The Evolutionary Locomotion of Tripedal and Quadrupedal Biomorphic Robots

Abstract

Evolutionary robots have become an interesting topic recently. These robots can achieve certain goals via evolutionary algorithms without specifying all the detailed actions. The robot interacts with the environment and receives natural feedback from the environment regarding the fitness of its goal. In this paper, we study biomorphic robots. Our robots have three or four legs, and share the goal of moving as fast as possible on a flat plane without an understanding of robot kinetics. The feedback was collected via a supersonic sensor and the evolutionary algorithm is a genetic algorithm. We found that the biomorphic robots can move faster with the motion sequence generated by the evolutionary algorithm than by the motion sequence assigned by human programmers. The structures of robots, in this case, the number of legs, might affect the speed convergence.