Full-speed domain position sensorless control strategy for PMSM based on a novel phase-locked loop

• Published in: Control engineering practice 2024-11, Vol.152, p.106058, Article 106058
• Main Authors: Yao, Guozhong, Yang, Zixian, Han, Shaojun, Wang, Zhengjiang
• Format: Article
• Language: English
• Subjects: I/F start-up; Luenberger observer; New PLL; SPMSM
• ISSN: 0967-0661
• DOI: 10.1016/j.conengprac.2024.106058

This paper proposes a full-speed-domain position-sensor-less control strategy for precise control under forward and reverse rotation conditions to address the weak convex polarity of surface-mounted permanent magnet synchronous motor (SPMSM). The strategy comprises several key stages: pre-positioning of the rotor, constant current variable frequency (I/F) start-up, construct the Luenberger State Observer, and utilization of an improved phase-locked loop (PLL) for position estimation. In the pre-positioning stage, a constant amplitude current is applied to drag the rotor to a predetermined position. Subsequently, the I/F start-up stage accelerates the motor to a predetermined speed before transitioning to the Luenberger observer for closed-loop speed control, which is based on an extended back electromotive force (back-EMF) a two-phase rotating coordinate system. The improved PLL for position and speed estimation features three components: a phase discriminator (PD), a voltage-controlled oscillator (VCO), and loop filter (LF). Experimental results demonstrate the efficacy of the proposed strategy, showing quick start-up response, speed estimation error below 2 RPM, rotor position estimation error under 0.6 degrees post-loop closure, stable tracking during rapid speed changes, and consistent accuracy and stability even under reverse rotation conditions, thereby meeting the control strategy’s objectives. •Propose a full speed domain sensorless control strategy.•Rapid and smooth switching of operating states.•Stable and precise control in both forward and reverse rotation.•Control strategies are experimentally verified using a hardware platform.