Nonlinear Trajectory Generation for the Caltech Multi-Vehicle Wireless Testbed

Jonathan Chauvin, Laure Sinegre, Richard M. Murray
2003 European Control Conference

The Caltech Multi Vehicles Wireless Testbed (MVWT) is a platform designed to explore theoritical advances in multi-vehicle coordination and control, networked control systems and high con�dence distributed computation. The contribution of this report is to present simulation and experimental results on the generation and implementation of optimal trajectories for the MVWT vehicles. The vehicles are nonlinear and spatially contrained with bounded input control. The trajectories are generated using the NTG software package developed at Caltech. Minimum time trajectories and the application of Model Predictive Control (MPC) are investigated. can actually follow i.e. trajectories that satisfy every constraint of the testbed. Those constraints can either be linear, like the boundaries of the testbed or nonlinear like the constraints on the input. The main di�erence and also di�culty in our case is that the system is not linearly controllable around its equilibrium. In Section 2 we will give a quick description of the systems properties and in Section 3 and 4 we will describe the progression which led us from the optimization problem to the implementation on the real vehicles. In Section 5 other optimization problems such as minimum time trajectory generation and model predictive control are investigated.

• Conference

 Paper: http://www.cds.caltech.edu/~murray/preprints/csm03-ecc.pdf

• Project(s): Template:HTDB funding::DARPA/SEC