Robust Filter Development to Pursue Robotic Localization and Mapping Problem

• Main Authors: Hamzah, Ahmad, Namerikawa, Toru
• Format: Conference or Workshop Item
• Language: English
• Published: 2011
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://umpir.ump.edu.my/id/eprint/3533/1/Robust_filter_development_to_pursue_robotic_localization_and_mapping_problem.pdf; http://umpir.ump.edu.my/id/eprint/3533/

This paper deals with the development of a robust filter known as H Filter , as an approach to provide alternative solutions in robotic localization and mapping problem. In this problem, normally the environment conditions and noise characteristics are unknown which makes the Kalman Filter approach suffers. H Filter is proposed due its advantage in estimating conditions in non-Gaussian noise which Kalman Filter cannot offer. The analysis is shown by comparing the performance between H infinity Filter and Extended Kalman Filter for the above problem. Based on the simulation analysis and preliminary experiments, there is a high reliability in estimation of H Filter compared to Extended Kalman Filter regarding the robot location and landmark positions. Hence, H Filter can provide better solution especially in specific environment conditions with unknown noise characteristics.