Impact of switching functions on sliding mode observer in sensorless PMSM drives

The Sliding Mode Observer (SMO) algorithm has emerged as a promising solution for estimating the position and speed of motors, effectively replacing traditional physical sensors. This paper focuses on investigating the impact of switching functions in SMO on the control performance of Permanent Magnet Synchronous Motors (PMSMs) under the Field-Oriented Control (FOC) strategy. Through simulations conducted in Matlab/Simulink, the study evaluates the performance of SMO using three types of switching functions: Signum, Sigmoid, and Hyperbolic Tangent (Tanh). The results indicate that the Signum function induces chattering, significantly degrading system stability and accuracy. In contrast, the Sigmoid and Tanh functions deliver superior performance, with the Tanh function achieving the most optimal results. These findings provide a solid foundation for selecting appropriate switching functions to optimize SMO in sensorless PMSM control, while also opening new prospects for minimizing estimation errors and enhancing system responsiveness.