Adaptive hierarchical multi-objective fuzzy optimization for circuit design

The outline of a methodology for adaptive hierarchical multiobjective function optimization in a fuzzy sense is presented, suitable for circuit optimization and able to mimic to some extent the designer's interactive circuit design process. It is an effort to generalize and go beyond the Taguchi methodology of variability-then-target optimization with the addition of automated selection of optimization variables. The methodology allows for the simultaneous and/or sequential optimization of an arbitrary number of statistical measures of an arbitrary number of circuit performances.<<ETX>>