Data-Driven Optimization of Electromagnetic–Mechanical Coupling of Permanent Magnet Topology for Vibration Suppression in PMDC Motors

This study proposes a vibration reduction strategy for a 12-slot, two-pole permanent magnet brushed DC (PMDC) motor used in automotive blower systems. A multi-parameter optimization framework combining finite element analysis and experimental validation is developed to address cogging torque, a critical source of electromagnetic vibration and acoustic noise. The influence of pole arc coefficient and permanent magnet eccentricity on cogging torque is systematically investigated using response surface methodology, identifying an optimal design with significantly reduced torque ripple and vibration. Furthermore, a machine learning model based on the random forest algorithm is introduced to predict cogging torque, air gap magnetic flux density, and output torque, achieving high accuracy and strong generalizability. The results confirm that the optimized motor structure suppresses resonance-induced noise near 7500 Hz, improving overall motor stability and acoustic performance. The proposed data-driven design approach offers a reliable and efficient pathway for vibration optimization in low-cost automotive PMDC motors.