Application Of Controller Design Methods Research Articles

Dynamic Model of a Multi-Evaporator Organic Rankine Cycle for Exhaust Heat Recovery in Automotive Applications

Abstract: This paper presents a dynamic model of an organic Rankine cycle (ORC) for exhaust heat recovery (EHR) in automotive applications. EHR converts the thermal energy of the exhaust gas into mechanical power. Dynamic components like evaporator, condenser, etc. are modeled based on physical conservation laws given as partial differential equations (PDE). The system involves two parallel evaporators coupled via an open t-piece which is handled numerically efficiently by introduction of algebraic constraints. An integral method improves the evaluation of the fluid properties which partly contain discontinuities. They can have negative effects on accuracy and simulation performance if evaluated inappropriately. The model is implemented by signal-based programming suitable for simulation in Matlab/Simulink. This enables convenient software-in-the-loop (SIL) engineering since many control unit producers provide rapid prototype programming via Matlab/Simulink. A simplified model is proposed for the increase of robustness and simulation performance or for application of controller design methods. Simulation shows both the entire ORC system compared measurements from a Daimler prototype truck, and comparison between complex and simplified model.

Design of periodic output feedback and fast output sampling based controllers for systems with slow and fast modes

AbstractIt is well known that the straight forward application of controller design methods to two‐time‐scale systems is susceptible to numerical ill‐conditioning and stiffness problems. However, it has been demonstrated in this paper that the presence of time‐scales can be exploited for achieving better computational efficiency in case of the design of controllers based on periodic output feedback and fast output sampling principles. In particular, an approach for design of periodic output feedback and fast output sampling controllers for a two‐time‐scale system by decomposition into a slow and a fast subsystem problems, has been presented. These smaller order problems are then separately solved and the results thus obtained for the slow and fast subsystems are combined to obtain the solution of the original design problems. The procedure has been illustrated with the help of a very simplified nuclear reactor model.Copyright © 2011 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society