Investigation of a Multiple Chromosome Evolutionary Algorithm for Bus Driver Scheduling and Other Problems

Genetic Algorithms have recently been used to solve various transport scheduling problems such as bus driver scheduling. These algorithms have so far produced good results on small problems, but as the problems become larger, they become more difficult to solve. A method of improving the performance of the genetic algorithm may be to `reinvent' it, using inspiration from complex biological systems to change both the way the problem is represented and the way the algorithm operates. The novel feature of this new representation is that it incorporates redundant information into the genetic model by using a genotype that contains multiple chromosomes. This alternative genetic algorithm could also be used to solve other problems that have traditionally been difficult for the standard algorithm to solve. This work investigates the use of such a genetic algorithm on various common problems before applying it to the bus driver problem. Acknowledgements I would like to thank my supervisor Da..