Optimal Design of Interior PM Synchronous Machines Using Randomly Guided Mesh Adaptive Direct Search Algorithm

Interior PM Synchronous Machines (IPMSM) has characteristic of nonlinear magnetic saturation and many local optimal, have been focused on reducing computation time [1]-[2]. Hence, randomly guided MADS (Mesh Adaptive Direct Search) is newly implemented to reducing the excessive computation time. Optimization capability of the proposed randomly guided MADS is validated through two benchmark functions. In particular, it has been forwarded to optimal design of IPMSM for maximizing Maximum Torque per Ampere (MTPA) based on FEA (Finite Element Analysis).