Soft Computing based Control Strategies to Improve the Performance of Wind Turbines in Uncertain Conditions

The power harvesting capacity of wind turbines is the major performance metric that depends on weather conditions and controller accuracies. This dissertation work presents a statistical estimation of wind power. It examines the impact of uncertainties on wind power generation. Also, new control approaches are developed to improve the output power of horizontal axis wind turbines. Statistically, the wind power capacity of the Adama II wind farm site is estimated. In 2019, the yearly average wind power density of the site is 532 W/m2 . The uncertain weather conditions are one of the many factors that affect the energy-yielding ability of wind turbines. Thus, this study statistically investigates uncertain weather conditions, parametric uncertainties, and analyses the performance of wind turbines using real-time data. Due to variations in an annual average air density and yearly average wind speed, and uncertainties in aerodynamic parameters of a wind turbine, the annual uncertainty in the output power is 32% of the rating of the selected 1.5 MW wind turbine. The linearized model of doubly-fed induction generator (DFIG) based wind turbine performance is estimated by using the system identification technique. The parametric uncertainty in this model and its effects on the dynamic and steadystate performances of the DFIG based wind turbine are examined. It caused instability in the dynamic model and influenced the steady-state performances. The output power of wind turbines that operate under uncertainty is improved by three new approach control strategies: model identification (MI) based, Adaptive Network Fuzzy Inference System (ANFIS), and the Mandani Fuzzy Inference System (MFIS). These three methods focused on improving the power conversion coefficient of wind turbines by incorporating control techniques of the turbine blade pitch actuator. The proposed control strategies are implemented and verified on a 1.5 MW DFIG based wind turbine using MATLAB/Simulink software. The results of the study show that the SI-based controller and MFIS based controller improve the output power of the wind turbine by 14.12% and 16.74%, respectively in comparison to factory test report data of the SANY SE7715 model wind turbine. The simulation results verify that the ANFIS based controller improves the output power of the Gamesa G80 wind turbine by 9.47% in comparison to data taken from the power curve of G80 manufacture (Gamesa). Thus, the researcher recommends the practical implantation of the proposed controllers to efficiently harness wind energy.