The software-based resolver-to-digital converter (RDC) is cost-effective and widely used to obtain an angular position. Angle estimation is an essential procedure in an RDC, but conventional angle estimation methods face difficulties in parameter tuning and lack analysis of the parameters. To overcome these problems, a novel angle estimation method based on an α-β-γ filter is proposed in this paper, which only requires tuning one parameter. The performance measures to gauge the tracking ability, noise filtering strength and transient response performance are derived. In addition, this paper analyzes the impacts of nonideal characteristics when using frequency-shifting technique. To suppress the impacts, an amplitude corrector is further proposed. Combined with the α-β-γ filter and the amplitude corrector, the proposed RDC has excellent performance in terms of noise rejection, fast position response and tracking ability. Experimental tests verify the effectiveness and feasibility of the proposed RDC system.

High-frequency injection (HFI) is widely adopted for the sensorless control of permanent magnet synchronous motors (PMSMs) at low speeds. However, conventional HFI strategies relying on fixed-frequency carrier modulation and square-wave injection concentrate current harmonic energy within narrow spectral bands, thereby inducing pronounced high-frequency motor vibrations and noise. To mitigate this issue, this paper proposes a noise suppression strategy based on synchronized periodic frequency modulation (PFM) of both the carrier and high-frequency square-wave signals. By innovatively synchronizing the periodic modulation of the triangular carrier in space vector pulse width modulation (SVPWM) with the injected high-frequency square wave, harmonic energy dispersion and noise reduction are achieved, substantially lowering peak acoustic emissions. First, the harmonic characteristics of the voltage-source inverter output under symmetric triangular carrier SVPWM are analyzed within a sawtooth-wave PFM framework. Concurrently, a harmonic current model is developed for the high-frequency square-wave injection method, enabling the precise derivation of harmonic components. A frequency-synchronized modulation strategy between the carrier and injection signals is proposed, with a rigorous analysis of its harmonic suppression mechanism. The rotor position is then estimated via high-frequency signal extraction and a normalized phase-locked loop (PLL). Comparative simulations and experiments confirm significant noise peak attenuation compared to conventional methods, while position estimation accuracy remains unaffected. This work provides both theoretical and practical advancements for noise-sensitive sensorless motor control applications.

Field-weakening control strategy extends the speed range of permanent magnet synchronous motor (PMSM) drives without modifying hardware, but suffers from complexity and inefficiency. This paper proposes a novel field weakening strategy based on power regulation and an adaptive observer. A power mapping model is formulated to achieve real-time generation of the dq-axis current reference. A discrete state space representation of the power loop demonstrates zero steady state speed tracking error through power regulation, while pole placement analysis provides insights for stability optimization. A speed adaptive observer integrates the sensorless operation and the field-weakening current trajectory generation without additional control loops, improving voltage utilization, while also reducing the torque ripple. Compared with conventional approaches, the proposed method achieves higher DC‑link voltage utilization and improves the load capability. Experimental results on a PMSM test platform validate its effectiveness.

In permanent magnet synchronous motor (PMSM) drives, compared to the traditional approach of using at least two current sensors, single dc-link current sensor control (SDCSC) offers cost reduction and a more compact drive system. The switching-state phase shift (SSPS) method can suppress the current reconstruction dead zone in the low modulation region and sector boundary regions for SDCSC. Still, due to the asymmetry of the SSPS method, reconstruction errors and current harmonics are introduced, which affect the bandwidth of the PMSM current loop. This paper analyzes the errors introduced by the SSPS method and proposes a robust dq-axis current observer to resolve this problem. Based on the theory of Lipschitz nonlinear systems, the tuning principle of the correction gain for the proposed observer is derived. A speed decoupling term is proposed to maintain the desired observer performance, and the computation of the optimal correction gain combining the speed decoupling term and the gradient method is given. A disturbance compensation component is further proposed to enhance the robustness of the proposed current observer. The proposed method suppresses the reconstruction errors of the SSPS method and improves the current loop bandwidth under SDCSC. Experiments are performed on a PMSM platform to verify the effectiveness and feasibility of the proposed method.

In active magnetic bearing (AMB) systems, harmonic currents induced by mass unbalance and sensor runout cause the system to generate harmonic vibrations. Conventional harmonic current suppression methods depend highly on accurate rotational speed information, which limits their application in scenarios without rotational speed sensors. To address this problem, this paper first analyses the main sources of harmonic vibration and then proposes an improved multiple second-order generalized integral frequency-locked loop (IMSOGI-FLL) algorithm. The algorithm achieves adaptive estimation of the rotational frequency by utilizing the frequency of the fundamental component in the disturbance signal and effectively suppresses the harmonic currents generated by the mass unbalance and sensor beating. In addition, in order to ensure the stability of the system in the operating range, this paper designs a simple phase compensator without parameter switching, which can improve the stability range of the system in the low-frequency band. The experimental results show that the proposed algorithm can not only accurately estimate the rotational frequency but also significantly eliminate the harmonic currents caused by mass unbalance and sensor runout, which provides an effective solution for harmonic current suppression in the case of no speed sensor.