Reluctance Network Modeling and Optimization of Brushless Doubly-Fed Machines for Wind Power Generation

Brushless doubly-fed reluctance machine (BDFRM) and brushless doubly-fed induction machine (BDFIM) have great potential to become third-generation electrical machines for wind turbine applications instead of doubly-fed induction machine (DFIM). However, due to their complicated operating principle, they have not yet been commercialized. In the idea of reducing the materials cost, increasing the performances of these machines, and allowing their commercial breakthrough, it is important to propose a new approach to the design and optimization of these kinds of machines. This paper contributes by combining the reluctance network (RN) model developed for those machines together with the particle swarm optimization (PSO) algorithm. Previously, a new RN model has been developed and validated by finite elements (FE) analysis for BDFRM modeling. This model simultaneously takes into account magnetic saturation, rotor motion, and stator winding distribution. It also offers an interesting trade-off between precision and computational time compared to the FE model. Therefore, the developed model is firstly extended to the BDFIM structure and then coupled to the PSO algorithm and the result is an accurate and fast BDFRM and BDFIM design optimization tool. This tool is then used to generate a 1kW machine design with optimized performances.