tailieunhanh - Robot Localization and Map Building Part 8

Tham khảo tài liệu 'robot localization and map building part 8', kỹ thuật - công nghệ, cơ khí - chế tạo máy phục vụ nhu cầu học tập, nghiên cứu và làm việc hiệu quả | 12 Consistent Map Building Based on Sensor Fusion for Indoor Service Robot Ren C. Luo and Chun C. Lai Intelligent Robotics and Automation Lab National Taiwan University Taipei Taiwan 1. Introduction Consider the actual applications of an intelligent service robot ISR it is expected that an ISR will not only autonomously estimate the environment structure but also detect the meaningful symbols or signs in the building it services. For example an ISR has to locate all the docking stations for recharging itself. For an ISR to lead a customer in the department store to any location such as the toy department or the nearest restroom it must have the essential recognizing and guiding ability for its service. For this purpose to carry out an applicable self-localization and map building technique for the indoor service robot becomes important and desirable. In recent years the sensing and computing technology have made tremendous progress. Various simultaneous localization and mapping SLAM techniques have been implemented. The principle of SLAM is derived from Bayesian framework. The EKF-SLAM Durrant-Whyte Bailey 2006 is based on robot state estimation. However EKF-SLAM will fail in large environments caused by inconsistent estimation problem from the linearization process Rodriguez-Losada et al. 2006 Bailey et al. 2006 Shoudong Gamini 2007 . A full SLAM algorithm is using sequential Monte Carlo sampling method to calculate robot state as particle filter Montemerlo et al. 2002 Montemerlo et al. 2003 . But the technique will grow exponentially with the increase of dimensions of the state space. Another full scan matching method is suitable for the environment reconstruction Lu Milios 1997 Borrmann et al. 2008 . But the pose variable will also grow enormously depending on the sampling resolution. Based on the practical needs of a service robot patrol in the building it is desirable to construct an information map autonomously in a unitary SLAM process. This chapter .