Simultaneous Estimation of Velocity and Mass Using an Augmented Extended Kalman Filter for Sliding Mode Control of Precision Magnetic Levitation Systems

초록

This paper proposes an augmented extended Kalman filter (AEKF) to enhance the tracking performance of sliding mode control (SMC) for precision magnetic levitation systems (MLSs). The AEKF simultaneously estimates the unmeasurable velocity and mass, thereby improving tracking accuracy during the sliding phase. Both simulation and experimental results demonstrate that the proposed observer-based AEKF-SMC approach effectively reduces tracking errors under parameter uncertainties compared with conventional differentiator-based SMC. Furthermore, the control scheme requires only a single position sensor and eliminates the need for low-pass filtering and differentiation, enabling real-time implementation on cost-effective microcontrollers such as the ESP32.

키워드