GA-BASED GAIT GENERATION OF SONY QUADRUPED ROBOTS

This paper presents a GA (Genetic Algorithm) approach to the development of locomotion gait for Sony quadruped robots. The selection of GA parameters such as the population size and recombination methods is made to be flexible and strive towards optimal performance autonomously. An interactive software environment with an overhead CCD camera is used to evaluate the performance of the generated gaits. The experimental results are given to show that the stable and fast gaits have been achieved. Keyword: Evolutionary Programming, Quadruped Robots, Genetic Algorithms, Gait Generation