A Rotating Sonar and a Differential Encoder Data Fusion for Map-Based Dynamic Positioning

• Published in: Journal of intelligent & robotic systems Vol. 29; no. 3; pp. 211 - 232
• Main Authors: Yang, Haeyong, Park, Kyucheol, Lee, Jang Gyu, Chung, Hakyoung
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Nature B.V 01-11-2000
• Subjects: Coders; Dynamical systems; Dynamics; Encoders; Errors; Mathematical analysis; Rotating; Sonar; Studies
• ISSN: 0921-0296; 1573-0409
• DOI: 10.1023/A:1008187932210

In this paper, a dynamic positioning system using a rotating sonar and a differential encoder is proposed. The method is implemented by employing an indirect feedback Kalman filter. The state equation is written for encoder propagation and its error characteristic. A measurement equation describes a map-based measurement equation based on rotating sonar sensor data. In other words, sonar data compensates for the system and navigation errors of the differential encoder. The positioning system calculates the position and headings of a mobile robot. The real-time calculation is performed by a map-based measurement update utilizing wide-angle beam characteristics of the sonar sensor and the Kalman filter. In addition, an observability analysis for the positioning system is performed. Experimental results show that the proposed hybrid positioning system successfully provides accurate position and headings in real-time. The position and heading errors arc bounded within few centimeters and within few degrees, respectively.[PUBLICATION ABSTRACT]