Evolution of automatic robot control with genetic programming

This paper investigates the problem of automatic control of a robot model in an arbitrary two-dimensional environment. The robot control is based on behavior evolution with the use of genetic programming (GP). A simulation environment is developed which allows automatic synthesis of different behaviors suited to the environment and user requirements. The influence that different robot properties and learning mechanisms have on the evolved behavior are studied in depth. The results show that the presented approach is able to produce effective control procedures.