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

net/publication/339409270

Power Efficient e-Bike with Terrain Adaptive Intelligence

Conference Paper · July 2019

DOI: 10.1109/ICCES45898.2019.9002178

CITATIONS READS

5 1,705

5 authors, including:

Haneesh K. M.

13 PUBLICATIONS   22 CITATIONS   

SEE PROFILE

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

Voltage Stability View project

Hybrid Electric Vehicles View project

All content following this page was uploaded by Haneesh K. M. on 24 February 2020.

The user has requested enhancement of the downloaded file.

Power Efficient e-Bike with Terrain Adaptive
Intelligence
D. S. H. Abhilash , Iqbal Wani, Kiran Joseph, Rashmi Jha, Haneesh K. M.
Electrical and Electronics Engineering
CHRIST (Deemed to be University)
Bangalore, India
[email protected], [email protected]

Abstract— Electric bicycles or e-bikes are gaining momentum of these researches [2]. E-bikes are mainly of two types; a
in the market as they are offering a smooth, noise less and pure electrical powered one with electrical motor which is
pollution free option for individual transportation in cities as well controlled through a throttle by the user to obtain different
as in countryside. E-bikes are usually with a battery powered power depending on the intention of the user and the second
electric motor drive with an additional option for pedaling. In one is a pedal assisted one or a pedelec, in which the power is
this work a low cost e-bike was designed and developed with a supplied by both the electric motor as well as pedalling power
brushless DC hub motor with controllers. For smart control, together. In any e-bike the range of the bike is determined by
smartphone was used a console and the e-bike can be controlled the battery capacity, motor power, driving terrain, human load
using a mobile application which was connected to the e-bike
and the control mechanism [3]. Pedelecs need pedalling to
through Bluetooth. The controller will pick the gradient of the
terrain and will control the power of the motor, which results in
actuate the motor operation and the comfort level is less
energy saving. Predicted range of the e-bike, speed, acceleration compared to a fully battery powered one, but will be having
and total distance covered were displayed in the console along higher range per charge. There were some studies conducted
with the geographical position on the map and throttle control on optimization of power in pedelecs [4].
options. The bike with the proposed control tested and the results Fully battery powered e-bikes with BLDC hub motor offer
were giving a reduction in current drawn from the battery. more comfort and can replace conventional fuel based two
wheelers in populated cities, especially for short distance
Keywords— Electric bicycle, E-mobility, Range prediction,
Intelligent control, Terrain intelligence.
commutation. Lower distance per charge and throttle control
are the main drawbacks of such vehicles [5]. There were few
techniques adopted to improve the efficiency of the vehicle
I. INTRODUCTION and thus the battery backup, like the regeneration during the
Electric bicycles generally known as e-bikes are the slowing down regimes [5]. The regeneration techniques were
bicycles integrated with electric motors for propulsion. These improved with the help of bi-directional controllers [6] and
e-bikes are battery powered with either pedal assist provision with EMC compatible controllers [7]. The BLDC motor based
or with throttle control to ride through different terrains. E- e-bikes can additionally save power during when the power
bikes offer more comfort and less noisy operation than usual demand is less, like in the case of climbing down the hill.
pedal powered bicycles, so the market for e-bikes is growing Since the controller is feeding the supply to the BLDC motor
all over the world. Increase in fuel prices and polices by based on the rotor position, there is an option to save power
various countries to reduce emissions are also driving the depending on the road conditions [8, 9].
market for e-bikes. Riding bicycles is a healthy and eco- In this work, we designed and developed a battery
friendly practice and e-bikes with pedalling option is adding powered e-bike with a BLDC hub motor. The rider can control
comfort and efficiency to that. Global opportunity and the e-bike using a hand controlled throttle. To improve the
industrial forecast report by Allied Market research shows that efficiency and the range, a gradient senor based controller is
the market of e-bikes expected to grow annually at a rate of 5 provided, which help in suitably giving the throttle based on
% to a total value of 23.83 billion USD in 2025 from the total the gradient of the road in the auto run mode. The rider‘s
value of around 20 billion USD in 2018. Market analysis also mobile phone is used as the console for the e-bike, which will
indicates that countries with higher population density are display the predicted range, speed, acceleration as well as the
having higher demand for e-bikes and the Asian countries are total distance covered. The bike tested in different terrains and
leading the trend [1]. Also the development in Lithium-ion the results showed improved power efficiency as well as
battery technologies after the year 2012 and the availability of longer range per charge.
low cost, efficient and light vehicles are also offer better
prospects for the market for e-bikes. This paper is organised as follows; in section II the design
of the e-bike is described, controller details are given in
Since e-bikes offer a pollution free and comfortable drive, section III. The gradient based control mechanism is described
lot of research and developments are happening in the in section IV, integrated e-bike with controllers is presented in
technological area which offer better solutions in mobility as section V, results and discussion are given in Section VI and
well as in sports goods category. Multiple fuel options, hybrid conclusion is given in section VII.
vehicle and improved efficiency are the major areas of many
Fig. 1. Forces acting on a bike, while climbing a slope Fig. 2. Block diagram of BLDC motor controller

current values. So over voltage and over current protections

II. DESIGN OF THE E-BIKE are also incorporated in the controller.
A single rider bike with some minor additional load
carrying capacity which can run for a medium trip length of 50 IV. GRADIENT BASED CONTROL MECHANISM
km is explored. To obtain better comfort and easier mobility,
The gradient of the road is measured by a gradient sensor;
a fully electric powered bike is preferred over pedelecs. If the
the sensor will feed different voltages based on the gradient of
rider needs to shift to pedalling mode, the motor can be
the track. If the gradient is negative, the sensor will feed zero
switched off and run as a normal bicycle. The design starts
and with increasing positive slope, it will give positive
with the propulsion motor, the specification of which depends
voltages. This voltage values are augmented by the throttle
on the weight of the rider and the other loads on the bike apart
output voltage values in the auto run mode. In negative slopes,
from the traction force. There are few options available for the
the motor will be switched off and in positive slopes the motor
motor type, but a BLDC can provide more torque, stability,
will be powered based on the voltage values fed by the sensor
higher efficiency as well as lower noise compared to DC
and processed by the microcontroller.
motors. Even though the control circuit of the BLDC is
complex, lack of commutators and smaller size will give The rider can opt for normal riding mode or auto run
BLDC motors an edge over its competitors. mode. As shown in Fig. 3, the modes are selected based on the
relay selection. In auto run mode, the gradient sensor will give
The following values were selected for the selection of
voltage values, which the microcontroller will process and
motor power. Mass of bike frame and accessories – 12 kg,
categorize into five different levels. For negative slopes it will
battery and controller – 3 kg, cyclist – 80 kg, load – 5 kg. So
give zero voltage and for the positive slopes between 10 0 and
total weight – 100 kg. As given in Fig. 1, the bike is climbing
500 it will a voltage between 3.4V-6.6V and which is
a slope with an inclination of θ, the traction force Fp has to
calibrated as comparing with the throttle voltage values. While
overcome the normal component of weight Fd and the friction
climbing down the slope, the motor will supply with zero
and windage force Fwf. If the linear speed of the vehicle V is
power and at higher positive slope motor will be running at
assumed as 20 km/h. An additional aerodynamic load of 20W
highest power. This control scheme is based on the terrain
is assumed when riding at V km/h [10]. The power is
intelligence and which was implemented to improve the power
calculated as 260 W and a 36V, 250 W, 250 rpm BLDC hub
efficiency and thus to improve the range of the bike [11]. The
motor with a maximum torque 32Nm selected to fix on the
system is compatible with the pedalling mode also, where with
rear wheel hub.
the help of a switch; we can totally disconnect the motor from
the battery. In case of the battery charge is exhausted the rider
III. BLDC MOTOR CONTROLLER can shift to pedalling mode just with toggling the switch.
BLDC motors offer higher torque over a range of speed
with higher efficiency than DC motors. The motor use a set of
MOSFET bridge to energize the phase windings according to
the position of the rotor with respect to the stator. The position
of the rotor is sensed through a set of Hall Effect sensors. In
three phase wound motor, three set of Hall Effect sensors are
used to precisely locate the position and these signal output
from the sensor will be used as the feedback for speed control.
For the e-bike speed control, a throttle is connected on the
handle, where the rider can adjust it with hand or fingers.
Similarly braking and ON-OFF control signals are also fed to
the controller through cables. Fig. 2 shows the outline of the
controller. In addition to the power control, the controller
should be capable of protecting against abnormal voltage and Fig. 3. Gradient sensor based auto run mode
V. E-BIKE INTEGRATED WITH CONTROLLERS
A normal pedal bicycle was converted into the e-bike by
fixing a 250W, 36V BLDC hub wheel motor in the hub of rear
wheel of the bike. Modifications were done on the wheel
holder as well as chassis to accommodate the motor in the
hub. The tire size was 66 cm and with a frame size of 45.7 cm.
The brake levers are replaced with hybrid brake levers which
are capable of controlling mechanically, also capable of
producing braking signals for the motor controller [12]. A 5V
electric thumb throttle was fixed on the right hand side of the
handle to control the acceleration. The throttle was provided
with a key to switch ON or OFF the electric motor mode.
The power for the motor is provided with a 37V, 10.4 Ah Fig. 5. The e-bike integrated with controllers and sensors
Li-ion battery packs. It was a custom designed pack with 40 Terrain intelligence of the e-bike was provided with a set
numbers of 3.7V, 2800 mAh batteries connected as 10 in of gradient sensors. These sensors were calibrated to measure
series and 4 such units connected in parallel. The batteries are the gradient of the road whether it is positive or negative.
with 2C rating and suitable for faster charging. A battery pack These sensors will provide voltage signals based on the slope
is provided with charge controllers and protective circuits. of the road and for better accuracy these sensors were fixed on
Total weight of the battery pack is around 3kgs and a metal the horizontal part of the frame. These signals were fed to the
holder is provided on the chassis for fixing it comfortably on microcontroller, which will give a voltage signal output
the bike. calibrated with the throttle signal values. When the bike is
To monitor the State of Charge (SoC) of the battery a climbing down the hill, the voltage signal output will be zero
voltage sensor and a current sensor were included in the power from the microcontroller and the motor will be disconnected
circuit. The voltage sensor was connected across a high from battery.
resistance potential divider circuit to reduce power losses and Fig. 6 shows the various modes of operation of the e-bike.
to obtain more reliable readings. The voltage and current The rider can select the pedalling mode with the help of the
sensor output were fed to a Arduino Uno microcontroller and ‗kill switch‘ provided. In normal mode, the rider has to apply
SoC calculations are done there. The current sensor output is the throttle based on the slope of the road. If the rider opts for
used to monitor the current values and to provide protection in the ‗auto-run‘ mode, the motor will be driven automatically by
case of abnormal operations. These sensor values are also used providing appropriate power calculated based on the gradient
to calculate the instantaneous power consumption. The Hall sensors.
Effect sensor outputs were used for speed measurement. Fig. 4
shows the outline of the microcontroller part. Modern e-bikes are provided with consoles to display the
measurements like speed, time etc. But such consoles are
Fig. 5 shows the actual view of the e-bike integrated with costly and not suitable for low cost e-bikes. In this project we
the battery pack, motor controller and sensors. The had developed a mobile phone application which will help the
components added to the e-bike were placed carefully to rider to use own smart phone as a console while riding. The
reduce the aerodynamic resistance as well as to improve the mobile application will fetch the predicted range, total
comfort of the rider. The controller and the microcontroller distance covered and the instantaneous power consumption of
circuits were provided with an enclosure to ensure safety and the motor in addition to the current taken from the supply
to have a better appearance. through Bluetooth from the microcontroller unit. Presently the
mobile phone application is developed for Android platforms,
but can be easily modified to make compatible for other
platforms also. A simple holder for the smart phone was
provided on the frame of the bike to safely hold the phone.

Fig. 4. Outline of the Microcontroller circuit Fig. 6. Flow chart for various modes of operation of the e-bike
VI. RESULTS AND DISCUSSION TABLE I PERFORMANCE OF 250W, 36 V BLDC MOTOR

The 36V, 250W, BLDC hub motor was tested for the Sl. Voltage Current Torque Speed Output Effici-
performance using a mechanical loading with spring dials. The No (V) (A) (Nm) (rpm) power ency
motor shown very good speed regulation with a maximum (W) (%)
speed of 275 rpm at very light load and with a minimum speed 1 37.2 1.1 0.55 275 15.85 38.72
recorded during the test was 224 rpm during the test. The 2 37.2 1.14 0.584 272 16.64 39.24
motor was capable of producing a maximum torque of 19.08 3 37.2 1.32 0.78 269 21.98 44.76
Nm. The maximum efficiency of the motor was found to be 4 37.2 1.56 1.12 264 30.98 53.38
81.54% which shows it is much superior to a DC motor. The 5 37.2 2.1 1.86 261 50.86 65.10
maximum efficiency recorded at around the half full load and 6 37.2 2.82 2.675 257 72.02 68.65
the efficiency at higher loads found to be in close range of 7 37.2 3.65 3.86 254 102.71 75.65
maximum efficiency. The load test values are tabulated in 8 37.2 4.48 4.86 252 128.30 76.99
Table I. Fig. 7 shows the variation of torque with the output
9 37.2 5.28 5.92 248 153.81 78.31
power. Fig. 8 and Fig. 9 show the variation of speed and
10 37.2 6.11 7.024 245 180.28 79.32
efficiency with the output power respectively. The vibration of
11 37.2 7.09 8.426 243 214.50 81.33
the e-bike with the BLDC was found to be very small even at
higher motor speeds. 12 37.2 8.21 9.864 241 249.04 81.54
13 37.2 9.54 11.048 238 275.46 77.62
14 37.2 10.52 12.98 234 318.20 81.31
15 37.2 11.86 14.42 232 350.47 79.44
16 37.2 12.98 16.02 230 386.01 79.94
17 37.2 14.44 17.88 228 427.08 79.51
18 37.2 15.9 19.08 224 447.74 75.70
The integrated e-bike was tested on different terrains and
found to be with easily controllable with thumb throttle. At
higher slope the motor takes higher current to supply the
higher power requirement. Noise and vibrations were less
compared to a normal pedal bicycle. The maximum speed
Fig. 7. Torque Vs Output Power curve recorded with a rider weighing 80 kg was 25 km/h. The
battery pack was capable of providing around 1.5 hours of
back up with a single charge. Around 40 km range with a
single charge make it a good choice for usual scenarios like
urban travel, countryside commutation, in tourism and in
sports.
The terrain intelligence system was introduced only in
auto-run mode. In auto-run mode no throttle control is
required and only braking signal to be applied whenever
required. In auto-run mode, the gradient sensor output was
used to generate the signal corresponds to throttle signal. In
usual throttle control, the throttle voltage signal was varying
from 3.4V to 6.6V. In auto-run mode also the voltage signal is
Fig. 8. Speed Vs Output Power curve calibrated to these levels using the microcontroller

TABLE II PERFORMANCE OF 250W, 36 V BLDC MOTOR

Slope of the terrain Voltage output from

(Degrees) microcontroller (V)

Less than -5 0
-5 to 10 2.4
10-20 2.8
20-30 3.2
30-40 3.5
Fig. 9. Efficiency Vs Output Power curve More than 40 3.66 (max)
program. Only four levels of voltage signals were provided in
TABLE III POWER SAVINGS IN TERRAIN INTELLIGENT MODE
the terrain intelligent system. In Table II the levels used were
consolidated. Presently only a rule based algorithm was used
Slope of the Current (A) Current (A) Power
to control the motor power. A more complex intelligent terrain Thumb Auto-run Mode saving (%)
algorithm can be used to control the motor. (Degrees) throttle mode
The terrain intelligence system was capable of saving
power consumed by the motor. In normal thumb throttle Less than -5 0 0 0.00
controlled mode, the throttle was applied based on the -5 to 10 3.4 3.3 3.03
assessment of the rider based on his comfort and the road 10-20 4.5 4.2 7.14
conditions. Most of the time the perception of the rider will be
fuzzy in nature and there will be definitely some scope for 20-30 5.6 5.2 7.69
optimization in throttle control. To verify the power savings in 30-40 6.4 6.0 6.67
the terrain intelligent control mode, the currents taken by the
More than 40 6.9 (max) 6.5 (max) 6.35
motor in throttle control mode and terrain intelligent mode
were compared. Since the power consumed by the motor is
directly proportional to the current drawn from a constant DC
voltage supply, the reduction in current while riding in same I. CONCLUSION
terrain with different modes can be seen as the difference in E-bikes are gaining popularity especially in populated
power. The e-bike was tested under some known slopes in cities. Riding comfort, long range per charge of battery and
different operating modes and the results are tabulated in capital cost are the major considerations while people
Table III. The reduction in current is evident in terrain selecting e-bikes in present market. In this work a BLDC rear
intelligent mode. The percentage power savings can be better hub motor powered e-bike was developed and tested in
if the e-bike run in a complex track with varying slopes. various terrains. A gradient sensor based throttle control
A mobile phone application was developed for Android provided in the auto-run mode was making it terrain intelligent
platform and the current, predicted range, speed and by providing throttle equivalent signals to the motor. The
instantaneous power consumption were displayed on it. The results showed that the terrain intelligent control is capable of
microcontroller fixed on the chassis is acquiring the voltage of saving power consumed by the motor while riding under
battery, current drawn from the power circuit and calculating different slope conditions. A mobile phone application
the above parameters and will send to the mobile application developed and which can display the current, predicted range,
through the Bluetooth module. Fig. 10 shows the display of instantaneous power of the motor can serve as a console for
mobile application. The rider can select the mode and drive the e-bike, thus the cost of the e-bike can be reduced.
accordingly. The predicted range is calibrated to give the Presently only a rule based terrain intelligent system was
minimum distance the bike will run before recharging the employed but a better algorithm can be tried to improve the
battery power savings. Regeneration and charging of battery while
pedaling are the future works to be added to the project.
ACKNOWLEDGMENT
The authors acknowledge the facilities provided by the E-
mobility Research & Development Facility at Faculty of
Engineering, CHRIST (Deemed to be University), Bangalore
for the successful completion of this work.

REFERENCES
[1] Salmeron-Manzano, E.; Manzano-Agugliaro, F., ―The Electric Bicycle:
Worldwide Research Trends‖, MDPI Energies, Vol. 11, pp 1894, 2018.
[2] C. Abagnale, M. Cardone, P. Iodice, R. Marialto, S. Strano, M. Terzo,
G. Vorraro, ―Design and Development of an Innovative E-Bike”, in
Proc. 71st Conference of the Italian Thermal Machines Engineering
Association, ATI2016, 14-16,September 2016, Turin, Italy, 2007, pp. 1–
6.
[3] P. Iodice, C. Abagnale, M. Cardone, S. Strano, M. Terzo, G. Vorraro,
―Performance Evaluation and Environmental Analysis of an Electrically
Assisted Bicycle Under Real Driving Conditions‖, in Proc. of the ASME
Fig. 10. Display of the mobile application 12th Biennial Conference on Engineering Systems Design and Analysis
(ESDA2014), ESDA2014-20438, Vol. 1, Copenhagen, Denmark, 25–27
June, 2014, pp. 33-40.
[4] M. Cardone, S. Strano, M. Terzo, ―Optimal power-assistance system for commutation used for e-bike," 2013 International Conference on
a new pedelec model‖, in Proc. IMechE, Part C: Journal of Mechanical Advanced Electronic Systems (ICAES), Pilani, 2013, pp. 316-318.
Engineering Science,vol 3, pp. 332-345, 2015. [9] C. Dow, S. Chang, V. Bui and P. Liu, "A Vibration Reduction System
[5] E. Ceuca, G. Brezeanu and V. Trifa, "Study for developing the energy for E-bikes," 2018 IEEE 9th Annual Information Technology,
recovering circuit for modern e-bike controller," 2015 International Electronics and Mobile Communication Conference (IEMCON),
Semiconductor Conference (CAS), Sinaia, 2015, pp. 245-248. Vancouver, BC, 2018, pp. 772-775.
[6] C. Dow, S. Chang, V. Bui and P. Liu, "A Vibration Reduction System [10] Roman Nadolski, Krzysztof Ludwinek, Jan Staszak , Marek Jaśkiewicz
for E-bikes," 2018 IEEE 9th Annual Information Technology, ―Utilization of BLDC motor in electrical vehicles‖, in Journal of
Electronics and Mobile Communication Conference (IEMCON), Przegląd Elektrotechniczny (Electrical Review), R. 88, NR 4a/2012
Vancouver, BC, 2018, pp. 772-775. [11] Chris Kiefer, Frauke Behrendt, ―Smart e-bike monitoring system: real-
[7] E. Ceuca, I. Pocan and O. Pop, "Electromagnetic compatibility analysis time open source and open hardware GPS assistance and sensor data for
of the assembly e-bike — Power electronic converter with recovery electrically-assisted bicycles‖, in journal of IET Intell. Transp. Syst., pp.
function," 2016 39th International Spring Seminar on Electronics 1–10,2015.
Technology (ISSE), Pilsen, 2016, pp. 398-401. [12] W. Chlebosz, G. Ombach and J. Junak, "Comparison of permanent
[8] V. N. Kumar, A. Syed, D. Kuruganti, A. Egoor and S. Vemuri, magnet brushless motor with outer and inner rotor used in e-bike," The
"Measurement of position (angle) information of BLDC motor for XIX International Conference on Electrical Machines - ICEM 2010,
Rome, 2010, pp. 1-5.

View publication stats