International Journal of Automotive Technology, Vol. 14, No. 5, pp.

773−778 (2013) Copyright © 2013 KSAE/ 073−13

DOI 10.1007/s12239−013−0085−0 pISSN 1229−9138/ eISSN 1976−3832

THREE-SPEED TRANSMISSION SYSTEM FOR PURELY

ELECTRIC VEHICLES

Z. ZHANG*, C. ZUO, W. HAO, Y. ZUO, X. L. ZHAO and M. ZHANG

College of Mechanical Science and Engineering, Jilin University, Changchun 130022, China

(Received 18 January 2012; Revised 23 October 2012; Accepted 16 January 2013)

ABSTRACT−This paper discusses the necessity of using a transmission system to improve the energy efficiency of purely
electric vehicles (EVs). The energy efficiency of an electric motor varies at different operating points to meet the output power
demand. The three gear ratios of a transmission system can maintain the motor speed within a stable region with relatively
high energy efficiency, while various vehicle speeds are needed. This work is based on a light EV prototype. The optimized
gear ratios of this transmission result in a considerably reduced energy consumption of 9.3% compared with conventional EVs
with single-speed reducers under the condition of the Urban Dynamometer Driving Schedule driving cycle. Thus, the
transmission system is necessary to improve the energy efficiency of EVs.

KEY WORDS : Electric vehicle, Transmission system, Motor efficiency, Gear ratio, Optimization

1. INTRODUCTION adjust the constant motor speed and variable vehicle speed.
A higher motor efficiency results in lower energy
Purely electric vehicles (EVs) have become a highly popular consumption (Andersson, 2004) and higher endurance
topic in the last decade. EVs have attracted extensive (Pearre et al., 2011). Electric variable transmissions (EVTs)
attention from researchers in the fields of new energy and are widely used in hybrid EVs for power splitting (Cheng
advanced automobile technology (Axsen and Kurani, et al., 2009; Kim et al., 2011). Suh et al. (2011) have
2008; Thomas, 2009). Instead of a conventional internal carried out the optimization for a heavy-duty hybrid
combustion engine (ICE), an electric motor can be used to electric bus using a continuously variable transmission
power a vehicle for environmental friendliness and low (CVT) to achieve high efficiency of the ICE and electric
operating costs (Bishop et al., 2011). motor. EVTs and CVTs can provide continuous speed
In a conventional vehicle, a transmission is used to ratios but suffer from bulky energy losses caused by
adjust the speed ratio from the ICE to the rest of the electromagnetic induction or friction. Gearbox transmissions
driveline (Zackrisson 2003), which can increase the have the highest mechanical efficiency (Lechner and
starting torque and extend the range of operating speeds Naunheimer, 1999). Researchers have recently started
(Chen and Cheng, 2006). The electric motor provides studying gearbox transmissions for EVs (Qin et al., 2011).
constant torque at a low speed and provides rated power at Some concept prototypes are available in the market, such as
the maximum speed (Rahman et al., 2000). Its high starting those by Vocis/Oerlikon Graziano and Antonov (Loveday
torque and wide speed range are considered to be sufficient 2011; McKeegan 2011). This paper proposes an application
to meet the driving demand. Thus, multi-speed transmissions of gearbox transmission in the EV field and performs the
are not essential and single-speed reducers are extensively optimization of gear ratios to obtain the maximum energy
used in commercial EVs nowadays. The vehicle speed can efficiency in drive cycles. The remainder of this article is
be regulated by a motor controller (Tolbert and Peng, organized as follows. In the next section, we describe the
1998). The dynamic and economic performances under vehicle specification and simulation environment. The
different driving conditions mainly depend on the control following section presents the investigation of a three-
strategies (Haddoun et al., 2007). speed transmission system and the optimization of gear
However, the efficiency varies as a motor works at ratios. Finally, the conclusions are given in Section 4.
different operating points, which describe the output speed
and torque of an electric motor (Pellegrino et al., 2012). 2. VEHICLE SPECIFICATION AND
The peak efficiency is obtained only within a relatively SIMULATION ENVIRONMENT
narrow speed region. Thus, a transmission is necessary to
2.1. Vehicle Specification
*Corresponding author. e-mail: [email protected] This work is based on a light EV prototype. A permanent

773
774 Z. ZHANG et al.

Table 1. Vehicle parameters.

Vehicle specifications Parameters
Vehicle mass (kg) 1098
2
Frontal projected area (m ) 2.0
Coefficient of aerodynamic drag 0.335
Wheel radius (m) 0.282
Rolling resistance coefficient 0.009
Max speed (km/h) ≥90
Rated power (kW) 25

magnet motor is used as the power source. Electric energy Figure 2. Output properties of the electric motor. The
is provided by a set of sealed valve-regulated lead-acid maximum torque curves are presented for the maximum
(VRLA) batteries. The parameters of the testing vehicle are tractive torque (blue cross), maximum continuous tractive
given in Table 1. torque (black cross), maximum regenerative torque (blue
The energy storage system on the EV prototype is a circle), and maximum continuous regenerative torque
Hawker Genesis 12V26Ah10EP sealed VRLA battery, (black circle), respectively.
which has 25 modules and weighs 275 kg. The open-circuit
voltage is set from 11.7 V to 12.89 V depending on the
state of charge (SOC) from 0 to 1.
The electric motor is modeled following the description
proposed by Bauch-Banetzky et al. (1999). The rated output
power is 25 kW, and the peak efficiency greater than 93%
can be achieved. The efficiency of the electric motor is
defined as:
Pout-
η = ------- × 100% (1)
Pin
In this work, when Pin = 0, η = 0 is assumed to simplify
the calculation. This motor can also work as a generator
while braking. The efficiency map is shown in Figure 1.
The efficiency significantly varies at different operating
points. An ideal efficiency of about 90% can be obtained at
a constant speed of nearly 2000 rpm. The efficiency contours Figure 3. Ideal motor speed curve versus output power
and maximum torque curves are presented in Figure 2. required.
To analyze precisely the relationship between the
efficiency and operating point, we present the ideal speed
curve with the highest efficiency as a function of the output single-speed reducer cannot adjust the change of the ideal
power required (Figure 3). The ideal motor speeds with motor speed, which limits the average efficiency during the
peak efficiency differ with increasing required power. A entire driving cycle. The ultimate objective of this work is
to keep the motor working steadily within the ideal
efficiency region under different driving conditions.

2.2. Simulation Conditions

The simulation test of the prototype vehicle is carried out
under the condition of the Urban Dynamometer Driving
Schedule (UDDS) driving cycle. This driving cycle
represents city driving conditions for light-duty vehicle
testing. The output power curve is calculated with the
ADVISOR2002 simulation tool. The statistics of power
distribution are shown in Figure 4. The motor is unloaded
(Pout = 0) almost 25% of the time in the overall cycle. For
Figure 1. Efficiency map of the electric motor. most of the remaining time, the output power fluctuates
 THREE-SPEED TRANSMISSION SYSTEM FOR PURELY ELECTRIC VEHICLES 775

Figure 6. Distribution of efficiency points of the EV (total

Figure 4. Distribution of output power (total of 1370 of 798 efficiency points; red denotes the three-speed trans-
operating points). mission and blue denotes the single-speed reducer). Points
with null input power are ignored.

from 0 kW to 10 kW, during which the ideal speed linearly

increases with the power.

3. THREE-SPEED TRANSMISSION SYSTEM

At the beginning the simulation test on the prototype is

conducted with the single-speed reducer. The maximum
vehicle speed of 90 km/h is achieved when the motor is
working at its peak speed of 6000 rpm. Thus, the gear ratio
i0 of the reducer is assigned to be 4.0039:1. Figure 5 shows
the distribution of operating points of the prototype vehicle
in a UDDS driving cycle with red crosses. The statistics of
the distribution of efficiency points with the single-speed
reducer are shown in Figure 6 as blue columns; points with Figure 7. Wheel speed curve (blue curve) and ideal motor
null input power are ignored. Notice that most efficiency speed curve versus time (red curve).
points are between 77% and 90%. The motor cannot
remain stable within the high efficiency region. Thus, a
three-speed transmission system is proposed to improve the
performance of the electric motor. The electric energy and wheels to adjust any possible speed ratio. The motor
consumption of the vehicle with the single-speed reducer is can work at the best speed with peak efficiency, which only
4.8931 × 103 kJ. depends on the output power required. Figure 7 presents
First, we consider an ideal CVT connecting the motor the wheel speed curve and ideal motor speed curve,
respectively. Correspondingly, the ideal speed ratio is
calculated at all time points, and 94.7% of the speed ratios
are found to be below 10:1. The remaining ratios are
assigned to be 10:1 for simplicity. Then the ideal speed
ratio curve versus time is obtained, as shown in Figure 8.
There remains a problem that the ideal speed ratio
appears continuous, but the gearbox transmission can
provide only three fixed gear ratios. The three gear ratios
should be optimized to achieve the maximum efficiency in the
overall driving cycle. We define a function η = Eff (pwr_r,
spd) to describe the efficiency η obtained by the motor at
the rotational speed to achieve the required power pwr_r.
The efficiency η can be calculated from the efficiency map
(Figure 1) by linear interpolation. For a fixed set of
Figure 5. Distribution of operating points of the prototype transmission with gear ratios i1, i2 and i3, the motor can
vehicle with a single-speed reducer (red cross). work at three different rotational speeds spd1, spd2 and spd3
776 Z. ZHANG et al.

Figure 8. Ideal speed ratio curve versus time.

to meet each vehicle speed veh_spd required by the UDDS.

Figure 9. Gear-shifting map. The blue, green and brown
The maximum efficiency is chosen to determine the best
points indicate the 1st gear ratio i1 = 9.2:1, the 2nd gear ratio
gear shift at each time during the driving cycle.
i2 = 5.4:1, and the 3rd gear ratio i3 = 3.2:1, respectively.
ηmax(pwr_r, veh_spd) = max(Eff(pwr, spd1), (2)
Eff(pwr, spd2), Eff(pwr, spd3))

The motor can convert electricity to kinetic energy to

power the vehicle, and it can also regenerate electric energy
during deceleration or downhill driving. In the regeneration
process, the kinetic power obtained from the wheels is also
defined as required power for consistency. In the driving
process, the output kinetic power is pwr+ and the input
electric power is pwr+/η. In the regeneration process, the
kinetic power into the motor is pwr− and the obtained
electric power is pwr−*η. In this paper, the regeneration
power is negative.
The symmetry of Figure 2 indicates that the motor has
the same efficiency with the same rotational speed and
power, regardless of driving or generation. As a result, the
maximum efficiency obtained in Equation 2 can simultane-
ously minimize the driving energy and maximize the
regeneration energy. The objective function of total energy
consumption during the driving cycle is calculated as:
1370

E = ∑ ∆t ⋅ (pwrj+ ⁄ ηmax (pwrj + , veh_spdj ) + pwrj- *ηmax (pwrj− , veh_spdj )) (3)

j=1

where spdj stands for the vehicle speed required at the time
j, and pwrj stands for the power required at the time j, with Figure 10. Gear-selection scheme during the UDDS
the subscripts j+ for driving and j− for regeneration, driving cycle. The blue curve denotes the original gear-
respectively. It is defined that pwrj- = 0 when driving and selection scheme from the gear-shifting map (a), and red
pwrj+= 0 when regenerating. The time step ∆t is 1s. The curve denotes that with a hysteresis strategy (b).
energy consumption values with all possible gear ratios are
calculated. The optimized gear ratios, namely, i1 = 9.2:1,
i2 = 5.4:1 and i3 = 3.2:1 are obtained to achieve the minimum can be selected at any time during travelling. In the UDDS
energy consumption. driving cycle, the gear-selection scheme is shown in Figure
To validate the gear-shifting strategy, the situations with 10 (a). However, the gear shift is too busy for practical use.
every possible required output power and vehicle speed are A hysteresis strategy is used in the gear-shifting assessment
calculate, where the power required depends on the angle program to avoid repeated shifts near the dividing line on
of acceleration pedal. Figure 9 presents the gear-shifting the gear-shifting map. The speed on the upshifting curve
map. Based on this gear-shifting map, the best gear ratio obtained from Figure 9 is increased by 20%, whereas that
 THREE-SPEED TRANSMISSION SYSTEM FOR PURELY ELECTRIC VEHICLES 777

the electric motor varies at different operating points. A

three-speed transmission is proposed to keep the motor
working steadily within the peak efficiency region to
reduce energy losses. The objective is defined as the
minimum energy consumption during the UDDS driving
cycle. The optimized gear ratios are confirmed as the 1st
gear ratio i1 = 9.2:1, the 2nd gear ratio i2 = 5.4:1 and the 3rd
gear ratio i3 = 3.2:1, respectively. The gear-shifting strategy
is subsequently obtained. Compared with the energy
consumption of a vehicle with a single-speed reducer, the
prototype vehicle with the three-speed transmission
consumes 9.3% less energy. In summary, the transmission
Figure 11. Motor efficiency curves during the UDDS driving
system is necessary to be applied on EVs for the ideal
cycle. The efficiency using the three-speed transmission
energy efficiency.
system (red curve) is almost always higher than that using a
single-speed reducer (blue curve). ACKNOWLEDGMENT−This work was supported by the
National Natural Science Foundation of China (No. 51175223).

REFERENCES

Andersson, C. (2004). On Auxiliary Systems in

Commercial Vehicles. Industrial Electrical Engineering
Department of Industrial Electrical Engineering and
Automation. Ph. D. Dissertation. Lund University.
Axsen, J. and Kurani, K. (2008). The Early US Market for
PHEVs: Anticipating Consumer Awareness, Recharve
Potential, Design Priorities and Energy Impacts.
Institute of Transportation Studies. University of
California. Davis. UCD-ITS-RR-08-22.
Bauch-Banetzky, D., Ruethlein, A. and Sonnenburg, R.
(1999). Permanent Magneterregte Synchronmaschine
mit Nd-Fe-B Dauermagneten fur den Einsatz in
Figure 12. SOC history curves of the battery with the Personenkraft Fahrzeugen. VDI Berichte 1459:
single-speed reducer and the three-speed transmission. Entwicklung Konstruktion Vertrieb HYBRIDANTRIEBE,
ISSN 0083-5560, ISBN 3-18-091459-9, Dusseldorf,
Germany.
on the downshifting curve is decreased by 20%. The new Bishop, J. D. K., Doucette, R. T., Robinson, D., Mills, B.
gear-selection scheme is presented in (b). The least and McCulloch, M. D. (2011). Investigating the
shifting interval is about 4s, which is acceptable for either technical, economic and environmental performance of
manual or automatic transmissions. The distribution of electric vehicles in the real-world: A case study using
optimized efficiency points is shown in Figure 6 (red electric scooters. J. Power Sources, 196, 10094-10104.
columns). The operating points with the three-speed Chen, C. H. and Cheng, M. Y. (2006). Design of a
transmission are more highly concentrated in the high multispeed winding for a brushless DC motor and its
efficiency region than those with the single-speed reducer. sensorless control. IEE Proc.-Electr. Power Appl., 153,
Figure 11 shows the motor efficiency curves during the 834−841.
UDDS driving cycle. The efficiency is promoted almost all Cheng, Y., Cui, S. and Chan, C. C. (2009). Control
the time with the reasonable gear-shifting strategies. The strategies for an electric variable transmission based
SOC history curves of the battery with and without hybrid electric vehicle. Proc. IEEE VPPC ’09, Sept.,
transmission are shown in Figure 12. The optimized energy 1296−1300.
consumption E is 4.4368 × 103 kJ, which is 9.3% less than Haddoun, A., Benbouzid, M. E. H., Diallo, D.,
that of the vehicle with the single-speed reducer. Abdessemed, R., Ghouili, J. and Srairi, K. (2007). A
loss-minimization DTC scheme for EV induction
4. CONCLUSION motors. IEEE Trans. Vehicle Technology, 56, 81−88.
Kim, J., Kim, T., Min, B., Hwang, S. and Kim, H. (2011).
In this work, a light EV model is built and simulation tests Mode control strategy for a two-mode hybrid electric
are carried out using the ADVISOR tool. The efficiency of vehicle using electrically variable transmission (EVT)
778 Z. ZHANG et al.

and fixed-gear mode. IEEE Trans. Vehicle Technology, Qin, D., Zhou, B., Hu, M., Hu, J. and Wang, X. (2011).
60, 793−803. Parameters design of powertrain system of electric
Lechner, G. and Naunheimer, H. (1999). Automotive vehicle with two-speed gearbox. J. Chongqing
Transmissions: Fundamentals, Selection, Design and University, 34, 1−6.
Application. Springer. Berlin. Rahman, Z., Ehsani, M. and Butler, K. L. (2000). An
Loveday, E. (2011). Vocis says two-speed electric vehicle investigation of electric motor drive characteristics for
transmission will boost battery life. http://green.autoblog. EV and HEV propulsion systems. SAE Paper No. 2000-
com/2011/08/19/vocis-says-two-speed-electric-vehicle- 01-3062.
transmission-will-boost-ba/ Suh, B., Frank, A., Chung, Y. J., Lee, E. Y., Chang, Y. H.
McKeegan, N. (2011). Antonov's 3-speed transmission for and Han, S. B. (2011). Powertrain system optimization
electric vehicles boosts efficiency by 15 percent. http:// for a heavy-duty hybrid electric bus. Int. J. Automotive
www.gizmag.com/antonov-3-speed-transmission-ev/ Technology 12, 1, 131−139.
19088/ Thomas, C. E. (2009). Transportation options in a carbon-
Pellegrino, G., Vagati, A., Guglielmi, P. and Boazzo, B. constrained world: Hybrids, plug-in hybrids, biofuels,
(2012). Performance comparison between surface- fuel cell electric vehicles, and battery electric vehicles.
mounted and interior pm motor drives for electric Int. J. Hydrogen Energy, 34, 9279−9296.
vehicle application. IEEE Trans. Ind. Electron., 59, 803− Tolbert, L. M. and Peng, F. Z. (1998). Multilevel converters
811. for large electric drives. APEC ’98, Anaheim, California,
Pearre, N. S., Kempton, W., Guensler, R. L. and Elango, V. Feb., 530−536.
V. (2011). Electric vehicles: How much range is required Zackrisson, T. (2003). Modeling and Simulation of a
for a day’s driving? Transportation Research Part C, 19, Driveline with an Automatic Gearbox. M.S. Thesis.
1171−1184. Royal Institute of Technology. Stockholm. Sweden.