Flexible design of reluctance synchronous machines using constrained genetic algorithm optimization

The output performance of a reluctance synchronous machine strongly depends on a well-designed machine draft. Since there are many different requirements concerning the output capability, the machine design has to be done considering all specifications and demands also determined by the field of application. However the machine design and optimization by genetic algorithm is a simple and comparatively fast method to achieve adequate results and is therefore presented within this publication. The outcome of the optimization of course depends on the algorithm provided design variables as well as the calculation model concealed behind. In this case an analytical model based on a simplified magnetic circuit analysis is used for fitness value computation. The calculation method provides good results regarding some major parameters like d- and q-axis inductances as well as d- and q-axis currents. Within the genetic algorithm optimization a flexible design of reluctance synchronous machines in terms of the specific requirements is easily accessible. As a result important conclusions regarding the stator and rotor geometry can be achieved.