Sensorless Super-Twisting SMO Based PMSM Drive With Improved DANF-PLL for Hybrid Three-Wheeler EV Application

In this work, an improved super-twisting sliding mode observer (ST-SMO) is proposed to enhance speed and position estimation in a PMSM drive. This approach leverages the advantages of both higher-order sliding mode observers and conventional sliding mode observers. Additionally, speed-dependent adaptive gains are incorporated, which ensure stability. The improved ST-SMO provides smooth and accurate speed and position estimation while effectively addressing challenges such as chattering and singularities. However, estimation errors may still occur due to DC offsets and noise in the back-EMF signal. To mitigate these issues, an advanced notch filter is integrated with improved ST-SMO to refine the estimated back-EMF signal. To achieve accurate speed and position estimation from the filtered back-EMF, an improved PLL is employed. This PLL incorporates a feedforward phase approach instead of frequency, enhancing both dynamic speed response and noise rejection capabilities. Additionally, a rooftop solar photovoltaic (PV) array is integrated in the system to assist the battery, extending vehicle's range and enabling battery charging. Solar PV array is connected to the DC link through a DC-DC Cuk converter. System effectiveness for three-wheeler (3W) electric vehicle (EV) applications is validated through simulation and experimental results, demonstrating the successful integration solar PV array with battery, in a sensorless PMSM drive.