Electric machine design by fuzzy fitness
Esa Koskimäki, J. A. C. Goos · 1996
. Electric machine designer often utilizes calculation libraries that consist of different calculation models. Dimensioning the machine means setting the various input parameters to the calculation models and checking the technical requirements of the result. Genetic algorithm is proposed as a substitute for the designer's heuristic optimizing process. The role of the designer is to choose the parameters optimized within tolerances. For evaluating the dimensioned machine fuzzy logic is used to a part of the fitness function. The tolerances are determined by fuzzy membership functions. Also, the importance of the parameters optimized can be assigned. A fuzzy fitness function is calculated as the weighted sum of the grade of membership functions and importance values. The goal is to automatize the dimensioning process at a rough accuracy level, and integrate that as part of a synchronous machine offer calculation system. The system solves a traditional configuration problem where all th...