Hierarchical Hybrid System Design on Berkeley UAV
T. John Koo, David Hyunchul Shim, Omid Shakernia, Bruno Sinopoli, Yuqi Ma, Frank Hoffmann, S. S. Sastry · 1998
This paper describes recent results on system design and implementation of Berkeley UAV. The system design deploys the architecture of a Flight Vehicle Management System, FVMS, which combines planning and control. The resulting hierarchical control strategy which involves the interaction of continuous dynamics and discrete events is a hybrid system. Three controller based on different control methodologies are designed for various types of man oeuvres and flight modes, and their performance are evaluated under simulation based on a nonlinear model. The FVMS interacts with a vision system which is responsible for detection and recognition of different types of hazardous waste barrels. The vision algorithm consists of three parts: filtering, segmentation, and recognition. A 3D virtual environment simulation , SmartAerobots, is developed as a visualization tool. A helicopter-based aerial vehicle has been constructed and the proposed algorithms are being implemented and verified. 1 Introdu...