1 A Review on Sensorless Speed Control of Interior Permanent Magnet Synchronous Motor IPMSM Abdulhasan Altaey 1 , Halis Altun 2 , Osman Bilgin 3 1 Electrical & Computer Engineering Ph.D. Student, Mevlana University 600912007@st.mevlana.edu.tr 2 Electrical & Electronics Engineering Assoc. Prof. Dr., KTO Karatay University haltun@mevlana.edu.tr 3 Electrical & Electronics Engineering Assoc. Prof. Dr., Selcuk University obilgin@selcuk.edu.tr Abstract In this paper a hybrid approach has been proposed for sensorless speed control and parameter estimation of interior permanent magnet synchronous motor (IPMSM). In this approach, Modal Reference Adaptive System (MRAS) and Extended Kalman Filter (EKF) are used in parallel. The rationale behind mixing EKF with MRAS is that MRAS is a second order state equation which means estimating more than two parameters are impossible, while controlling of IPMSM requires estimation of many parameters which change simultaneously and might be estimated using EKF. As EKF is time consuming algorithm, an FPGA implementation is being proposed. The MRAS and two reduced order EKF would work in parallel on FPGA platform. We expect to obtain a wide speed control range using maximum torque per ampere MTPA in the low speed range and Field Weakening in the high speed range. 1. Introduction To achieve best performance of Interior Permanent Magnet Synchronous Motor, a variety of sensorless methods have been proposed by researchers. However during motor working cycle, the temperature increases and the load changes so the parameters change continuously according to the load and environment conditions. Also the armature current demagnetizes the permanent magnet [1]. All these changes will certainly affect the sensorless operation. Hence it is necessary to correctly estimate the parameters simultaneously during sensorless control. The rotor position can be sensed from the position related signals e.g. back EMF, third harmonics back EMF. If the motor is star connected, the summation of the input three phase voltages results in a component related with the third harmonic back EMF and theoretically is not effected by noise and motor parameters, hence third harmonic EMF is preferred to fundamental back EMF [2]. The third harmonic flux-linkage which results from the integration of the third harmonic back EMF travels in the air gap with the same synchronous speed as the fundamental component flux linkage and related to rotor position [7]. However this method cannot be applied to the IPMSM due to the fact that the back EMF in IPMSM depends not only on the position information of the rotor, but also it depends on the direct and quadrature axis stator currents (1) & (2), hence depends on the load conditions. The using of the extended electromotive force EEMF needs other strategy [5] and below are the general equations for EEMF in the stator stationary axes: (1) (2) where ω re angular velocity in electrical degree. Θ re rotor position in electrical degree. R S stator resistance. L d , L q direct and quadrature axis stator inductances. λ m flux linkage generated by the permanent magnets. i d ,i q stator direct and quadrature axis currents in rotating reference frame respectively. i α , i β stator currents in stationary stator reference frame. E ext extended electromotive force EEMF There is a differential term of iq in the EEMF equation (2). This means that even when the motor’s velocity is near zero, the EEMF is not zero if the q-axis current iq is changing. This property will be useful for standstill and low-speed drives [5]. The sensorless estimation can be done in different methods: i) Signal Injection Based method in the low and zero speed: Signal Injection Based Method is used in the low and zero speed only, by superimpose a high frequency voltage to the stator voltage. The rotor saliency affects the magnitude of the high frequency stator current and from which the rotor position can be identified [3]. However unwanted torque may 722 Otomatik Kontrol Ulusal Toplantısı, TOK'2015, 10-12 Eylül 2015, Denizli