In this paper, a pole-changing (PC) vernier permanent magnet (VPM) motor with a special S-Iron-S-N-Iron N rotor structure is proposed. By this rotor design, both the 1st and 3rd PM magnetomotive force harmonics can be effectively utilized for PC operation. Most importantly, the proposed rotor is adaptively designed to reconfigure the d-q axis magnetic circuits before and after PC. Accordingly, a reverse saliency effect can be obtained in the post-PC mode, which contributes to higher torque density and an extended operation region. In addition, specific rotor barriers are introduced as additional design parameters to further regulate torque outputs and characteristic currents in both PC modes, thus widening the whole operation range. The electromagnetic performances of the proposed motor are analyzed by finite element analysis (FEA). A prototype is manufactured and tested, and the experimental results agree well with the theoretical analysis and FEA predictions, verifying the effectiveness of the proposed structure.
For synchronous machines, Park's circuit models are formulated as coupled $dq$-frame circuits based on the rotor's perspective. Under balanced stator voltage conditions and when the rotor electrical speed matches the stator frequency, Park's models are analytical: the state variables remain constant at steady state, enabling linear model extraction through Jacobian linearization. However, when these conditions are not satisfied-for example, when the stator voltage is unbalanced, contains harmonics or resonant components, and/or when the rotor electrical speed differs from the stator frequency-Park's models are no longer analytical and do not readily provide insights or facilitate straightforward analysis. In this paper, we derive dynamic phasor-impedance circuit models of synchronous generators, with the effect of rotor saliency and rotor damping windings included. The circuit models enable unified steady-state and dynamic analysis under a wide range of operating conditions. Three applications of the proposed models are demonstrated, including negative-sequence impedance calibration, average torque calibration during machine starting, and generator's frequency-domain impedance characterization for subsynchronous resonance (SSR) stability prediction.
Some nominal-frequency-anchored (NFA) controls enable grid-following (GFL) renewable energy converters to ride through deep voltage sags in weak grids. However, anchoring the frequency to its nominal value leads to a sustained frequency mismatch and monotonic power-angle divergence when the grid frequency deviates. To address this limitation, this paper proposes adaptive frequency anchoring (AFA) to support both low-voltage ride-through (LVRT) and low-frequency ridethrough (LFRT). AFA retains the original NFA mechanism and adds a frequency anchoring loop that adjusts the phaselocked loop (PLL) frequency reference based on the active power deviation, enabling the converter to accommodate grid frequency changes. Theoretical analysis demonstrates the effectiveness of AFA in both centralized and decentralized implementations. A quantitative design method is also developed to determine the gain of the frequency anchoring loop. Electromagnetic transient (EMT) simulations and controller-hardware-in-the-loop (CHIL) experiments validate the effectiveness of AFA.
This paper analyzes the fundamental cause of performance degradation in extended back-EMF (EEMF)-based sensorless control under synchronous PWM operation and proposes a compensation method to address this issue. In synchronous PWM, the control period varies in real time depending on operating conditions and changes abruptly, particularly during pulse number transition intervals. If these characteristics of the variable control period are not properly reflected in the sensorless control algorithm, a time-reference mismatch arises between the voltage and current signals used in the EEMF observer. As a result, rotor position estimation errors accumulate, leading to degraded synchronization and sensorless control performance as well as the occurrence of current surges. The proposed method systematically compensates for the time delay characteristics of the input variables caused by the variable control period, thereby maintaining the position estimation accuracy of the EEMF observer and ensuring continuity of the current response even during pulse number transitions. In addition, the proposed approach improves synchronization performance and torque production. Experimental results obtained using a 332-kW IPMSM validate that the proposed method effectively preserves synchronization and suppresses current surges during sensorless operation, confirming its suitability for reliable operation in high-power propulsion systems.
This paper proposes an improved high-precision semi-analytical (SA) model based on the harmonic modeling (HM) method for magnetic field prediction in permanent magnet synchronous motors (PMSMs) with fully closed-slot. To overcome the limitation of conventional geometric simplification that directly neglects the highly saturated tooth-shoulder region, a layered geometric approximation is developed, in which the inclined tooth-shoulder is represented by multiple equivalent annular sectors based on the area-equivalence principle while remaining compatible with the HM formulation. An adaptive relaxation-based nonlinear solution algorithm is further incorporated to account for the nonlinear permeability of the iron core and improve convergence stability and efficiency. The effectiveness of the proposed SA approximate model is validated through finite element (FE) analysis and prototype-based experiments, and the influence of the number of concentric sectors on solution accuracy and computation time is systematically investigated. Moreover, the proposed approximation methodology can also be extended to other electromagnetic structures with irregular radial or tangential geometries.
