AFOSR Grant FA 9550 - 05 - 1 - 0401 “ Control and Information Architecture for Coordinated Networked Systems ” Principal Investigator :

A unifying framework for the design and operation of networked coordinated systems has been researched with developments focusing on application in the management of autonomous vehicles which are subject to large external disturbances, such as wind gusts. The benchmark problem is the use of intervehicle communication and active control to effect collision avoidance in multi-vehicle systems with significant environmental disturbances. The central aim of the work was to combine constrained optimal control, achieved via so-called Model Predictive Control methods, with limited capacity communication link resource assignment to achieve collision avoidance with a nominal fleet formation and a specified level of external disturbances. The core results concern the use of the covariance or quantified uncertainty of the estimate of the other vehicles’ positions as the link between communications resource assignment – more bits of communications means more accurate position estimates – and collision avoidance requirements – close vehicles in the formation require more accurate position information to avoid collision. A computational tool is derived. The work has been presented at AFRL.