Co-evolutionary Robotics using two kinds of Neural Networks

摘要

The evolutionary robotics approach has been attracting lots of concern in the fields of robotics and artificial life, since it can automatically synthesize controllers by taking the embodiment and the interaction dynamics between the robot and its environment. In this paper, we propose a novel evolutionary robotic approach in the quasi-ecosystem. The proposed approach has two kinds of neural networks. Moreover, the proposed evolutionary robotic approach employs the combination of BP learning and GA search. Experimental results show effectiveness of the proposed approach.