On low-cost and high precision localization of snowplow robot: A fuzzy and visuality approach

摘要

High precision localization is essential to autonomous control of unmanned vehicles (mobile robots). Most existing methods for localization are either complex and costly, or bear low accuracy and poor robustness. In this work, a structurally simple and functionally effective localization system is designed and developed for snowplow robots with successful on-site application. The major hardware involved in the system is a regular camera mounted on a rotating platform in the robot, together with an artificial beacon placed nearby the boundary of working envelope of the robot. Simple fuzzy algorithms are utilized to control the camera, allowing the images of the surrounding area to be continuously scanned and collected by the camera. The autonomous localization of the robot is achieved by calculating the relative distance between the robot and the artificial beacon. Numerous experiments and no-site tests demonstrate and confirm that the designed localization system works well with high precision.