Research On Simultaneous Location And Map Building Of Mobile Robot For Virtual Reality

This paper researches on Simultaneous Localization and Map Building (SLAM) of Autonomous Mobile Robots(AMR), in order to build a complete system that makes robots have the ability of autonomous map development in unstructured environment. FastSLAM based on particle filter is a highly efficient simultaneous localization and mapping algorithm. Robot updated information needs 0 (M log K) time. This is much simpler to the traditional method based on the Kalman filter (need to 0 (K2) time).Firstly, the paper summarizes the SLAM problem so far from a proposal to develop a simple process, research institutions, application environment, the characteristics of the corresponding sensors, and analysises of the traditional method of the basic tenets of SLAM with the process.Secondly, this paper makes a overall modeling to the Hebut II-wheeled mobile robot, representing the odometer model mainly used for robot positioning, the body model and motion control orders used to measure the location of road signs perceptron sensor model.Mobile robot body has a distribution of ultrasonic distance sensor array perceived as a tool for measuring the distance from the landmarks, a milestone of its own robot-assisted positioning tools. Body Motion Control traverse through unknown environment, useing FastSLAM algorithm to maintain the two-dimensional planar map data.Thirdly, the paper highlights the history of the development of SLAM problems, the basic theory, various research methods and their comparison ; and it elaborates on this issue to create maps used by the algorithm method FastSLAM principle, the particulate filter methods, sampling principles and calculation processes, and FastSLAM algorithm with the corresponding data association and maps that way. By FastSLAM algorithm calculation, the robot completed its trajectory and the environment in determining the location of the signs.Finally, we developed a control software, which will control algorithm applied to Hebut II robot, and make a lab experiment. The experimental results show that algorithm is fast and accurate, which proves that particle filter based FastSLAM solution to the simultaneous map building localization problem for region-coverage autonomous mobile robot is indeed possible the method has higher computational efficiency and better localization accuracy.