Enhanced Sensor-Less Control of DFIG-Based Generators Using Hybrid MRAS-ANN Observer and PSO Parameter Optimization

This study presents a robust sensor-less Sliding Mode Controller (SMC-l) for a wind turbine system with a Doubly Fed Induction Generator (DFIG). The control system focuses on managing the stator's active and reactive power through an SMC-l controller and utilizes a bidirectional converter for power transfer between the DFIG's rotor and the grid. A Model Reference Adaptive System (MRAS) observer, enhanced with an Artificial Neural Network (ANN) controller instead of the traditional PI regulator, is implemented to estimate mechanical parameters. The Particle Swarm Optimization (PSO) algorithm fine-tunes the parameter gains of the proposed controller. Simulation results demonstrate the improved robustness and effectiveness of the sensor-less SMC-l control system with the MRAS-ANN observer compared to conventional MRAS methods, positively impacting power generation.