SBIR-STTR Award

In this research effort, GTD Unlimited will design tools for efficiently specifying, representing, and analyzing the interactions between control systems in System of Systems (SoS). Our approach will be to model the problem as an uncertain system and analyze the resulting models using control methods. Control methods can be related back to the standard Nyquist criterion for stability, ensuring the overall system stability. The central benefit of this approach is that the outputs passed back and forth between interconnected systems/simulations can be modeled as an uncertain system. The range of expected values for the parameters passed between the coupled systems can be specified along with ranges for possibly destabilizing intersystem gains, time-delays, and lags to any Hardware in the Loop (HwiL) systems. Control methods allow for the assessment of stability margins, performance margins, and conditions that will cause a system to destabilize. Approved for Public Release 16-MDA-8620 (1 April 16)

In this research effort, the GTD team will design tools for efficiently specifying, representing, and analyzing the interactions between control systems in System of Systems (SoS). Three approaches will be considered and integrated. The first two are based in the rigorous application of H_8 control theory to the SoS design problem. The third will use complexity measures to generate a metric that can be used to predict the probability of performance problems due to controller interactions or to compare designs based on the probability of interactions. The first approach models the problem as an uncertain system and bounds the stability regions of the system using H_8 and active agent control methods. H_8 methods allow for the assessment of stability margins, performance margins, and conditions that will cause a system to destabilize. The second approach will be to use the concepts of a generalized frequency variable applied to a system of systems described as a network. Previous application of this work has focused on control of active agents. The SoS can be modelled as a network of subsystems that behave as active agents. This approach provides a framework for the rigorous analysis of stability, observability, controllability and reachability. Approved for Public Release | 17-MDA-9219 (31 May 17)