International Journal of Energy Engineering 2013, 3(2): 74-96
DOI: 10.5923/j.ijee.20130302.05
A GA-ANFIS Self Regulating Scheme for Induction
Motor Filter Compensation
Ahmed M. Othman
1,*
, Adel M. Sharaf
2
1
Electrical Power & Machine Department, Zagazig University, Zagazig, 44519, Egypt
2
Department of Electrical and Computer Engineering, School of Science and Engineering, Habib University, Karachi, Pakistan
Abstract The paper presents a Novel Green Plug-Filter Compensation Scheme developed by the Second Author and
controlled by an integrated Genetics algorithm (GA) with Adaptive Neuro-Fuzzy Inference System (ANFIS) controller
developed by the First Author for gain adjusting of a PID tri-loop stage control scheme applied on single phase Induction
motor. A Tri Loop dynamic error controlled technique is used to reduce inrush current conditions, improve energy
utilization, ensure soft starting, reduce inrush current as well as effectively ensure motor dynamic speed tacking.The
proposed technique is used to adjust the feeding of PWM switching of GP-FC by finding the optimal control gain settings
that dynamically minimize the global dynamic error. Digital simulations are provided to validate the effectiveness of this
device in improving the power quality and system stability.
Keywords Adaptive Neuro-Fuzzy Inference System (ANFIS), Dynamic Green Plug-Modulated Filter Compensator,
Genetic Algorithm (GA), Green Power Filter, Tri-loop Dynamic Control
1. Introduction
A novel Green Plug-Modulated Filter Compensator
(GP-FC) is used as green energy efficient plug
compensation scheme. The Simple modulated
filter/Capacitor compensation scheme, developed by the
Second Author and controlled by an integrated Genetics
algorithm (GA) with Adaptive Neuro-Fuzzy Inference
System (ANFIS) controller developed by the First Author,
is a member of a family of Energy efficient, Soft Starting
Switched/Modulated FACTS based Compensation Devices
for single phase and three phase motorized, inrush and
nonlinear loads. Active and Reactive powers have direct
impact on the energy efficiency, efficient utilization, power
factor, power quality and voltage profile of the system.
There are many techniques to supply and compensate for
load reactive power requirements needed by
nonlinear/inrush/motorized type loads. So fast control
action is needed. By building on technologies developed for
FACTS and LC Switched Compensators and high-power
CSI and VSI-electronic converters and drives, it is offered a
number of advantages in control of power systems,
including speed and accuracy of the controlled response.
Advanced control and improved semiconductor switching
of these devices have provided distinguished solutions for
power quality enhancement[1]-[5].
* Corresponding author:
ahmed_othman80@yahoo.com (Ahmed M. Othman)
Published online at http://journal.sapub.org/ijee
Copyright © 2013 Scientific & Academic Publishing. All Rights Reserved
The functions of the Energy Efficient, soft starting and
reactive compensation green plug scheme can be power
factor correction, power quality enhancement, efficient
utilization, dynamic voltage control, inrush current
reduction and dynamic speed reference tracking. GP-FC
may be used in AC power system for various applications,
from controlling reactive power to the system to improving
voltage regulation and power factor by reducing
transient/inrush content in voltage and current supplied to
the dynamic nonlinear/inrush type motorized load. Beside
the power demand requirements, some contingencies and
negative sequence /ripple content are among other factors
that are crucial when it happens to power quality
issue[6],[7]. The most important point is to find the optimal
dynamic self- regulating switching patterns for the GP-FC
devices, this selection can be adapted by AI trends. In
recent years, AI theory applications have received
increasing attention in various areas of power systems such
as operation, planning, and control. The effect of different
controllers as conventional and adaptive AI controllers can
be compared to conclude the most effective controller.
The main benefit in using the genetics algorithm is the
ability of GA to reach the optimal solutions, which
guarantees that if the GA is well self designed and trained
that will lead to the most achieved level from the desired
performance for the system under any problem space.
About the Adaptive Neuro-Fuzzy Inference (ANFIS), we
can first state that ANFIS is a merging system between the
neural network system and fuzzy logic system. Therefore, it
has the resultant benefits of both systems. The fuzzy system
is a very efficient tool in the controlling actions and the