New Solutions for Kidnapped Robot Problem in Markov Localisation Algorithm

摘要

In order to let autonomous robot do not lose its way, two problems should be solved. Firstly, if the robot is kidnapped or slipped from a position to another position, it can find its position quickly. Secondly, if the robot is put into a featureless symmetrical environment, it can find its position as well. These situations are often encountered in real world. In this paper, we have new solutions for these situations. A modified Markov Localization Algorithm is given for robot to solve the kidnapped problem. A proposal, in which a robot is equipped with a gyroscope or compass, is offered to solve the problem in a featureless symmetrical environment.