This paper aims at achieving global optimal solution of complex problems, such as traveling salesman problem (TSP), using extended version of real coded genetic algorithms (RCGA). Since genetic algorithm (GA) consists of several genetic operators, namely selection procedure, crossover, and mutation operators, that offers the choice to be modified in order to improve the performance for particular implementation, we propose three new crossover techniques for Real Coded Genetic Algorithms, which will improve the quality of solution as well as the rate of convergence to the optimum solution. Methods proposed for crossover operators are inspired by asexual reproduction commonly observed in nature. In this regard, new crossover techniques proposed incorporates the concept of Boltzmann's distribution (BD) for escaping local optima by allowing hill-climbing moves and Metropolis Algorithm (MPA), where, survival of offspring is tested before transit to new generation. Finally, these three methods are compared on various aspects like rate of convergence and quality of final solution among each other and against other randomized algorithms.
[1]
Kalyanmoy Deb,et al.
Optimization for Engineering Design: Algorithms and Examples
,
2004
.
[2]
Lawrence. Davis,et al.
Handbook Of Genetic Algorithms
,
1990
.
[3]
John Geraghty,et al.
Genetic Algorithm Performance with Different Selection Strategies in Solving TSP
,
2011
.
[4]
Eugene L. Lawler,et al.
Traveling Salesman Problem
,
2016
.
[5]
David Beasley,et al.
An overview of genetic algorithms: Part 1
,
1993
.
[6]
David E. Goldberg,et al.
Genetic Algorithms in Search Optimization and Machine Learning
,
1988
.
[7]
Frank Neumann,et al.
Proceedings of the Genetic and Evolutionary Computation Conference 2016
,
2016,
GECCO 2016.
[8]
Gregory Gutin,et al.
The traveling salesman problem
,
2006,
Discret. Optim..