Scalable Optimization for Learning in Control

PI: Anders Hansson (LiU), co-PI:Anders Rantzer (LU); with Johan Löfberg (LiU), Richard Pates (LU)

Large-scale engineering applications put new demands on control theory, as most existing methods for analysis, design and verification do not scale well with increasing complexity. Furthermore, new powerful algorithms for machine learning are increasingly being used for control engineering purposes, further adding to the complexity of analysis and verification.  To counteract this, there is a strong demand for scalable optimization methods and corresponding information interfaces. Important applications areas are autonomous transportation, manufacturing and robotics.   The purpose of the proposed project is to address the complexity challenges by developing and exploiting new optimization algorithms suitable for parallel and/or distributed implementation.

Project number: B13