International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]

Volume 3, Issue 5, September-October 2014

ISSN 2278-6856

Vehicle Stability Maintenance Through Online

Estimation of Parameter Uncertainties
1

J.Ramprabu and , 2G.Sindhuja

Assistant Professor, Department of Electrical and Electronics Engineering,

Kumaraguru College of Technology, Coimbatore -641049.

Second Year, M.E. (Embedded Systems), Department of Electrical and Electronics Engineering,
Kumaraguru College of Technology, Coimbatore -641049.

ABSTRACT
A Robust system to maintain the vehicle stability for active
safety is focused in this paper. The online observance and
estimation of parameter uncertainties such as road gradient
and vehicle mass and its convergence with the real-time results
is the proposed idea. The system is implemented through a
sensor based hardware system and a mathematical model that
proves the theoretical estimation of parameter uncertainties.
The observed parameters prove to maintain the vehicle
stability and the Yaw stability. They are compared with realtime results from sensor-fixed system that resembles a car.
The sensor results converge with the standard results. The
simulation of this system is done using MATLAB.

KEYWORDS:- vehicle stability, active safety, vehicle

mass, road grade

1. INTRODUCTION
Ensuring the vehicle stability and providing active safety
vehicles has become the priority of automotive industries
today. The key parameters to maintain the vehicle stability
includes Vehicle mass, road gradient, air drag friction, tire
pressure measurement etc. Research and development
incite this fact. The insight of road gradient and vehicular
mass seems to be the promising criterions in vehicular
technology. They support the production of excellent active
safety vehicles and ensure efficient usage of fuels. [1]
intents Global Positioning System (GPS) to measure the
road gradient on a ground vehicle. The estimated values
are then harnessed with engine torque information in order
to predict vehicle mass, rolling resistance and aerodynamic
drag. An algorithm based approach for road grade
estimation coupled with real-time sensor data from vehicle
and GPS is dealt in [2]. Per Sahlholm et al. [3] proposed a
method that utilizes a piecewise linear road model derived
from highway design methodologies. Here the sensors
found on standard heavy duty vehicles are used to calculate
Road grade. A two-stage estimation strategy is used by M.
McIntyre et al. in [4] to find road gradient and vehicle
mass. Adaptive least-squares estimation strategy forms the
first stage to determine the vehicle's mass and an estimate
for a constant road grade. Lyapunov theory is the
significant estimation strategy employed. In [5] a gradient
benchmark system incorporates an inclinometer sensor and

Volume 3, Issue 5, September-October 2014

an acceleration-based fault correction algorithm are used.

This strategy is used to inherit accurate of the road
gradient using a sensorless estimation method. Research
in [6] shows an active on-line estimation scheme for slope
of road and vehicle mass estimation. It combines an
extended Kalman filter (EKF) to generate on-line
parameter estimates along with a model predictive
controller (MPC) in order to control vehicle speed
trajectory. Simultaneous estimation of vehicle mass and
road grade is done using two different for heavy-duty
vehicles in [7]. an observer to estimate mass and timevarying grade and a recursive time-varying least square
method with forgetting are deployed successively. Vehicle
manufacturers consider mainly on the active safety
technologies that involves yaw rate control, traction
control, Roll stability control etc. [10] and [11] deals with
antilock braking systems. [8] and [9] research upon vehicle
stability control These technologies owe improved control
of the vehicle dynamics. Any stability control systems in
vehicular technology require information on the vehicles
inertial parameter to perform prior system calibration.
Vehicle mass is such a criterion. Variable loads are
common in SUVs and trucks. In the case variable load
vehicles, the online estimation of vehicular mass is of
utmost importance. M. N. Mahyuddin et.al. provide a
novel approach of road grade and online vehicular mass
estimation using only the vehicles velocity and the driving
torque achieved based on a novel adaptive nonlinear
observer through sliding mode term. Practical results from
three different heavy duty ground vehicles and six road
traversals are presented in [14]. The results obtained help
in deriving about direct power demand making use of it
gearbox systems. Thus we can regulate the consumption of
fuel in engine. Yaw stability, traction control and side-slip
angle maintenance are highly important parameters during
downhill travel especially when the path is steep [6]. These
considerations demand the knowledge of vehicle mass and
road gradient for producing active safety hill riding
vehicles. [14] and [2] takes aid of the GPS in gaining
knowledge about the road gradient. But this proves to be
an expensive technology when compared to [7] which
deploys only sensors to arrive at results. There are research
on sensorless approach in vehicles mass and road gradient
estimation [1], [4], [5], [7], [13], [19]. They also involve
model-based approach in converging to real-time results. It

Page 132

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]
Volume 3, Issue 5, September-October 2014
is found that there is instability and infinite growth in [22]
and [23] which is due to the presence of an unstable
integrator. The classical RLS algorithm basically delays
converge fast, exponential convergence in particular,
excluding finite-time convergence [21], [24]. [36]
Overcomes these issues through Persistent Excitation
condition. This paper is organized as follows: In Section 2,
a system design is presented in a generic form. Section 3
presents the practical results obtained through hardware
implementation and simulation. Section 4 presents
conclusion of the paper.

ISSN 2278-6856

Figure 1 Block diagram of the Car model

2. SYSTEM DESIGN
The proposed system is implemented using a hardware test
system that resembles a car. The simulation is done in
MATLAB and results are obtained. The theoretical model
for parameter estimation is developed through Lyapunov
theory. The hardware implementation involves sensors that
can provide the real-time values of parameters that are
dealt.

Start

Initial parameters of the

vehicle are observed

Vehicle runs on Slope

2.1 Theoretical model

Lyapunov theory proves to be the accurate and promising
algorithm to find parameter uncertainties and maintain the
stability of the vehicle. The differential equations formed
clearly explain the state of the dynamic vehicular system
under study. The state of the vehicular model in this
proposed system is represented as

Where x is mass of the vehicle, a is the velocity of the

vehicle and is the force of engine. The air drag, viscous
friction and rolling friction are avoided in this calculation
as the system under test is a prototype and not a ground
vehicle. The objective of developing a robust method to
perform parameter estimation resides in deriving the
theoretical model, which extends as the future scope of this
work.

Sensor results from the

vehicle are observed

Vehicle load is varied in

calculated steps

Parameters are compared

between the estimated and
sensor values

Stop

Figure 2 Flowchart of the proposed system

2.2 Hardware Implementation

The hardware of this system involves a car model that has
an Electronic Control Unit (ECU) and necessary sensors
interfaced to ECU to obtain the real-time parameter values.
The ECU has TMS570LS20216 microcontroller of
Hercules family. This is a microcontroller best suited for
automotive active safety applications, produced by Texas
Instruments. It is a 32-bit RISC Microcontroller involving
ARM-Cortex-R4F based high-performance CPU with
floating point unit. The Dual Core in Lockstep feature
makes this microcontroller suitable for Safety Critical
Applications. The vehicle mass weighs around 7kgs. A
motor driver and a DC motor are interfaced in the
prototype. The inclinometer is interfaced to the ECU to
measure the inclination in the pathway.

Volume 3, Issue 5, September-October 2014

Figure 3 Vehicle on a road with different grade

The motor is operated by the ECU through the motor
driver. The microcontroller employed in the ECU has
Controller Area Network (CAN) which makes the
communication easier between peripherals attached and
the ECU. The motor gets its driving force from the ECU.
The slope profile used for testing is given below.

Page 133

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]
Volume 3, Issue 5, September-October 2014

estimation of parameters. The application of Lyapunov

theory with back stepping approach is taken for
consideration for theoretical model development.

Table 1 Test Slope Profile

Slope
Profile
1

5o

REFERENCES

10o

15o

[1] H. Bae, J. Ryu, and J. Gerdes, (2001) Road grade and

vehicle parameter estimation for longitudinal control
using GPS, in Proc. IEEE ITSC, pp. 166171.
[2] H. Jansson, E. Kozica, P. Sahlholm, and K.-H.
Johansson,(2006) Improved road grade estimation
using sensor fusion, in Proc. 12th Reglermote,
Stockholm, Sweeden, , pp. 16.

3. PRACTICAL RESULTS
16

14

Real Value
Sensor Value

<---Angle =15degree---->

[3] P. Sahlholm and K. Johansson, (2010)Segmented road

grade estimation for fuel efficient heavy duty
vehicles, in Proc. IEEE Conf. Decision Control, , pp.
10451050.

12

10

Road Grade

ISSN 2278-6856

<---Angle = 10degrees---->
8

[4] M. McIntyre, T. Ghotikar, A. Vahidi, X. Song, and D.

Dawson, (Sep. 2009)A two-stage Lyapunov-based
estimator for estimation of vehicle mass and road
grade, IEEE Trans. Veh. Technol., vol. 58, no. 7, pp.
31773185.

<----Angle = 5 degrees---->

0
0

10

20

30

40

50

60

70

80

90

100

Time (s)

Figure 4 Sensor values and Real values of Road Gradient

vs Time

[5] S. Mangan and J. Wang, Development of a novel

sensorless longitudinal road gradient estimation
method based on vehicle can bus data, IEEE/ASME
Trans. Mechatronics, vol. 12, no. 3, pp. 375386,.
[6] V. Winstead and I. Kolmanovsky, (Jun. 2007)
Estimation of road grade and vehicle mass via model
predictive control, in Proc. IEEE CCA, 2005, pp.
15881593.
[7] A. Vahidi, M. Druzhinina, A. Stefanopoulou, and H.
Peng, (2003,)Simultaneous mass and time-varying
grade estimation for heavy-duty vehicles, in Proc.
Amer. Control Conf., vol. 6, pp. 49514956.

Figure 5 Sensor values and Real values of Vehicle Mass

vs Time
The standard test values and real-time values of the
parameters are compared from the sensor results. The
results seem to converge well with the standard test values.
The simulation is done using MATLAB.

4. CONCLUSION
An effective estimation system to predict the road gradient
and vehicle mass is presented in this paper. The test values
for parameters are prefixed for comparison with the real
time values. The test values are compared with the realtime results obtained from the sensors fixed in the vehicle
under test. The simulation of results proves the
convergence of test parameters with the real-time results.
Vehicle stability can be easily controlled with these
parameters. Future scope of this research extends to
finding a robust theoretical model that can minimize the
convergence error and increase the accuracy of the online

Volume 3, Issue 5, September-October 2014

[8] R. Daily and D. Bevly,(Apr. 2004) The use of GPS for

vehicle stability control systems, IEEE Trans. Ind.
Electron., vol. 51, no. 2, pp. 270277.
[9] J. Stephant, A. Charara, and D. Meizel,(Apr. 2004)
Virtual sensor: Application to vehicle sideslip angle
and traversal forces, IEEE Trans. Ind. Electron., vol.
51, no. 2, pp. 278289.
[10] Y. Hori, (Oct. 2004) Future vehicle driven by
electricity and control-research on four-wheel-motored
UOT electric march II, IEEE Trans. Ind. Electron.,
vol. 51, no. 5, pp. 954962.
[11] W. Wang, I. Li, M. Chen, S. Su, and S. Hsu, (May
2009 ) Dynamic slip-ratio estimation and control of
antilock braking systems using an observer-based
direct adaptive fuzzy-neural controller, IEEE Trans.
Ind. Electron., vol. 56, no. 5, pp. 17461756.
[12]

R. Frampton and P. Thomas,(Mar. 2007)

Effectiveness of electronic stability control systems in
Great Britain, Dept. for Transp., Veh. Safety Res.
Centre (VSRC), Loughborough, U.K., Tech. Rep.

Page 134

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]
Volume 3, Issue 5, September-October 2014
[13] H. Fathy, D. Kang, and J. Stein, (Jun. 2008) Online
vehicle mass estimation using recursive least squares
and supervisory data extraction, in Proc. Amer.
Control Conf., Seattle, WA, USA, pp. 18421848.
[14] P. Sahlholm and K. Johansson, (Nov. 2010) Road
grade estimation for look-ahead vehicle control using
multiple measurement runs, Control Eng. Pract., vol.
18, no. 11, pp. 13281341.
[15] P. Lingman and B. Schmidtbauer, (Jan. 2002) Road
slope and vehicle mass estimation using Kalman
filtering, Veh. Syst. Dyn., vol. 37, no. 1, pp. 1223.
[16] J. Park, J. Kong, H. Jo, Y. Park, and J. Lee, (Sep.
2001) Measurement of road gradients for the
development of driving modes including road
gradients, Proc. Inst. Mech. Eng. D, Autom. Eng.,
vol. 215, no. 9, pp. 997986.
[17]

S. Mangan, J.Wang, and Q.Wu, (2002)

Measurement of the road gradient using an
inclinometer mounted on a moving vehicle, in Proc.
IEEE lnt. Symp. Comput. Aided Contr. Syst. Design,
pp. 8085.

[18] J. Barrho, M. Hiemer, U. Kiencke, and T. Matsunaga,

(Jul. 2005) Estimation of elevation difference based
on vehicles inertial sensors, in Proc. 16th IFAC
World Congr., Praque, Czech Republic, p. 1906.
[19] A. Vahidi, A. Stefanopoulou, and H. Peng, (Jan.
2005) Recursive least squares with forgetting for
online estimation of vehicle mass and road grade:
Theory and experiments, Int. J. Veh. Mech. Mobility,
vol. 43, no. 1, pp. 3155.
[20] A. Vahidi, A. Stefanopoulou, and H. Peng, (Jan. 2003
)Experiments for online estimation of heavy vehicles
mass and time-varying road grade, in Proc. IMECE,
pp. 16.
[21] P. Iaonnou and J. Sun, (1996) Robust Adaptive
Control. Englewood Cliffs, NJ, USA: Prentice-Hall.
[22] V. Adetola and M. Guay, (Apr. 2008) Finite-time
parameter estimation in adaptive control of nonlinear
systems, IEEE Trans. Autom. Control, vol. 53, no. 3,
pp. 807811.
[23] V. Adetola and M. Guay, (Sep. 2010) Performance
improvement in adaptive control of linearly
parameterized nonlinear systems, IEEE Trans.
Autom. Control, vol. 55, no. 9, pp. 21822186.
[24] S. Bittanti, P. Bolzern, and M. Campi,(Dec. 1990)
Recursive least-squares identification algorithms with
incomplete excitation: Convergence analysis and
application to adaptive control, IEEE Trans. Autom.
Control, vol. 35, no. 12, pp. 13711373.
[25] S. P. Bhat and D. S. Bernstein, (May 1998)
Continuous finite-time stabilization of the
translational and rotational double integrators, IEEE
Trans. Autom. Control, vol. 43, no. 5, pp. 678682.

Volume 3, Issue 5, September-October 2014

ISSN 2278-6856

[26] C. Edwards and S. Spurgeon, (1998) Sliding Mode

Control: Theory and Applications. Boca Raton, FL,
USA: CRC Press.
[27] J. Foreman, M. Yazdouni, S. Russo, G. Chan, A.
Spiers, G. Herrmann, and P. Barber,(2009) Hardware
in the loop validation of a gradient and weight
estimation algorithm and longitudinal speed control
using a laboratory model car, in Proc. ICSE,
Coventry, U.K.
[28] G. Herrmann, S. Ge, and G. Guo, (Aug. 2008)
Discrete linear control enhanced by adaptive neural
networks in application to a hdd-servo-system,
Control Eng. Pract., vol. 16, no. 8, pp. 930945.
[29] J. Na, G. Herrmann, X. Ren, M. Mahyuddin, and P.
Barber, (Sep. 2011) Robust adaptive finite-time
parameter estimation and control of nonlinear
systems, in Proc. IEEE ISIC, pp. 10141019.
[30] P. Sahlholm, (2008) Iterative road grade estimation
for heavy duty vehicle control, M.S. thesis,
Elektrotek. Syst., Kungliga Tekniska Hogskolan,
Stockholm, Sweden.
[31] P. Sahlholm and K. Johansson,(2008) Road grade
estimation for look-ahead vehicle control, in Proc.
17th IFAC World Congr., pp. 33803385.
[32] S. Sastry and M. Bodson, Adaptive Control: Stability,
Convergence and Robustness. New York, NY, USA:
Dover, 1989.
[33] N. Shimkin and A. Feuer,(Sep. 1987) Persistency of
excitation in continuous system, Syst. Control Lett.,
vol. 9, no. 3, pp. 225233.
[34] J.-J. Slotine and W. Li,(Oct. 1990) Applied Nonlinear
Control. Englewood Cliffs, NJ, USA: Prentice-Hall.
[35] T. Soderstrom and P. Stoica, (1989) System
Identification, M. Grimble, Ed. Englewood Cliffs, NJ,
USA: Prentice-Hall.
[36] M N Mahyuddin, J Na, G Herrmann, X Ren, and P
Barber, (June 2014) Adaptive Observer-Based
Parameter Estimation With Application to Road
Gradient and Vehicle Mass Estimation , IEEE
Transactions on Industrial Electronics, Vol. 61, No. 6,
pp. 2851-2863.

AUTHORS
Prof.J.Ramprabu works as an Assistant
professor in the Department of EEE,
Kumaraguru College of Technology. He has an
Academic experience of 8 years. He is
currently Pursuing research in Anna University Chennai
and his area of working is renewable energy and embedded
system. He is a Member of the Indian Society for Technical
Education and Member of System society of India.

Page 135

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]
Volume 3, Issue 5, September-October 2014

ISSN 2278-6856

G.Sindhuja received B.E in Electronics and

Communication Engineering from Sri
Shakthi Institute of Engineering and
Technology, Coimbatore. She is presently
pursuing Post Graduation in Embedded
Systems from Kumaraguru College of Technology.

Volume 3, Issue 5, September-October 2014

Page 136