Towards automatic level selection for design-aimed simulation

We describe methods by which a simulation environment can automatically select the appropriate level of detail for each model that composes a system definition. Objectives of the automatic selection process are to ensure that the accuracy of the computed system response is compatible with the corresponding stage of an iterative design process, and that the simulation is completed with the least possible computational effort. Three methods were analyzed, tested, and compared: a frequency response method, a knowledge-based method and a step response method. The three methods are illustrated by a case study of a transformer-fed induction motor.

[1]  Madhav Erraguntla,et al.  Simulation modeling at multiple levels of abstraction , 1998, 1998 Winter Simulation Conference. Proceedings (Cat. No.98CH36274).

[2]  Theodore Wildi,et al.  Electrical Machines, Drives, and Power Systems , 1990 .

[3]  E. Santi,et al.  Multilevel device models developed for the virtual test bed (VTB) , 2004, Conference Record of the 2004 IEEE Industry Applications Conference, 2004. 39th IAS Annual Meeting..

[4]  Torbjörn Thiringer,et al.  Comparison of reduced-order dynamic models of induction machines , 2001 .

[5]  Christiaan J. J. Paredis,et al.  Modeling and Simulation Methods for Design of Engineering Systems , 2001, J. Comput. Inf. Sci. Eng..

[6]  Ned Mohan,et al.  Advanced Electric Drives: Analysis, Control and Modeling Using Simulink , 2001 .