Multi-objective optimization of linear generator based on global sensitivity analysis for enhanced linear range extender performance

This paper proposes a comprehensive optimization method for improving the performance of linear range extenders by enhancing the linear generator (LG). The method combines global sensitivity analysis, neural network, response surface method, and multi-objective optimization algorithm. The LG's structural parameters, parameter constraints, and optimization objectives are identified. A global sensitivity analysis is conducted using Sobol's method to evaluate the comprehensive effects of structural parameters on output performance. For different sensitivity parameters, corresponding prediction and optimization models are developed, and multi-objective optimization is performed. Results show that the induced electromotive force of the optimized LG increases significantly, and its waveform is closer to the ideal sinusoidal curve. The harmonic distortion rate reduces from 18.07 % to 8.67 %. The output power and volumetric power density are greatly increased. Although the copper and core losses increase, the power generation efficiency still increases. Furthermore, the detent force and motor thrust fluctuations are significantly reduced. The peak-to-peak detent force decreases from 163 N to 62 N, and the thrust fluctuation drops from 19.01 % to 9.13 %. Meanwhile, the motor thrust increases by approximately 59 %. These results demonstrate significant improvements in output performance, load capacity, and operational stability of the LG, validating the effectiveness of the proposed optimization method.