Embedded Systems for Feedback Mixing Control in Fluid Flow

Northeastern University, Boston MA

Investigators

Abstract

Mixing in turbulent flows play an essential role in diverse engineering applications. Beneficial non-equilibrium dynamics impact on engine efficiency, durability, stability, and emissions, on reduced drag in aircrafts and in naval vessels. The unique focus of this research is on feedback regulation of active mixing control. Technical challenges are an outgrowth of the complexity and sensitivity of fluid dynamic systems, and require an integrated approach to modeling, information technology, control theory, algorithm development, and software and hardware implementation. Modeling will focus on dominant coherent structures, periodicity, and concepts of Hamiltonian energy. Those provide also possibilities for efficient parallel implementation on an embedded system. Intended models will range from lumped point vortex models, to purely phenomenologically derived I/O correlations of main structures. Hybrid models arise naturally as a design framework, subject to (nearly) periodic vortex generation and elimination. On a finer time scale, hybrid models will be used as a means to parallelize spatially distributed control and observer implementation. This project includes collaborations with industry, government and academe. If successful, this high risk-high yield undertaking will include both technology transition in critical applications and developments of integrated embedded systems and control design methods for interesting classes of complex, nonlinear systems.