Zero acceleration algorithm with state of motion identifier for position estimation of wheeled robot

• Main Author: Tan, Zhi Zhong
• Format: Thesis
• Published: 2019
• Subjects: Q Science (General); TJ Mechanical engineering and machinery
• Online Access: http://eprints.sunway.edu.my/2397/

Position estimation is an irreplaceable element in navigation and autonomous technology. Position estimation system using inertial sensor poses a critical challenge to deliver accurate positioning information. The sensor noise and uncertainties causing drift in velocity and discrepancy in estimated position. The solution to this problem is crucial for robot’s navigation especially for a mobile service robot in commercial buildings and households. In this research, a state of motion identifier framework is proposed to apply a novel kinematic motion constraint to position estimation system. The proposed framework consists of zero acceleration algorithm which responsible to detect the time and duration of a wheeled robot when it is travelling at constant speed (zero acceleration), zero velocity update to detect the still phase of the wheeled robot and improved drift correction. The result of the evaluation proven that the proposed framework is capable of reducing the velocity drift as well as the discrepancy between estimated position and actual measurement. The proposed frame also shows the ability to complement the weakness of zero velocity update in high speed and long dynamic duration conditions.