3 March 2022 in IEEE Access
Accurate machine parameters and rotor position information are essential in vector-controlled motor drive systems. However, machine parameter variations by various factors such as the current and the temperature degrade the performance of vector control. Also, a position sensor such as an encoder and a resolver increases the drive system cost. This paper proposes model predictive current control (MPCC) with the online parameter estimation for synchronous reluctance machines controlled by a high-frequency signal injection position-sensorless method. This approach removes the need for accurate knowledge about the system and eliminates the need for the position sensor. The proposed method adopts a recursive least-square (RLS) to estimate the electrical machine parameters in real-time. The estimated parameters are used for the deadbeat continuous control set (CCS) MPCC and the position-sensorless control. The high-frequency signal injection method is modified to be suitable for the proposed CCS-MPCC method, ensuring stable operation in the low-speed regions. Simulation and experimental results are provided to verify the performance of the proposed control method.
2 February 2024 in IEEE Transactions on Energy Conversion
The induction motor drive with indirect field-oriented control (IFOC) is widely adopted due to its simple structure and easy implementation. The rotor time constant plays the most crucial role in IFOC. However, the rotor resistance varies with temperatures, and the rotor inductance changes with the current magnitude in practice. Furthermore, an accurate estimation of such parameters is challenging. The parameter mismatch between the actual and the estimated rotor time constant deteriorates the performance of the IFOC vector control due to an error in the estimated rotor flux position. This paper presents the theoretical analysis of the effects of the inductance errors in the conventional Q-MRAS method, which is widely used to compensate for the rotor time constant. Next, this paper proposes the dual MRAS consisting of a novel stator flux-based MRAS and the conventional Q-MRAS. The proposed dual MRAS method can simultaneously compensate for the error of inductance and rotor time constant, thereby improving the performance of IFOC. The robustness and effectiveness of the proposed method have been validated through simulations and experiments.
12 May 2025 in IEEE Transactions on Industry Applications
Back-EMF-based sensorless vector control for a permanent magnet synchronous motor (PMSM) requires an additional startup method because the position information cannot be properly extracted from the back-EMF at low-speed regions. A current-frequency (I-F) control is typically adopted as the startup strategy due to its simple structure and stable control. However, the performance of the conventional I-F control deteriorates when the deadbeat model predictive current control (MPCC) is used for current regulations. Also, the I-F control requires a significantly high current for stable operation, resulting in low efficiency. This paper proposes a model reference adaptive system (MRAS) based I-F control to achieve maximum torque per ampere (MTPA) control and provides an online motor parameter compensation for a smooth transition from I-F to sensorless vector control. As a result, the sensorless control can be effectively operated in the deadbeat model predictive current controlled surface-mounted PMSM (SPMSM) drives. Simulation and experimental results are provided to verify the performance of the proposed method.