Efficiency Improvement Method Based on Comprehensive Loss Regression Model for Doubly Salient Electromagnetic Generation System

The doubly salient electromagnetic generator (DSEG) is a competitive candidate for more electric aircraft applications due to its high reliability. This article proposes a comprehensive loss management method based on current regulations with angular position control (APC) for DSEG systems. The loss components are divided and analyzed to accurately determine the best-conducting angle in the APC method for minimized total loss. An efficiency regression model based on random forest (RF) is also proposed to address the uncertainty of loss distribution and the coupling of multiple losses. The Bayesian optimization (BO) algorithm is used to optimize the RF parameters and improve its prediction accuracy. The efficiency optimization method is proposed based on the BO-RF efficiency model. The loss distribution is investigated based on analysis and experiments. A control method is proposed to reduce the overall losses and achieve optimal efficiency for the DSEG system by regulating the winding currents, which effectively predicts the optimal conducting angle for maximum efficiency in wide operation ranges.