High-frequency vibration and noise induced by pulsewidth modulation (PWM) deteriorate the acoustic performance of permanent magnet synchronous motor (PMSM) drives. Periodic-carrier-frequency PWM (PCFPWM) is an effective method that can disperse the high-frequency harmonics to a wider frequency band, thereby suppressing the peak harmonic. This article presents a detailed study on how exactly PCFPWM affects both the low-frequency and high-frequency vibrations of PMSMs. It is revealed that PCFPWM actually deteriorates low-frequency vibrations of the motor, due to unwanted side effect of additional ac components in d- and q-axis currents and consequent increased torque pulsations. But, this can be remedied by quasi-proportional resonant control of corresponding current harmonics. For high-frequency vibrations, the article thoroughly studies the effects of PCFPWM on current harmonics and the zero-order electromagnetic (EM) force acting on motor stator teeth. Compared with the experimental vibration results, it is found that high-frequency zero-order EM force is the direct and major cause of EM vibrations induced by PCFPWM. This article provides the theoretical basis for future optimization of PCFPWM based on the zero-order EM force as a direct vibration metric.
31 January 2023 in IEEE Transactions on Industrial Electronics
This article introduces a spatial-harmonic model for analyzing high-frequency vibrations induced by PWM-current harmonics in inverter-fed surface permanent magnet synchronous motors (PMSMs). The model provides a fresh perspective on calculating the high-frequency vibrations of motors with a fixed inverter switching frequency. Electromagnetic (EM) forces along stator boundaries are first calculated utilizing the permeance distribution function. Then, the concept of spatial-order-excitation transfer functions is introduced to facilitate the modelling of the motor vibrational response under distributed excitation forces. Two approaches to obtain spatial-order-excitation transfer functions (SOETFs) are subsequently investigated, one by finite-element analysis while the other by experimental identification. Finally, high-frequency vibrations predicted by the proposed model are validated by experiments, showing a high degree of accuracy.
8 March 2021 in IEEE Transactions on Power Electronics
Switching-frequency vibration caused by pulsewidth modulation (PWM) is a critical shortcoming of permanent magnet synchronous motor (PMSM) drives in certain applications. This article reveals the mechanism for suppressing high-frequency vibration of multiphase surface permanent magnet (SPM) motors via PWM carrier phase shifting (CPS). The article first introduces a 12-phase SPM drive with three different modes of CPS. Then, it presents an analytical method to calculate high-frequency current harmonics as well as flux and force distributions along stator boundaries. By analyzing the calculated current harmonics and flux distributions under different CPS modes and parameters, the article discovers that CPS greatly alters the spatial distribution of high-frequency stator electromagnetic forces and, thus, significantly changes their zero-order components, which are the major contributors to high-frequency motor vibrations. Finally, the optimal CPS mode and parameter are determined analytically and are validated by experimental results.
31 May 2023 in IEEE Transactions on Power Electronics
High-frequency vibration caused by pulse width modulation (PWM) is a critical problem of permanent magnet synchronous motor (PMSM) drives for low-noise applications. Periodic carrier frequency PWM (PCFPWM) can effectively reduce high-frequency vibrations by dispersing concentrated harmonics over a wider frequency band. However, the resulting vibration spectra are normally far from ideal and still contain several local peaks because the uneven vibration frequency response of the motor is not taken into account. To address this problem, this paper proposes an optimal PCFPWM scheme that minimizes vibration spectral peaks of the motor by considering its vibration frequency response function (VFRF). A time series of PWM carrier frequency is carefully devised such that the amplitude of the induced electromagnetic (EM) force is inversely related to the VFRF of the motor, thereby shaving peaks in the vibration spectra. Finally, the optimal PCFPWM scheme is experimentally verified and demonstrates improved vibration suppressing performance.
18 December 2019 in IEEE Transactions on Power Electronics
This paper introduces a new method to analyze electromagnetic forces induced by PWM current harmonics in inverter-fed surface permanent-magnet motors (SPMs). The concept of the permeance distribution function (PDF) is first introduced with its properties and relation to machine inductances fully explained. Utilizing the concept of PDFs, two methods are proposed to calculate the flux distribution on stator boundaries by armature currents and rotor permanent magnets, respectively. By superposing the stator flux distributions from both sources, the final stator electromagnetic force distribution is obtained. The method is validated by finite-element analysis, demonstrating good accuracy and very fast computation time
5 October 2021 in IEEE Transactions on Transportation Electrification
This article presents a simple and nonintrusive method to identify and compensate for current measurement errors in interior permanent magnet synchronous motor (IPMSM) drives. Each IPMSM has a distinct steady-state short-circuit current characteristic. By controlling the motor to operate at its steady-state short-circuit current during coasting down, the inverter can be used to smoothly switch the IPMSM into a three-phase short-circuit state. Then, the $d - q$ axis measured currents are compared with the true short-circuit characteristic of the IPMSM. The distortion pattern in the $d - q$ axis currents contains all the needed information to correct for current sensor errors in IPMSM drives. For a three-phase motor, the proposed method can correct for single-current-sensor error or dual-current-sensor error with good accuracy. The effectiveness of the method is validated by both simulation and experiment.