Auton Robot
DOI 10.1007/s10514-015-9530-4
A sliding mode controller with a nonlinear disturbance observer
for a farm vehicle operating in the presence of wheel slip
Javad Taghia
1
· Xu Wang
1
· Stanley Lam
1
· Jay Katupitiya
1
Received: 26 January 2015 / Accepted: 7 December 2015
© Springer Science+Business Media New York 2015
Abstract A sliding mode controller with a nonlinear dis-
turbance observer is proposed and developed to control a
farm vehicle to accurately track a specified path. The vehi-
cle is subjected to lateral and longitudinal slips at front and
rear wheels. The unpredictability of ground contact forces
which occur at the wheels while traversing undulating, rough
and sloping terrains require the controllers to be sufficiently
robust to ensure stability. The work presented in this paper
is directed at the practicality of its application with both
matched and unmatched uncertainties considered in the con-
troller design. The controller is designed using an offset
model derived from the kinematic model and its operation
is verified by simulation and field experiments. In the sim-
ulations, the kinematic model based controller is used to
control both a kinematic model and a dynamic model of
a tractor to verify the performance of the kinematic model
based controller. The proposed controller is compared with
two other nonlinear controllers, namely, back stepping con-
trol and model predictive control. In the field experiments, the
three controller were used to control the physical tractor to
Electronic supplementary material The online version of this
article (doi:10.1007/s10514-015-9530-4) contains supplementary
material, which is available to authorized users.
B Javad Taghia
taghia.javad@gmail.com; j.taghia@unsw.edu.au
Xu Wang
xu.wang3@student.unsw.edu.au
Stanley Lam
s.lam@unsw.edu.au
Jay Katupitiya
j.katupitiya@unsw.edu.au
1
School of Mechanical and Manufacturing Engineering,
University of New South Wales, Sydney, NSW 2052,
Australia
follow a specified path. Simulation and experimental results
are presented to show that the proposed controller demon-
strated the required robustness and accuracy at all times.
Keywords Sliding mode control · Path tracking ·
Autonomous ground vehicles · Autonomous farming
1 Introduction
The use of autonomous systems in many industries is rapidly
advancing. In the agricultural industry, the application of
smaller modular autonomous vehicles is promising in deliv-
ering productivity advantages while they are operating with
high accuracy on agricultural terrain. Among these are the
ability to lay the crop highly accurately according to a pre-
specified crop layout plan and to use the knowledge of crop
location to manage the crop right throughout the cropping
season. Similarly, knowing the crop locations also means
knowing where crop has not been planted and hence her-
bicides can be accurately applied to where they are needed
such as the inter-row space. The high accuracy autonomous
path tracking capabilities not only bring in the cost sav-
ings, but also considerable ecological and environmental
advantages. In general, large-scale agricultural activities are
time-consuming, tedious, hazardous and the proposed path
tracking accuracy is unable to be achieved by human opera-
tors. Without a sophisticated crop localization system, high
precision path-tracking is essential for reliable maneuvering
of an autonomous system in the field among the growing
crop to ensure that the crop is not damaged. The aim of this
study is to develop controllers that can deliver such accuracy.
Although the focus of this paper is on farming vehicles, more
specifically tractors, the presented approach is applicable to
123