Abstract

Over the past 25 years, a new form of optimization and search technique was refined using the same theories developed to explain human evolution, called the Genetic Algorithm. The algorithm is based on the premise that, when individuals exist in a competitive environment where resources are limited, only the fittest individuals will survive. This thesis attempts to use genetic algorithms to solve a complex optimization problem, which involves a real world simulation where trucks compete to make a profit, by buying and selling commodities in a country filled with different types of dealers. The genetic algorithm is used to evolve the trucks over generations, by modifying the strategies that they use to control their behaviour, in an attempt to produce more profitable trucks. The project was implemented using the C++ programming language. The purpose of the thesis and its implementation was to see if one could use object oriented techniques, such as inheritance and dynamic binding to achieve the genetic variation of the trucks. The results of the project are moderately successful. This is likely due number of strategies available to the trucks. By increasing the numb algorithm may produce trucks of superior genetic structure.