Current students

One of major challenges for the control of an autonomous vehicles is related to its lateral dynamics. Often, traditional approaches to lateral vehicle control are based on models of the vehicle dynamics that consider the steering wheel angle to be the control input, while neglecting possible delays and steering actuator dynamics. More recent approaches rely on computationally intensive and complex non-linear modelling and control strategies. The starting point of this work is defining a feedback linearization law that transforms a non-linear single-track vehicle model into a linear system using a non-linear feedback. This enables a flexible control design phase as the linear analysis and design tools can be easily applied to such a system. The feedback linearization law will then be augmented to consider also the steering actuator dynamics and delays as they are not negligible and often the source of instability in the system. The framework will mainly be composed of three layers: linearization, robustification and trajectory tracking. The design of each layer will aim to attain desired performance level within a operating limit while being computationally efficient and simple.