Vibration noise reduction in high-speed dual-structure six-phase switched reluctance motor via magnetic circuit analysis and genetic algorithm-optimized back propagation for slot design

The high-speed dual-structure six-phase switched reluctance motor (HDSSRM) exhibits pronounced vibration and acoustic noise, arising from inherent characteristics of the switched reluctance motor (SRM) and the non-uniform circumferential arrangement of its dual-structure six-phase configuration. This paper first analyzes the magnetic flux circuits to compare the effects of long and short magnetic circuits on radial force and operating conditions, and then mitigates the radial force associated with the short magnetic circuit. Next, the Maxwell stress tensor method is employed to examine how modifications to the rotor tooth geometry and the introduction of slots at the stator tooth tips reduce the radial magnetic pull force, and the slot parameters are optimized using a genetic algorithm–optimized back-propagation (GA-BP) neural network. Finally, experiments validate the proposed noise-reduction strategy, and finite-element analysis is used to compare the motor’s torque and radial-force characteristics.