Integrated Mobile Robot Control
Abstract
Real time mobile robot controllers have usually been designed with an emphasis on control theory ignoring the importance of system integration. This report demonstrates that useful mobile robots require a real-time controller with a wide range of capabilities in addition to control theory. These capabilities include: position estimation, mapping and tracking of paths, human interfaces, fast communication, multiple client support, and monitoring vehicle status for safety and debugging. An architectural framework supporting these capabilities has been designed and implemented. Using this framework, individual modules such as a position estimator, a path tracker, a mapper, network serves and other crucial elements have been successfully integrated into a controller of the Navlab autonomous vehicle. Issues regarding the use of an Inertial Navigation System for position estimation are discussed and the positioning data is compared to dead reckoning techniques. Three path tracking strategies: pure pursuit, quintic polynomial fit, and control theory approach are analyzed and compared. In addition the operation of a smooth velocity for the Navlab is discussed. The context of the research, the architecture, its implementation, and other performance results from experiments are discussed.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1990
- Accession Number
- ADA230117
Entities
People
- Omead Amidi
Organizations
- Carnegie Mellon University