No abstract

This paper presents the analysis of a surface-mounted permanent-magnet (SMPM) machine for the sensorless control scheme based on the high-frequency fluctuating voltage signal injection method. A simplified high-frequency model of an SMPM machine in the estimated rotor reference frame is developed and a sensorless rotor position and speed estimation algorithm is described. To support this, the high-frequency impedances of an SMPM machine are analyzed by finite-element analysis (FEA) and compared with measurement results using a pulsewidth-modulation (PWM) inverter system under various injection conditions. The results of the FEA and measurements are coincident with each other with some errors due to the nonlinear behavior of the PWM inverter and the SMPM machine. The analysis results give physical insights into selecting the injection conditions for sensorless operation of the SMPM machine even though adjustments considering nonlinear behaviors of PWM inverters are required in the actual operation for the desired performance. The experimental results of speed and position control using a commercial SMPM machine are presented based on the analysis of the SMPM machine for the sensorless control algorithm.Index Terms-High-frequency signal injection, sensorless operation, surface-mounted permanent-magnet (SMPM) machine. NOMENCLATURE SymbolsCoefficient of rotor position information signal (A). Equivalent proportional gain of a bang-bang controller in rotor position estimator. Proportional and integral gains of proportional-integral (PI) controller in rotor position estimator, respectively. Band of bang-bang controller. Resistance ( ).

No abstract

In this paper, a novel voltage modulation scheme is described. With the effective voltage concept, the actual switching times for each inverter arm are deduced directly as a simple form. With the help of the presented zero sequence voltage allocation algorithm, the proposed PWM method has the high performance voltage generation capability exactly same as the conventional space vector PWM method with reduced effort. As well as giving a detailed explanation of the new PWM algorithm, the paper presents the comparison results with the conventional method such as space vector PWM and sinusoidal PWM method. In the proposed PWM method, the execution time can be reduced more than 25% as compared with the conventional space vector PWM method, and memory size will be minimized to 15% of the conventional PWM method.

No abstract