Application of Neuro-Fuzzy Control Systems for Increasing the Energy Efficiency of Wind Turbines

Wind has the variable nature and possesses a huge amount of available energy to be transformed into the usable energy with the desired maximum efficiency. Therefore, forecasting of wind turbine power generation, increasing of energy efficiency and optimization of wind turbine power output are of great importance. Modernization of wind turbine control systems stimulates solution of these problems, including the development of systems based on artificial intelligence and computer technologies. Now, special hybrid systems developed with the use of neural networks and fuzzy logic are popular. Combination of these methods of finding optimal solutions is used for modernizing control systems. It includes benefits of fuzzy logic and neural networks and hides their drawbacks, thus having the following main advantages: clear knowledge representation, retraining in the case of changing the input values. On the one hand, these models of control systems can be developed in the form of demonstrative and understandable fuzzy production rules. On the other hand, these rules are produced with the help of methods of neural networks that is more convenient and not so difficult to calculate with the possibility of self-learning. The paper presents the simulation results for the neuro-fuzzy controller integrated into the rotor speed control system for the wind turbine wheel. Obtained results have shown that combination of fuzzy logic and neural networks allows optimizing the WT output power that directly influences on its efficiency.