A Dynamic Self-Adjusting System for Permanent Magnet Synchronous Motors Using an Improved Super-Twisting Sliding Mode Observer

To enhance the robustness of sensorless control in permanent magnet synchronous motors (PMSMs) under parameter mismatches, this paper proposes a novel sliding mode observer (SMO) that automatically adjusts the error factor. The purpose is to enable the precise observation of rotor position in PMSMs while simultaneously suppressing chattering and simplifying the design process. First, an SMO based on an adjustable error factor is designed, which reduces chattering and eliminates the need for a low-pass filter (LPF). The impact of the error factor within the SMO is then analyzed, including its effects on the estimation of current, speed, and position, and a method for determining the error factor based on these estimated values is introduced. This method uses a neural network algorithm to balance chattering suppression with high control accuracy. Finally, a neural network-based self-adjusting SMO model is proposed to automatically adjust the error factor based on motor operating conditions. Simulation and experimental results demonstrate the feasibility and effectiveness of this approach.