Designing and optimizing axial flux motor for electric ships

This paper proposes a hybrid sensitivity analysis and response surface optimization method for the design of an axial flux permanent magnet (AFPM) motor with a yokeless and segmented armature for electric ships’ applications. The initial design parameters are determined using formulas and design experience, and then refined through iterative finite element simulations. Ansys Maxwell is employed for the construction of a finite element model with suitable meshing, with the objective of validating the magnetic circuit design and performance indicators. Taguchi orthogonal experiments are conducted to calculate sensitivity indices, distinguishing between strong- and weak-sensitive parameters. The BBD method is employed for experiments on strong-sensitive parameters, while the NSGA-II algorithm is utilized to identify Pareto optimal solutions. Finite element simulations are conducted to compare motor performance before and after optimization, thereby confirming the effectiveness of multi-objective optimization.