The development of intelligent control system for animal robot using multisensor fusion

Abstract

With the rapid growth of computer technology, the applications of robots are not limited to industrial automation. Entertainment robots will be one of the most important emerging applications in the 21/sup st/ century. The paper presents an intelligent control system for an animal robot using multisensor fusion technology to deal with touching, collision, and falling conditions. After analyzing the walking of a four-legged animal robot, we use two-level multisensor fusion techniques to fuse sensory data from IR proximity, mercury switch, touch sensor, thin-film switch, and limit switch. The fused multisensor data is used to determine the contact position of patting collision, and the standing-up direction after falling. Experimental results indicate that the proposed method can effectively fuse multisensor data for the control of the animal robot.