In this paper, a new rotor position estimation method of an interior permanent-magnet synchronous motor (IPMSM) is proposed. It is motivated by a conventional sensorless control based on the high-frequency rotating voltage signal injection in the stationary reference frame, but extracts the rotor position information from the relation between injection voltages and induced currents in the high-frequency voltage model of an IPMSM. The rotor position can be directly calculated from currents obtained by transforming the rotation matrix in the stationary reference frame. It is also calculated by using the induced current envelope through the all-pass filter. Unlike the conventional heterodyning demodulation process, the proposed method provides the accurate rotor position information without a time delay of the rotor position by the band pass filter or low pass filter. Thus, the overall dynamic performance of IPMSM is enhanced by the accurate rotor position estimation than conventional methods. The various experimental results confirm the effectiveness of the proposed method.

This paper proposes a model predictive control based on three voltage vectors for an interior permanent-magnet synchronous motor (IPMSM) in a rotating reference frame. The proposed method can obtain the reference voltage vector quickly by predicting only one voltage vector during one sampling period using the characteristics of the deadbeat control. In addition, to obtain better steady-state performance, the optimal vector duration ratios can be obtained from the cost function to minimize the error between the predicted reference voltage vector and the synthesis vector by using the switching sequence relationship of a symmetrical 3-vector, which is different from applying a single voltage vector in the conventional finite-control-set model predictive control (FCS-MPC). To verify the proposed method, the experimental results are compared with the conventional FCS-MPC and two-vector-based MPC.

This paper presents an improved saliency-based position sensorless drive of an interior permanent-magnet synchronous machine (IPMSM) with a single current sensor, used for low cost applications. The phase current, reconstructed from a dc-link current, shows a reconstruction error resulted from the difference of the sampling point from that of a full current sensor drive. This reconstruction error results in degradation of the overall sensorless control performance. In this paper, we analyze the impact of the reconstruction error depending on the injection frequency for a high frequency voltage signal injection sensorless control. Based on this analysis, we propose an improved position sensorless drive that employs a new current reconstruction method. In order to minimize the reconstruction error, the proposed current reconstruction method predicts the dq-axis currents with the same sampling point as a full current sensor drive. Several experimental results are provided to verify the analysis of the reconstruction error and the proposed method's improved sensorless control performances.

This paper proposes a new position sensorless method for a direct torque control system using the inductance inflection point of a switched reluctance motor (SRM). The SRM essentially requires knowledge of the rotor position in the torque control system; this is because it generates torque by exciting the stator windings according to the rotor position. This rotor position information is an important factor used to obtain phase commutation point and to calculate instantaneous torque. Various rotor position detection methods have been used, but most of the methods increase the construction cost by requiring additional hardware and use many parameters for high-order analysis modeling, causing problems such as increasing the number of calculations. Therefore, in this paper, we propose a method using the inductance inflection point to solve these problems. It is possible to estimate the position of the rotor by detecting where the inductance inflection point occurs due to the rotor and stator overlapping with each other, and the phase commutation point and instantaneous torque can be calculated. The calculated instantaneous torque is used for direct torque control and torque ripple reduction through the torque-sharing function. The proposed method is verified through simulation and experiments.

Recently, an omnidirectional wireless power transfer (WPT) has attracted attention as a solution to the limitations of free-positioning and mobility of conventional WPT. However, the efficiency of the omnidirectional WPT is reduced due to the low mutual inductance relative to the transmitter size. Therefore, this article proposed an optimal magnetic structure to improve efficiency without increasing the size of the transmitter. Four magnetic structure candidates (spherical, cubical, crossed bar, three-orthogonal plane core) are presented and their performances are compared using finite element analysis software. A Pareto optimal set is selected using a multiobjective optimization design, and the optimal magnetic structure is determined considering the weight of the objective variable and the Tchebycheff distance. The optimal magnetic structure and omnidirectional WPT prototype are built for performance verification. The experimental results are consistent with the simulation results and demonstrate that the optimal magnetic structure improves the efficiency in the three-dimensional space. When the power transmission distance is 150 mm, the system efficiency is improved up to 18.6%.