Decentralized Architecture for Generic Real-Time Distributed Engine Control Modeling and Simulation for Hardware-in-the-Loop (HIL), Model-Based control and PHM Research Development

摘要

This paper presents a brief overview as well as some preliminary research results on the advantages of Decentralized architecture in the Distributed Engine Control and Simulation studies. A generic turbine engine control and simulation model developed by the Air Force Research Laboratory is used in the Distributed Engine Control and Simulation Laboratory facility at The Ohio State University to demonstrate the relative merits of a decentralized partially distributed architecture with a data concentrator for robustness against data loss and time delays. In addition to the standard PID control law, new control laws based on Model Predictive Control are implemented in a decentralized framework and the resulting performance is evaluated in the simulator facility. This type of research is deemed highly beneficial to turbine engine control industry and FADEC manufacturers. In addition, incorporation of engine health management via diagnostics and prognostics into the control framework is addressed. Considerable insight into the issue of integration of control and PHM is provided. It is shown that Decentralized architecture is beneficial on various counts such as stability robustness to parameter perturbations, engine health management integration, and compensation against communication time delays and data loss.