Distributed constraint fulfillment in cooperative control

Staff Mentor:

dr.ir. T. Keviczky (Tamas)


Distributed control; Multi-agent systems; Distributed and large-scale systems; Optimal and model predictive control


Decentralized control techniques today can be found in a broad spectrum of applications ranging from robotics and formation flight to civil engineering. Their importance for dynamically decoupled systems arises from the abundance of networks of independently actuated systems and the necessity of avoiding centralized design when this becomes computationally prohibitive or would require unrealistic expectations regarding information exchange.

Most real-life cooperative control problems involve coupling constraints between subsystems. However, if coupling constraints are present, ensuring feasibility in a distributed control scheme without introducing excessively conservative assumptions is a challenging problem. The conservativeness introduced by e.g. a pure robust constraint fulfillment procedure in predictive control or other worst-case approaches suggest that new methods are needed, which rely on the cooperation between neighboring subsystems or exploit other degrees of freedom in the problem formulation, such as the interconnection graph structure.

The aim of this MSc project is to design feasible cooperative distributed control laws for large-scale systems composed of dynamically decoupled subsystems with coupling constraints (e.g. collision avoidance in satellite formation flight). The MSc project will consist of carrying out a literature survey on controlled invariant sets, reachability and various approximation schemes, designing control laws for negotiation-based robust constraint fulfillment, defining case studies using Matlab/Simulink and implementing them in a simulation environment.

© Copyright Delft Center for Systems and Control, Delft University of Technology, 2017.