Optimal design of pma-synrm for e-bike traction motor using adaptive interpolation robustness optimization algorithm

This paper proposes an adaptive interpolation robustness optimization algorithm (AIROA) for the robust optimal design of a permanent magnet assisted synchronous reluctance motor for electric bicycle (EB) traction. The AIROA can drastically reduce the number of function calls to converge to an optimal solution by interpolating the problem regions. Especially, the AIROA effectively adjusts the number of generated samples using the diversification strategy with novel initial sampling strategy. Moreover, in the intensive searching strategy, the calculation time for generating a surrogate model and the accuracy of the solutions of the surrogate model can be enhanced as the AIROA adjusts the resolution of the interpolation near the roughly found solutions. At the end of the algorithm, a robustness test strategy is conducted, and a robust optimal solution is determined considering the accuracy of the interpolated uncertainty band of both global and local solutions. The superior performance of the AIROA was verified by two mathematical test functions. The applicability of the practical motor design was verified by applying the AIROA to the optimal design of the EB traction motor and successfully deriving the robust optimum design. The prototype motor was manufactured, and further validation was conducted by comparing the experimental results.