R-FL-C Model for Design Optimization of PM Generators

A novel recursive fuzzy-logic classifying (R-FL-C) approach for the design optimization of permanent magnet generators is presented. This approach is used to cut the search space and to avoid local minima during the optimization process. The search space database is established using knowledge obtained off-line from finite-element state-space models. The application of the approach to a prototype aerospace type PM generator resulted in a toothless stator design with lower weight and improved performance. The results are validated by comparison to test data