Fuzzy hybrid navigation of an active mobile robotic assistant: A multisensory fusion approach

Abstract

This paper presents methodologies and techniques for multi-sensory fuzzy hybrid navigation of an active mobile robot assistant for the elderly people in known, cluttered and dynamic indoor environments. For self-localization, the geometric and topological representations are combined together to maintain the advantages of them and counteract their disadvantages. By using RFID system, the estimation of position hypothesis for similar regions in a known environment can be reduced. The global optimal path composed of via points is determined by using Dijkstra's dynamic programming approach. Fuzzy adaptive speed control law is designed to make the robot adapt to various users' speeds. Safe fuzzy hybrid navigation is achieved by fusing adaptive speed, goal-seeking and obstacle- avoidance from the sensing data of laser scanner and ultrasonic rangers. In addition, Experimental results are conducted to show the effectiveness and merits of the proposed methods.