Robot Planning based on Behavior Tree and Knowledge Graph

Abstract

Task planning refers to ordering executive actions to ensure the efficient and successful completion of the robot task. Unlike previous methods, we propose new ideas, based on Knowledge, that can make robot strategy, in addition to automatically enhancing behavior tree building. Various knowledge bases are applied to our robot brain system, serving as robot planning and used for lab exhibitions. According to And Or Graph and graphing database, a special knowledge graph is designed to simplify the robot's planned routes. We analyze the proposed approach and validate our ideas through experiments on human guide robots.