Design of a Permanent Magnet Synchronous Generator using Interactive Multiobjective Optimization

Abstract

We consider an analytical model of a permanent magnet synchronous generator and formulate a mixed-integer constrained multiobjective optimization problem with six objective functions. We demonstrate the usefulness of solving such a problem by applying an interactive multiobjective optimization method called NIMBUS. In the NIMBUS method, a decision is iteratively involved in the optimization process and directs the solution process in order to find her/his most preferred Pareto optimal solution for the problem. We also employ a commonly used noninteractive evolutionary multiobjective optimization method NSGA-II to generate a set of solutions that approximates the Pareto set and demonstrate the advantages of using an interactive method. This study is the first one to consider an interactive approach for the design of a permanent magnet synchronous generator. Thus, we promote the further usage of interactive multiobjective optimization methods in the design. Further, we see that these methods could also be very useful in the teaching of electrical machines.