Metamodeling of Rotating Electric Machinery

摘要

This work sets forth an approach whereby a system designer can rapidly estimate the size versus loss characteristics of a class of electric machinery, and the impact of this tradeoff at the system level. In essence, the methodology provides the first physics-based multiobjective scaling law for a Pareto-optimal front capturing the tradeoff between size and loss as a function of speed and requirements (output power or torque). It is the first time such a representation has been achieved. This is accomplished through the metamodeling of a normalized optimization based machine design framework. The method is applied to a permanent magnet ac machine.