Rotor Position Estimation Strategy Using Artificial Neural Network for a Novel Design Transverse Flux Machine

Türker, Çiğdem Gündoğan; Kuyumcu, Feriha Erfan

doi:10.5370/jeet.2015.10.5.2009

Cited by 1 publication

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…9, whereT sq andT sin are the torque ripple rates resulted from square wave current and sinusoidal current respectively, and T a to T e represent the output torques of A-phase to E-phase. In this study, the torque ripple rate is calculated according to (3).…”

Section: Control By Two Typical Currentsmentioning

confidence: 99%

Torque Ripple Suppression of Building-Block Transverse Flux Permanent Magnet Motor by Current Compensation and Variable Parameter Control Based on Real-Time Inductance

2020

View full text Add to dashboard Cite

Building-block transverse flux permanent magnet motor (B-TFPMM) has the advantage of electromagnetic decoupling and high torque density, but it has the strong nonlinearity and the large torque ripple. To solve the problem, the nonlinear characteristics of B-TFPMM stator back electromotive force (EMF) e pm , winding inductance L and single-phase torque T sp are analyzed firstly, and the nonlinear dynamic models involving e pm , L, T sp , stator current and rotor position are established by three-dimensional finite element method (3DFEM). Secondly, apply the square wave current and sinusoidal current to stator windings respectively, and the result shows that using sinusoidal current would acquire smaller torque ripple rate when the load torque is more than 30 N • m. Furthermore, the variation rule of five-phase current utilization is obtained and the maximum current utilization model is established. Based on these, the current compensation method is developed and then the torque ripple rate could be reduced by 10% approximately. After that, combining with 7 nonlinear dynamic models of B-TFPMM, a closed-loop control system based on the compensation method and variable parameter PID algorithm is built. The ratio of the real-time winding inductance to sampling period is chosen as the proportional coefficient of controller. The results indicate that compared with the control mode which combining uncompensated current and traditional constant parameter PID algorithm, the control mode of using compensated current and variable parameter PID algorithm would reduce the torque ripple rate by 57% to 65% and increase the average torque by 13% to 33% in the meantime. INDEX TERMS Building-block transverse flux permanent magnet motor, torque ripple suppression, current compensation, nonlinear dynamic model, variable parameter control.

show abstract

Section: Control By Two Typical Currentsmentioning

confidence: 99%