliding-mode control is
insensitive to the changing of parameters of the control object. It has good
anti-interference ability and good dynamic performance, and can response quickly.
Therefore, it is suitable for both linear and non-linear systems. But, the algorithm of
sliding-mode control has static error adjustment so it is hard to ensure the accuracy in
static. It also has high instantaneous switching frequency on sliding-mode surface, which
will affect the control precision of braking torque. To solve this problem, this paper uses
fuzzy rules to do effective assessment for the switching gain according to sliding-mode
234 Z. Zhou et al.

reaching condition to eliminate the interference, thus to eliminate chattering. The control
diagram of vehicle braking system is shown in Figure 18. In the figure, sopt is the optimal
slip ratio; s is the real slip ratio calculated from the sensor signal of vehicle speed and
wheel speed. By using sliding-mode control with switching gain fuzzy adjusting, the real
slip ratio will track the optimal slip ratio, so as to ensure the vehicle braking system
follows the optimal slip ratio.

Figure 18 Structure of fuzzy-sliding mode control

4.4 Design of sliding-mode controller

In vehicle EMB system, four wheels are controlled by using the same control method
separately. To simplify the analysis, aerodynamic drag force and rolling resistance are
neglected; the motion equations are described as follows:
Vehicle motion equation is
dv
m = − Fx ( s ). (15)
dt
Wheel motion equation is
dw
Jω = Fx ( s ) R − Tb (t ). (16)
dt
Wheel longitudinal friction force is
Fx ( s ) = µ ( s )mg . (17)

In equations (15) and (17), m is the one-fourth mass of total vehicle mass; v is the vehicle
speed; Fx(s) is the wheel longitudinal friction force; Jω is the wheel inertia; ω is the wheel
angle speed; R is the wheel radius; Tb(t) is the braking torque; µ(s) is the adhesion
coefficient between wheel and road surface and g is the gravitational acceleration.
In the condition of braking, slip ratio can be defined as
v − vω v − ω R
s= = . (18)
v v
In equation (18), vω is the wheel linear speed.
Taking derivative of both sides of equation (18), we can obtain that
• •
• (1 − s ) v − ω R
s= . (19)
v
Substituting equations (15)–(17) to equation (19), we can obtain that
 Integrated control of electromechanical braking 235

• µ ( s )mg 1 − s R 2  Tb (t ) R
s=−  + + . (20)
v  m Jω  Jω v

Select the braking torque as the control object, and the control target is to find out the
control rules, which can maintain the real slip ratio s oscillating around optimal slip ratio
sopt. Define the sliding-mode surface as
δ (t ) = se (t ) = sopt − s(t ). (21)

When the braking system moves along the sliding-mode surface, δ (t ) = 0 , so

differentiate equation (21) and combine with equation (20), the equal control braking
torque can be described as
∧ Jω N µ
T beq = (1 − s ) + N µ R. (22)
mR
In equation (22), N = mg, which is the gravity of vehicle.
When braking system state is outside the sliding-mode surface, monitor control
braking torque Tb is added to ensure that the system state can reach at sliding-mode
surface (Liu, 2005),
~
Tb = F (t ) ⋅ sgn(δ ). (23)

In equation (23), sgn(⋅) is a sign function; F(t) is used to compensate system interference
to make sure the sliding mode exist condition can be satisfied. So, the output of control
system is
∧ ~ Jω N µ
Tb = T beq + Tb = (1 − s ) + N µ R + F (t ) sgn(δ ). (24)
mR
In sliding-mode control law equation (24), switching gain F(t) is the reason of causing
chattering, because system interference is time variant, and to decrease chattering, F(t) is
also time variant. To eliminate the chattering caused by system interference, this paper
uses fuzzy rules to estimate switching gain according to sliding-mode reaching condition.

4.5 Design of fuzzy logic controller

To eliminate the high-frequency chattering caused by time variable gain F(t), the fuzzy
logic rules can be designed as follows:
•
If δ δ > 0, then increase F (t ) (25)
•
If δ δ < 0, then decrease F (t ) (26)

With equations (25) and (26), fuzzy logic system can be designed about δ δ and
∆F(t). In this system, δ δ is the input and ∆F(t) is the output. Membership function
of input and output of fuzzy logic system is shown in Figures 19 and 20. In both
figures, NB means negative big, NM means negative middle, ZO means zero,
236 Z. Zhou et al.

PM means positive middle and PB means positive big. The control rules are shown in
Figure 21.
^
Using integration method to estimate the upper boundary of F (t ) ,
^ t
F (t ) = G ∫ ∆Fdt. (27)
0

In equation (27), G is the scale coefficient, G > 0

•
Figure 19 Membership function of input δ δ (see online version for colours)

Figure 20 Membership function of output ∆F (t ) (see online version for colours)

Figure 21 Fuzzy rules (see online version for colours)

5 Analysis of simulation results

To test the braking performance under different braking conditions, a special drive cycle
is constructed as shown in Figure 22. In the six phases, the acceleration is always
 Integrated control of electromechanical braking 237

1.5 m/s2, but the deceleration is varied, from −0.5 m/s2 to −12 m/s2. The required and
actual deceleration is shown in Figure 23. It can be seen that when the deceleration is
under −8 m/s2, the braking system can supply required braking torque according to road
adhesive force. But when the deceleration is greater than −8 m/s2, because of the
limit of road adhesive force, the actual deceleration cannot arrive at the required one.
During the whole braking process, the slip ratio and adhesive coefficient are shown in
Figures 24 and 25, respectively. From Figure 24, we can see that the slip ratio is always
smaller than 1, which means that the wheels are never locked during the whole cycle.
It also proves that the anti-lock braking control strategy used in this paper is feasible.

Figure 22 Constructed drive cycle (see online version for colours)

Figure 23 Required and actual deceleration (see online version for colours)

Figure 24 Slip ratio (see online version for colours)

Braking torque distributed between front and rear axle and between regenerative
braking and friction braking are shown in Figures 25 and 26, respectively. The
results show that the braking distribution method used in this paper is feasible and
reasonable.
238 Z. Zhou et al.

Figure 25 Braking torque between front and rear axle (see online version for colours)

Figure 26 Braking torque between regenerative braking and friction braking (see online version
for colours)

6 Conclusion

This paper proposed a coordinated control of regenerative braking and EMB in a PHEV.
The main objective is to maximise the regenerative braking energy while maintaining the
safety and stability of the vehicle during braking. To achieve this goal, from the aspect of
hardware, considering the braking torque can be controlled precisely and quickly, EMB
system is used as the frictional braking system. From the aspect of control method, first,
an I-curve is used to distribute the total braking torque between front and rear axles so
more braking torque is distributed to the front axis. It there offers more possibility to
capture more regenerative energy. Second, fuzzy logic control is used to distribute
braking torque between regenerative braking and friction braking on the front axis.
Comparing with look-up-table-based method, fuzzy logic control strategy will distribute
more braking torque to regenerative braking than to friction braking, which means
more regenerative energy can be captured. To keep safety and stability of the vehicle,
fuzzy-sliding mode control strategy is used to achieve the function of anti-lock braking.
Standard driving cycles are used to test the performance of the control strategy.
Simulation results indicate that the fuzzy-sliding mode control can keep the real slip ratio
 Integrated control of electromechanical braking 239

following the optimal slip ratio closely and quickly, so the braking distance can be
shortened and safety and stability of vehicle maintained.

References
Ahn, J.K., Jung, K.H. and Kim, D.H. (2009) ‘Analysis of a regenerative braking system for hybrid
electric vehicles using an electro-mechanical brake’, International Journal of Automotive
Technology, Vol. 10, No. 2, pp.229–234.
Chen, C. and Liu, J. (2002) ‘The study of slip ratio based anti-lock brake system based on fuzzy
control’, Journal of Wuhan University of Technology, Vol. 24, No. 10, pp.50–53.
Chen, J. (1995) Theory and Practice of Antilock Braking System, Institute of Technology Press,
Beijing.
David, I.J. (2001) Switched Reluctance Motor Drives-Modeling, Simulation, Analysis, Design, and
Applications, CRC Press, Boca Raton, Florida, USA.
Gao, Y. and Ehsani, M. (2001) Electronic Braking System of EV And HEV – Integration of
Regenerative Braking, Automatic Braking Force Control and ABS, SAE Technical Paper,
2001-01-2478, pp.1–7.
Gao, Y., Chen, L. and Ehsani, M. (1999) Investigation of the Effectiveness of Regenerative Braking
for EV and HEV, SAE Technical Paper, 1999-01-2910, pp.1–7.
Huang, S-J.,Huang, K-S. and Chiou, K-C. (2003) Development and Application of a Novel Radial
Basis Function Sliding Mode Controller, Mechatronics,Vol. 13, pp.313–329.
Jonas, H. (2007) ‘Maximisation of brake energy regeneration in a hybrid electric parallel car’,
Int. J. Electric and Hybrid Vehicles, Vol. 1, No. 1, pp.95–121.
Krishnamurthy, P., Lu, W.and Khorrami, F. (2009) ‘Robust force control of an SRM-based
electromechanical brake and experimental results’, IEEE Transactions on Control Systems
Technology, Vol. 17, No. 6, pp.1306–1317.
Li, J. Yu, F. and Zhang, J. (2001) ‘Fuzzy control of anti-lock braking system based on road
automatic identification’, Transaction of Agricultural Machinery, Vol. 32, No. 5,
pp.26–29.
Line, C., Manzie, C. and Malcolm, C. (2008) ‘Good. Electromechanical brake modeling and
control: from PI to MPC’, IEEE Transactions on Control Systems Technology, Vol. 16, No. 3,
pp.446–457.
Liu, J. (2005) MATLAB Simulation of Sliding Mode Control, Tsinghua University Press, Beijing.
Simpson, A. (2006) ‘Cost-benefit analysis of plug-in hybrid electric vehicle technology’,
22nd International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium and Exhibition
(EVS-22), Yokohama, Japan, pp.1–15.
Wang, J., He, H. and Chen, Y. (2004) ‘Control method of ABS based on adhesion coefficient
– slip curve’, Vehicle Technology, Vol. 6, pp.13–16.
Yang, K., Li, J. and Li, Y-d. (2009) ‘Study of EBD/ABS based on electromechanical brake
system’, Journal of System Simulation, Vol. 21, No. 6, pp.1785–1788.