Optimal drive train management of wind turbine using LiDAR-assisted predictive control strategy

In horizontal axis wind turbines, the maximum power extraction under extreme wind condition is tedious. The region-3 operation of turbine induces structural turbulence and fatigue torsional stress on drive train arrangement. The drive train encompassing gear box and energy transfer shafts are susceptible to sudden acceleration and de-acceleration inducing torsional torque on drive train. The wind turbine payback is a long-run process. To ensure sustained improvement in the performance of turbine the contradiction between the strict MPPT causing fatigue loading and fatigue stress damping causing reduced power extraction that is being resolved in region-3 operation. The conventional feedback controller fails to yield optimal converging point. Hence, this work proposes a preview-based Hybrid Multiple Point Model Predictive Controller which optimally controls the MPPT of wind turbine and maintains the fatigue load by active maneuvering of high-speed shaft brake. The proposed controller includes the fatigue loading variable in its usual MPPT control law. The simulations are performed using NREL FAST 5 MW wind turbine model interfaced with MATLAB Simulink. The credibility of the proposed controller is validated using the power curve and fatigue life analysis report and is found to optimally resolve the contradictory cost functions of maximum power production and fatigue stress at an operating point. Also, the results of the proposed LiDAR-based controller are compared with contemporary feedback controllers. The real-time wind data from wind monitoring station Kayathar in Indian state of Tamil Nadu are taken for analysis, and the research proposals are made for wind turbines in that region.