Hybrid GA-PSO multi-objective design optimization of coupled PM synchronous motor-drive using physics-based modeling approach

The torque ripple of a PM synchronous motor (PMSM) originates from two sources; the cogging torque and mismatch between the back EMF and phase current waveforms. The first source is only related to the magnetic design parameters of PMSM and the second source is related to the effect of the driving circuit on the magnetic design parameters. Therefore, achieving a minimum torque ripple requires a hybrid magnetic design of PMSM dynamically when it is connected to the driving circuit. From the motor design point of view, a minimum torque ripple can be achieved with a minimum cogging torque and a proper back emf waveform. This sometimes causes an undesirable phase current of machine which in turn increase the total harmonic distortion and the RMS value of the phase current. As a results, this leads to a lower motor efficiency. Although a desirable trade-off between performance measure of the machine and magnetic design parameter can help achieve a more efficient design This paper deals with an optimal design of PMSM motor geometry to achieve minimum torque ripple, minimum RMS value of phase current, and minimum total harmonic distortion of phase currents simultaneously. A multi-objective function is formed as a combination of these parameters. A physics-based phase variable model is used to couple the motor to the driving circuit. The physical behavior of PMSM was calculated by a non-linear transient FE analysis with motion. A mixed Genetic-particle swarm algorithm is developed and used as an optimization procedure. The results before and after optimization show the expected performance improvements while reducing magnet material and copper size.