Analysis of Convergence Effect Via Different Genetic Operations D.F.Fam & S.P. Koh & S.K. Tiong & K.H. Chong

Department of Electronic & Communication Engineering, Universiti Tenaga Nasional,
Km 7, Jalan Kajang-Puchong, 43009 Kajang, Selangor.
[email protected],[email protected],[email protected],[email protected] Abstract: Genetic Algorithms (GAs), Evolution Strategies (ES), Evolutionary Programming (EP) and
Genetic Programming (GP) are some of the best known types of Evolutionary Algorithm (EA)where it is
a class of global search algorithms inspired by natural evolution. In this research, genetic algorithm is
one of the optimization techniques used to maximize the performance of solar tracking system .This
paper presents analysis of convergence effect via different genetic operations used in Genetic Algorithm
as explained in the introduction and methodologies. Simulation Results will demonstrate the ability of
GA to produce different solutions via different genetic operations to maximize the performance of solar
tracking system.

Index Terms—genetic algorithm, solar tracking, genetic operations

1. Introduction

The basic principles of GA was developed by
John Holland [1] They have since been
reviewed and the concepts have been applied
on a wider range [2],[3],[4] in today’s world.
The GA is derived from Darwin’s theory of
Natural Selection. A GA mimics the
reproduction behavior observed in biological
populations and employs the principal of

interface card for maximum tracking. The
system tracks the sun with +/-10
0
in both axes.
The tests and analyses explained that the solar
tracking system using GA increases the output
voltage to 7.084% in comparison to that with
no GA [5].
Syamsiah Mashohor et al. evaluated the best
combination of GA parameters to optimize a
solar tracking system for PV panels in terms
of azimuth angle and tilt angle. Simulation
results demonstrated the ability of the proposed
GA system to search for optimal panel
positions in term of consistency and
convergence properties. It also has proved the
ability of the GA-Solar to adapt to different
environmental conditions and successfully track
sun positions in finding the maximum power by
precisely orienting the PV panels.[6]
However, recent researches for GA based solar
tracking system are based on the traditional GA
algorithm structure which is shown as below:
// populations //
t=0
Step 1= Initialize P (t)
Evaluate P(t)
While (Solution NOT found OR Max
Generation NOT Reached)
Do

in the early evolution time. If the highly fit
individuals are local optima areas, then newly
generated offspring from the parents are also
near the local optima areas.
In this coming methodology section, an
explanation of different genetic operations
will be studied and results section will show
the best genetic operations in preventing
premature convergence problem. 2. Methodology

Methodology part is divided into few sub
sections below:
1) Conventional crossover and mutation
2) Crossover only
3) Clone and selective mutation 2.1 Conventional Crossover and Mutation

Using conventional method of having
crossover and mutation in Genetic Algorithm
will affect its performance. One of the typical
problem is Premature Convergence Problem
[11,12].Most individuals in a prematurely
converged situation are located at some local
optimum areas and they can’t get out of the
local optimum areas because the exploration

the crossover point, the following offspring are
produced:

Offspring1: 11001|111
Offspring2: 00100|010
Crossover can not generate quite different
offspring from their parents because it uses
acquired information from their parents.
2.3 Clone and selective mutation

In most function optimization problems, their
input variables are encoded into the binary
strings of individuals. Since the binary strings
represent binary numbers for each variable, the
higher the bit position of string is, the larger the
bit weight has. From this, it is helpful to mutate
some part of strings of individuals according to
their fitness. That is, if an individual has low
fitness, then we mutate the most significant part
in order to largely change because we regard
the individual to be far from the global
optimum. Otherwise, we mutate the least
significant part in order to do fine tuning
because the individual has high probability to
be near global optimum. This selective
mutation can make genetic algorithms fast
approach to the global optimum and quickly get


Table 1: Conventional GA simulation
parameter
Simulation Parameter Value
Maximum Generation
Population, p
o

Chromosome length
Selection Method
Crossover Rate, p
c

Mutation Rate, p
m

Mutation Point, m
p

No.BestChromosomes
Kept, k
b

Crossover Type
50
10
8
Roulette Wheel
80%
0.025

8
Roulette Wheel
80%
1
Dynamic

Table 3: Conventional GA simulation
parameter with clone and selective mutation
Simulation Parameter Value
Maximum Generation
Population, p
o

Chromosome length
Selection Method
Crossover Rate, p
c
Elitism Rate, E
c
Selective Mutation
No.BestChromosomes
Kept, k
b

Crossover Type
50
10

0.08
Generation
Fitness value
Best: 0.018654 Mean: 0.018654Best fitness
Mean fitness

Graph 1 : Best Fitness Value- 0.018654 using
conventional GA

10 20 30 40 50 60 70 80 90 100
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Generation
Avergae Distance
Average Distance Between Individualsdata1

Graph 2 : Average distance between

Graph 3 : Best, worst and mean score for each
generation using conventional GA 1 2 3 4 5 6 7 8 9 10
0
0.5
1
1.5
2
2.5
3
3.5
4
Selection Function
Individual
Number of childrenstate.SelectionGraph 4 : Number of children that is produced
by each individual using conventional GA

0 5 10 15 20 25 30 35 40 45 50
0.01

0.08

0.1
0.12

0.14

0.16

Generation

Best, Worst, and Mean Scores
Best Score
Median Score
Worst Score