Machine Coded Genetic Algorithms For Real Parameter Optimization Problems

Abstract

   

In this paper, we introduce a new encoding-decoding strategy for the floating-point genetic algorithms and we call the genetic algorithms which use this strategy Machine Coded Genetic Algorithms. We suggest applying classical crossover and mutation operations on the byte representations of real values which are already encoded in memory. This is equivalent to use a 256-unary alphabet with 8 genes for a single real value. Use of byte representations makes the classical genetic operators interpretable in floating-point chromosomes and increases the search capabilities in a wide range without losing accuracy. This strategy also decreases the computation time needed for the genetic operators. Simulation studies show that our strategy performs well on many test functions by means of converging to global optimum and time efficiency.

Key Words

 

 

: Genetic algorithms, Chromosome encoding, Real parameter optimization.

Keywords

• Genetic algorithms, Chromosome encoding, Real parameter optimization.

References

1. Deb, K., “Multi-Objective Optimization using Evolutionary Algorithms”, John Wiley & Sons, (2004).
2. Elsayed, S.M., Sarker, R.A., Essam, D.L., “Multi- operator Based Evolutionary Algorithms for Solving Constrained Optimization Problems”, Computers & Operations Research, 38: 1877-1896 (2011). Fogel, D.B., Ghozeil, A., “A