International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-2, July 2019
4205
Published By:
Blue Eyes Intelligence Engineering
& Sciences Publication
Retrieval Number: B3404078219/19©BEIESP
DOI: 10.35940/ijrte.B3404.078219
Abstract: This paper attempts to use GSO algorithm to tune a
PID controller that can be used to control a satellite using
reaction wheels. These have a higher order transfer function and
the controller will be more difficult to tune due to this. To do this,
a satellite is chosen which controls its attitude using reactions
wheels. An axis is chosen and the reaction wheel along this axis is
taken into consideration. A PID controller is then attached to this
system. PID (Proportional-Integral-Derivative) are one of the
most popular controllers used due to their broad applications and
easy to design nature. The PID controller is tuned using
Glowworm Swarm Optimization (GSO) algorithm and then the
system is checked against a step input. The optimization of the
controller is done by minimizing the time weighed absolute error
cost function.
Index Terms: Glowworm Swarm Optimisation (GSO)
Algorithm, Optimisation, PID Tuning, Satellite attitude control
I. INTRODUCTION
Satellites today carry a wide variety of loads and there are
many launched each year with new methods and new
hardware for attitude control. Reaction wheels have proven to
be one of the best methods for attitude adjustment of a
satellite using no extra fuel, but only electricity to run its
motors. Electricity, which can be readily made available using
solar power available on almost all satellites. In references [4]
and [8], they have attempted to optimize a PID controller for 2
different cases using Fruit Fly Optimization algorithm. In the
papers [12], [2] and [3], PSO algorithms are used to tune PID
controllers used in different systems. The results they
achieved are positive, however the system transfer functions
were of a lower order than the ones in the case of a satellite.
Reference [9] tests the PSO algorithm on various fitness
functions and finds it to be on par with the other, though
sometimes it could not arrive at the optimum value. It
highlights the problems of choosing the correct fitness
function and shows how increasing the generation size after a
certain level only results in marginal improvements in the
result. Reference [1] compares PSO with Zeigler Nichols and
PSO obtains better results.
This paper is an attempt to control a satellite using GSO,
which can find the global optima, as opposed to PSO, which
may sometimes arrive on the local optima and miss the global
Revised Manuscript Received on July 15, 2019.
Pulkit Sharma, Department of Aeronautical and Automobile
Engineering, Manipal Institute of Technology, Manipal Academy of Higher
Education, Manipal, Karnataka, India 576104.
Vishnu G Nair, Department of Aeronautical and Automobile
Engineering, Manipal Institute of Technology, Manipal Academy of Higher
Education, Manipal, Karnataka, India 576104.
optimum. This paper is an attempt to control a satellite’s
attitude using Glowworm Swarm Optimisation. The same is
then done using Genetic Algorithm for comparison. The
satellite is assumed to be a rigid body and the dynamics is
expressed in the form of a MIMO system. One of the SISO
systems is then selected from these. This SISO system is then
connected with a classic PID controller. The optimization is
carried out using GSO, which is used to find the optimal
values of the PID constants.
II. METHODS
In a chosen satellite with reaction wheels, one axis is chosen
and our PID controller is attached to it. This controller is then
tuned using GSO algorithm to find the optimum values of the
PID constants. This is then compared with the constants
obtained using genetic algorithm. This is done by deriving the
dynamics equation of the satellite and reaction wheels, the
disturbance is modelled and then attached with it the transfer
function of the PID controller.
A. Glowworm Swarm Optimisation (GSO)
This is an evolutionary optimization that is used to find the
global maxima or minima of a given multi-modal function. It
is based on the glowworm in which the less luminous glow
worms move towards the more luminous glow worms that are
nearby [6].
The algorithm of GSO consists of several steps defining how
each glowworm behaves and how they are used to determine
the peaks of a multi-modal function. There are three basic
mechanisms at work [6].
Fitness broadcast is carried in the luminescent pigment of the
glowworm, called luciferin. It is proportional to the value of
the function at their current position and this works under the
assumption that the luciferin sensed by other glowworms does
not change with distance [6]. A glowworm tends to move
towards its neighbour if the neighbour glows brighter than
itself. In case of multiple such neighbours, a probabilistic
mechanism is used [6]. A glowworm will take another
glowworm to be in its neighbourhood only if it is within its
neighbourhood radius. This radius is modulated by using a
heuristic and it can only vary within a certain range [6].
GSO starts by placing n well dispersed glowworms in the
workspace. All the glowworms initially contain an amount l
o
of luciferin. Each cycle consists of three phases, namely the
luciferin update phase, the movement phase and the
neighbourhood range update
phase. These phases together
form the GSO algorithm [6].
Tuning of PID Controller using Glowworm
Swarm Optimisation on a Satellite Attitude
Control Reaction Wheel
Pulkit Sharma, Vishnu G Nair,