AASS Seminar

Who: Dr. Joshua Marshall, Queen's University, Canada
What: Seminar: "Learning-Enhanced Controller Designs for Off-Road Robotic Vehicles"
When: April 3, 2025, starting at 13:00
Where: T101, Teknikhuset. 
Remote attendance: https://oru-se.zoom.us/j/65096623293

Abstract: Control for off-road autonomous driving can be challenging given the uncertainty of encountered terrains and difficulties in accurately modelling vehicle dynamics. This talk introduces two approaches to handling these difficulties by enhancing model-based controller designs with learned elements.  The first involves the application of iterative learning control (ILC) on a load-haul-dump machine (LHD) in underground conditions, which was work that began almost a decade ago and was conducted in partnership with Epiroc AB.  The second is more recent and involves a high-performance path following algorithm that combines Gaussian processes (GP) based learning and feedback linearization (FBL) with model predictive control (MPC) for ground mobile robots operating in off-road terrains, referred to as GP-FBLMPC. The algorithm uses a nominal kinematic model and learns unmodeled dynamics as GP models by using observation data collected during field experiments. Outdoor experiments using a Clearpath Husky A200 mobile robot show that the proposed GP-FBLMPC algorithm's performance is comparable to existing GP learning-based nonlinear MPC (GP-NMPC) methods with respect to the path following errors. The advantage of GP-FBLMPC is that it is generalizable in reducing path following errors for different paths that are not included in the GP models training process, while GP-NMPC methods only work well on exactly the same path on which GP models are trained. GP-FBLMPC is also computationally more efficient than the GP-NMPC because it does not conduct iterative optimization and requires fewer GP models to make predictions over the MPC prediction horizon loop at every time step.

More about Dr. Marshall and his research