The Evaluation of Provided Methods in SLAM Problem and a Method Development in Order to Use in Multi Robot

Abstract

Abstract. Simultaneously Localization and Mapping (SLAM) in mobile robots is anissue which is of utmost importance to Robotic Science researchers. The goal is to determine the position of intelligent mobile robots and their navigation to provide the overview map of the environment when no knowledge of it exists. So far, various methods have been presented in group robots to solve the SLAM problem. After about six years of research on mobile robots in Yazd Robotic Association and of applying available algorithms on rescue real robots, we will review major approaches on a robot and will introduce a new method that has been applied on two platforms of manual and autonomous robots in IranOpen 2011 competitions. We provide a method based on ROA-BlackWelized Particles filter (RBPF) algorithm and Forgetting curve in Multi robots which, compared to the limitations of collecting environmental data in the past two years, has provided satisfactory results. In the end, we simulate the consequence method in order to provide the robot's overview and local maps from the environment in PlayerStage software.