Search

The longitudinal end effect of permanent magnet linear synchronous machines induces parameter asymmetry, imparting time-varying, and nonlinear characteristics to the machine model. In parameter estimation, parameter asymmetry introduces additional estimated parameters and results in nonlinearity and coupling among the parameters. To enhance the modeling precision and estimation performance, this article proposes a full electrical parameter estimation method considering asymmetry. First, the universal asymmetric equations are formulated for modeling parameter asymmetry. After these equations are subjected to reparameterization, the linear regression formulations with model order reduction are yielded. Second, the dynamic regressor extension and mixing method is employed to achieve full parameter estimation. The quasi-steady state caused by the end effect ensures that the estimation equations incorporate sufficient excitation, thereby meeting the stability conditions without the necessity of external excitation. Finally, the comparative, disturbance, and control application experiments are carried out, which validate the effectiveness and merits of the estimation method.

This paper proposes a novel magnetization state (MS) control method to eliminate the unintentional demagnetization (UD) of low-coercive force permanent magnet (PM) for variable flux memory machine (VFMM). In the proposed method, the amplitude of demagnetization d-axis current pulse is determined from a look-up table with the relationship between the demagnetization d-axis current pulse and the intersections of torque-speed curves under different MSs, and subsequently used as the condition for the MS manipulation. A PM flux linkage observer is adopted to monitor the UD all the time and the amplitude of d-axis current is controlled without exceeding the amplitude of the demand demagnetization d-axis current pulse to avoid the UD in the whole speed range. The inverter nonlinearity is compensated to improve the estimation accuracy of the PM flux linkage. The experiment validation is conducted on a hybrid magnet memory machine system to illustrate the UD issue and to verify the PM flux linkage estimation method and the proposed MS control method.

This paper investigates the interaction effects of permanent magnet (PM) tolerances and rotor eccentricities of 12-slot/10-pole PM machines by finite element method (FEM) and phasor analysis. It is found that the relative location between diversified PMs and eccentric rotor have a significant effect on cogging torque. (a) When a weak (e.g., thin thickness) PM is close to the minimum airgap, rotor static eccentricity and PM tolerances have a strengthening effect to deteriorate cogging torque, but rotor dynamic eccentricity and PM tolerances have a weakening effect to reduce cogging torque. (b) When a strong (e.g., thick thickness) PM is close to the minimum airgap, rotor static/dynamic eccentricity and PM tolerances always have a strengthening effect to deteriorate cogging torque. The experiments verify the FEM-predicted cogging torques and the interaction effects between PM tolerances and rotor eccentricities on cogging torque in the 12-slot/10-pole PM machine.

This paper proposes a simple inductance asymmetry compensation method for minimizing the second-order harmonic current responses and the corresponding position errors in high frequency signal injection sensorless controlled dual three-phase interior permanent magnet synchronous machines (IPMSMs). It is found that the second-order harmonics of HF current responses in two sets of dual three-phase windings due to self-inductance asymmetry have opposite phase angles in the double dq control method. Thus, a simple current summation-based position error compensation method including an online calculation technique of harmonic amplitude ratio is proposed. The proposed compensation method can not only suppress the second-order position error effectively but also maintain a high signal-to-noise ratio of the sensorless control system. It is validated by experiments under different inductance asymmetry conditions.

This paper investigates the ability for elimination of unwanted magneto-motive force (MMF) harmonics in fractional slot dual m-phase single-layer PM machines. By doubling the number of slots and applying stator shifting, the 1st order and one unwanted higher order winding MMF harmonic can be suppressed. Then, through the addition of a second converter operating at a specified phase shift one higher order harmonic can be eliminated. This work identifies the required shift angle of the second winding set as well as converter phase shift in any dual m-phase machine to achieve elimination of an unwanted winding MMF harmonic. This methodology is validated by studying 3- and 5-phase machines through analytical modelling. Then two 3-phase machines, 12-slot/10-pole (12s/10p) and 12s/14p, are investigated using FEA. It is found that the 24s/10p dual 3-phase machine can achieve higher torque than the conventional 3-phase 12s/10p but the 24s/14p dual 3-phase performs worse. However, in both cases the significant suppression of unwanted harmonics reduces the rotor iron losses and PM eddy current loss. As a result, the efficiency of each machine is increased by about 3% over its initial 12-slot counterpart. In particular, the PM eddy current losses are reduced by over 95% leading to a substantial reduction in thermal demagnetization risk. The dual 3-phase single-layer machine is also found to achieve higher torque than two similar topology dual 3-phase double-layer machines while having equivalent PM eddy current losses. Both analytical and numerical results have been validated by a series of tests. A prototype machine has been manufactured to validate EMF, static torque, and efficiency measurements of the proposed machine.

This article proposes a dual 3-phase SPM machine with star-delta windings. The machine is based on the widely adopted 12 s/10 p configuration and employs a number of methods to suppress the two largest core loss causing MMF harmonics. Firstly, stator shifting is used to reduce the 7th order harmonic, which is then completely suppressed through the use of a second 3-phase converter operating at a 15° phase shift with regard to the first one. Secondly, each of the winding sets employs 3-phase star-delta winding that are able to completely cancel the 1st order MMF harmonic. Additionally, these methods increase the amplitude of the torque producing harmonic and so yield a machine with better torque performance than a conventional 3-phase machine. Analytical modelling is used to demonstrate the cumulative impact on winding MMF harmonics. FEA is then used to validate the analytical predictions and to investigate other machine performance. The proposed machine demonstrates comparable average torque and efficiency to a 12-slot dual 3-phase machine, but offers a substantial reduction in torque ripple and PM eddy current losses. This helps improve the machine performance and reduce the risk of thermal demagnetization. A prototype machine has been manufactured and EMF and static torque measurements validate the expected performance of the proposed machine.

In this article, contributions of stator and rotor PMs in dual-permanent magnet (PM) machines with different stator/rotor pole number combinations are identified and investigated based on a multi-torque component decomposition method. The relationship between additive and cancellation effects of unmodulated rotor PM field component, contributions of stator and rotor PMs, and stator/rotor pole number combinations is revealed by analytical derivation and finite element analysis. It is found that the rotor PMs contribute more in torque production than the stator PMs when the rotor pole number is lower than the stator pole number due to the additive effect of unmodulated rotor PM field component, while the stator PMs contribute more in torque production than the rotor PMs when the stator pole number is lower than the rotor pole number due to the cancellation effect of unmodulated rotor PM field component. The dual-PM machines with fewer rotor pole number than stator pole number exhibit better electromagnetic performance since the torque production capability of PMs is improved by utilizing the additive effect of unmodulated rotor PM field component. A guideline for selecting stator/rotor pole number combinations to improve torque density is provided. Two prototypes are manufactured and tested to validate the conclusions.

With the aid of vector analysis and finite element method, this paper investigates the influence of slot and pole number combinations on cogging torque in permanent magnet (PM) machines with interactive effects between tooth bulge and rotor eccentricity, accounting for various slot/pole number combinations (Ns/2p), e.g., 2p = Ns ± m where m is an integer (including 2p = Ns ± 1 and 2p = Ns ± 2), Ns/2p = 12/8, and Ns/2p = 12/4. It is firstly found that m greatly influences the interaction of two tolerances on cogging torque. The degree of interaction is reduced with m, and it is larger in “−m” than in “+ m” machines. Secondly, for the machines having 2p = Ns − m, there is a weakening interaction of cogging torque when the static eccentric rotor is close to outward bulged teeth, whilst there is a strengthening interaction of cogging torque when the static eccentric rotor is close to inward bulged teeth, but vice versa for 2p = Ns + m machines. However, in both 2p = Ns ± m machines, cogging torques due to dynamic eccentricity and tooth bulge always have a strengthening interaction. The measured cogging torques of the prototypes verify the predicted results.

This paper investigates the influence of arbitrary self-inductance asymmetries in one set three-phase windings on high-frequency pulsating sinusoidal signal injection based sensorless control of dual three-phase interior permanent magnet synchronous machine and the adaptive notch filter is utilized to compensate arbitrary inductance asymmetries. It is found that the demodulated current responses in both two sets generate 2nd- and 4th-order positive and negative sequence error signals because of arbitrary self-inductance asymmetries in one set three-phase windings. These current error signals cause 2nd- and 4th-order estimated position errors in high frequency pulsating sinusoidal signal injection-based sensorless control method. The adaptive notch filter based compensation method can suppress both 2nd- and 4th-order estimated position errors. The compensation method for inductance asymmetries is validated by the measured results.

This article proposes a novel analytical fault model for permanent magnet machines with interturn short-circuits that considers the influence of various factors on the fault current. These factors include mutual inductances between the faulted turns and the remaining healthy windings (faulted and healthy phases), load current, and pulsewidth modulation (PWM) harmonics introduced by the drives. These three factors have been largely neglected by the published methods in literatures but could have significant influence on fault current depending on operating conditions. The investigation in this article shows that, without using the proposed fault model, the conventional fault model that only considers the back-electromotive force in the short-circuited loop and the self-impedance of the short-circuited turns, would underestimate the fundamental fault current by more than 45%. In addition, the proposed model establishes the relationship between the PWM ripple current in the short-circuited turns and that in the faulted phase. By doing so, the fault current including both the fundamental and PWM ripple components can be accurately predicted. The accuracy of the proposed method has been fully validated by a series of experiments.

A strategy of current harmonics and unbalance suppression based on the vector space decomposition control with adaptive linear neuron (ALN) controller for the dual three-phase permanent magnet synchronous machine (DTP-PMSM) drive system is proposed in this article. According to the conventional vector space decomposition theory, the compensation voltages from the ALN controller in the dq-frame and dqz-frame are employed to suppress the 5th and 7th current harmonics and unbalances. Compared with the VSD control using only the PI controller, the current harmonics and unbalance are sufficiently suppressed in the proposed strategy. In addition, unlike the resonant controller discretized by conventional Tustin transformation, the ALN controller has no resonant pole shift to the desired resonant frequency. Therefore, the harmonic suppression effect of the proposed strategy exhibits better performance than the VSD control with the resonant controller, especially in the high-speed region. Finally, the feasibility and effectiveness of the proposed strategy are validated by experimental results under both steady and transient states.

A novel space vector modulation (SVM) strategy for a permanent magnet synchronous machine drive, operating at low switching-to-fundamental frequency ratio, is presented in this article. The proposed strategy is based on applying more than two active vectors in a switching period, aiming to consider the expected rotation and possible movement from one sector to another of the voltage reference vector within that period. The method is simple as the calculation of the duty cycles is still based on the expressions for a conventional SVM. Current distortion and efficiency calculations are also provided, comparing the proposed SVM with conventional SVM and six-step modulation. The results show that the proposed method generates a current distortion similar to the conventional SVM at high frequency ratios but much lower at low frequency ratios. Moreover, comparing with six-step modulation, the current distortion generated by the proposed SVM is always lower, independent of the frequency ratio. On the other hand, the efficiency of the inverter obtained with the proposed SVM is higher than the one obtained with conventional SVM. The effectiveness of the proposed strategy is validated via simulations and experiments.

Compared with conventional non-modular machines, the flux gaps in alternate stator teeth of the modular machines can increase the winding factor and provide flux focusing effect and hence can increase the torque/power density. In addition, the flux gaps can be used as cooling channels to improve the thermal performance. This article investigates an advanced cooling approach using ferrofluid (FF) as coolant to fill in the cavity around end windings and flux gaps. The multiphysics modeling in this article has shown that the influence of the flux gaps on machine thermal performances comes from two aspects: 1) the gravitational effect, the flux gaps allow more self-circulating coolant to transfer heat to the housing. This helps to reduce machine temperature by around 5 °C for a flux gap width of 2 mm; and 2) the magnetic body force, i.e., the thermomagnetic effect. This is very effective in non-modular machine cooling, but its efficiency slightly deteriorates in the modular machines. This is because the self-circulation of FF in the flux gaps due to the gravitational effect has been negatively affected by the thermomagnetic effect.