Eggert H., Guth H., Jakob W., Meinzer S., Sieber I., Süß W.
Institut fur Angewandte Informatik, DE
Keywords: analytical model, desgn optimization, design variants, heuristic search technique
At systems level, however, the extreme complexity of FEM models (many network nodes, long computer times) requires the use of models determined analytically. As the complexity of the search space of such analytical models becomes very high already if only a few sizes of a microsystem are treated, manually controlled simulations for only a few design variants, as a rule, will not result in optimum systems designs. Part-automated design optimiza on can be achieved by replacing human operators by a tool which studies the parameter space of the systems parameters. Human activities, in this case. are reduced to predefining an evaluation (a description of the quality goals and priorities). The tool then directs the search into that part of the parameter space in which optimum design variants can be found. Part-automated design optimization does not depend on a type of model, as only values of formal parameters are exchanged between th mulator and the optimization tool. The convergence reli ity of a traditional numerical method is compared with a heuristic search technique in an example of design optimization.
Journal: TechConnect Briefs
Volume: Technical Proceedings of the 1998 International Conference on Modeling and Simulation of Microsystems
Published: April 6, 1998
Pages: 344 - 349
Industry sector: Sensors, MEMS, Electronics
Topic: Modeling & Simulation of Microsystems
ISBN: 0-96661-35-0-3