APPROXIMATE SENSOR FUSION IN A NAVIGATION AGENT

摘要

A multiple sensor fusion model for a navigation agent based on rough integration is given in this paper. A rough measure of sensor signal values provides a basis for a discrete form of rough integral. This integral computes a form of ordered weighted average using a weighting factor determined by a classifier in the form of a Bet of "ideal" sensor values. In this paper, the focus is on classifying sensor signals relative to a classification interval of interest in guiding the navigation of a mobile robot. A navigation agent "looks" for rough integral values representing sensor signals to determine appropriate movements in a particular region of space. A navigation algorithm used by an agent to govern the movements of a mobile robot is given.