Intelligent vector control for a novel mechatronic modeling of a 1.5MW variable speed WECS using the bicausality concept under a real wind flow: A Bond Graph Approach

Abstract

Enhancing the dynamic behaviors of Wind Energy conversion Systems (WECS) based on a doubly fed induction machine (DFIG) is one of the most attractive challenging in the field of maximizing renewable energy's output power, improving its quality and ensuring its efficient grid integration. For this purpose, a novel mechatronic modelling of a 1.5 MW variable-speed WECS based on a doubly fed induction machine using the Bond graph Approach is proposed, and three robust control strategies are displayed to regulate: the DFIG's active and reactive powers for the networks side converter (NSC), maximizing the extracted power and controlling the pitch angle for the load area. The flux oriented vector control laws presented in this contribution are based upon the independent control of magnetic flux and the electromagnetic torque inside the DFIG machine. They are derived from the integrated model's Inverse Bond Graph (IBG). This robust control strategy based on the bicausality concept enabled us to elaborate the Vector Control model that relies on a Radial Basis Function RBF-VC to be trained online with an optimal dataset under some real wind profiles. The robustness and effectiveness of the proposed model and RBF-VC controller are verified. The simulation results were conducted to confirm the validity of the proposed technique. Hands-on experience is carried out by means of the 20-sim software package.

Keywords

• Wind turbine, Bond graph modeling, Mechatronic control
• DFIG, WECS -RBFNN
• ANN online training – Bicausality- inverse bond graph