Application of Hill Climbing Method in Position Angle Compensation for SPMSM

Zhe Song,Weihong Zhou,Xi Xiao,Jiayue Zhou
DOI: https://doi.org/10.1007/s10489-024-05594-9
IF: 5.3
2024-01-01
Applied Intelligence
Abstract:Surface-mounted permanent magnet synchronous motor (SPMSM) is widely used in the industrial field with excellent performance, and the rise of artificial intelligence has also promoted its development. In order to improve the estimation accuracy and response of speed and position angle in the SPMSM sensorless control system, a novel sliding mode control model reference adaptive system (NSMC-MRAS) observer based on variable step hill climbing method for online optimization is proposed. Firstly, an MRAS adaptive observer was constructed, and a conventional SMC controller is used instead of the PI regulator to improve the robustness of the SPMSM parameters. Aiming at the problem of chattering caused by the sign function sgn(s), an improved NSMC-MRAS sensorless control strategy using the continuous function sigmoid(s) is proposed, which effectively suppresses the chattering phenomenon. To address the position estimation errors caused by non ideal factors such as control delay, filtering phase shift, and parameter deviation, the artificial intelligence variable step hill climbing method is used for online optimization and adaptive compensation. The experimental results show that the proposed NSMC-MRAS sensorless control strategy based on variable step hill climbing method for online optimization can quickly and accurately estimate the speed and position angle of SPMSM, improved the control performance and intelligence level of the system.
What problem does this paper attempt to address?