Commande avancée d’une machine asynchrone multi-phasées

Abstract

This thesis investigates advanced approaches to enhance the control of the Dual Star Induction Machine (DSIM), a critical component in industrial applications that demand both robustness and efficiency. Initially, we reviewed conventional control strategies, such as Indirect Field Oriented Control (IFOC) and Direct Torque Control (DTC), which, despite their effectiveness, heavily rely on precise knowledge of system parameters and are sensitive to disturbances.

To address these shortcomings, we developed innovative solutions, including the Fractional PI Controller (FOPID) and Sliding Mode Control (SMC), further optimized using advanced algorithms like Particle Swarm Optimization (PSO) and Grey Wolf Optimization (GWO) to ensure superior performance.

The research findings highlighted the effectiveness of FOPID regulators and higher-order sliding mode controllers (STSMC and TOSMC) in mitigating the adverse effects of chattering while enhancing stability and robustness against parameter variations. Simulations revealed significant improvements, including faster response times and greater control accuracy, alongside a noticeable reduction in torque ripples and current fluctuations. Moreover, the robustness of these proposed methods was validated even under diverse dynamic conditions, showcasing their adaptability and suitability for modern industrial environments with stringent performance requirements.