This thesis presents and discusses a potential method for solving the dynamic obstacle avoidance problem using contemporary work with artificial neural networks (ANNs) and genetic algorithms (GAs) in combination with an imitation of a biological genetic process called segmental duplication. ANNs, GAs and segmental duplication are merged in the project to form SDNEAT, a type of evolutionary artificial neural network (EANN) system based on NeuroEvolution of Augmenting Topologies, or NEAT. The system is then used to develop an artificial neural network system that attempts to navigate environments incorporating both static and dynamic obstacles.
