Trajektorieföljning med garantier på prestanda och säkerhet.

Important results from the project

The project goal is the development of vehicle motion control algorithms, which come with safety guarantees. Such algorithms can be, for example, deployed within an autonomous driving system to make sure that the vehicle follows an intended path, with a guaranteed maximum lateral deviation. Such guarantee must hold in presence of non nominal operating conditions like, e.g., a deflated tire, variable vehicle load, and/or road banking.

Expected long term effects

The project was meant to produce a streamlined procedure to guide a control engineer in the design and tuning of motion control algorithms. Such goal has been achieved to a large extent. A Matlab toolbox is available at Zenuity, which, starting from data find a set of controllers, along with a number of control invariant sets, where the controllers are guaranteed to satisfy a set of constraints, ranging from comfort (low priority) to safety (high priority) constraints.

Approach and implementation

The project has been developed along an academic and industrial tracks working on the same problem from two different perspectives. The academic track has developed new theoretical results, which have aimed at contributed to the MPC theory for control problems where performance guarantees are of concern. The industrial track has more focused on an approach that could readily result in a toolbox to be used within the company.