Model predictive traction and steering control of planetary rovers

摘要

Results of the ESA project RobMPC (Robust Model Predictive Control for Space Constraint Systems) could successfully demonstrate that model predictive control (MPC) is definitively applicable for space systems with high dynamics like wheeled vehicles exploring a planetary surface. In the context of RobMPC a rover control hierarchy for guidance, trajectory control as well as traction and steering control was implemented. Controller verifications and robustness tests were performed using a functional engineering simulator (FES) including a multi-body dynamics model of ESA’s EGP rover and the vehicle-terrain contact physics. The latest validation step is the MPC implementation on a real-time computer system controlling the ExoMars breadboard rover at DLR’s planetary exploration lab.