Distributed Control for Complex Mission Scenarios With Non-Holonomic Agents - An LPV Approach

This thesis considers cooperative control of non-holonomic autonomous agents which are modeled as linear parameter-varying (LPV) systems. Two approaches are considered. First, the formation control approach; here a coupled information flow filter is presented that enables feedback of local agent data into the consensus protocol. Moreover, a simultaneous design and optimization of filter and local controller is proposed. Secondly, extension of flocking control to LPV agents. A separation in design is achieved by designing a flocking filter that receives data from and provides reference trajectories to local agents. This approach is illustrated with source seeking and level curve tracking scenarios.