Smoothing of wind farm output power using prediction based flywheel energy storage system

Being socially beneficial, economically competitive and environment friendly, wind energy is now considered to be the world's fastest growing renewable energy source. However, the stochastic nature of wind imposes a considerable challenge in the optimal management and operation of wind power system. Wind speed prediction is critical for wind energy conversion system since it greatly influences the issues related to effective energy management, dynamic control of wind turbine, and improvement of the overall efficiency of the power generation system. This thesis focuses on integration of energy storage system with wind farm, considering wind speed prediction in the control scheme to overcome the problems associated with wind power fluctuations. In this thesis, flywheel energy storage system (FESS) with adjustable speed rotary machine has been considered for smoothing of output power in a wind farm composed of a fixed speed wind turbine generator (FSWTG). Since FESS has both active and reactive power compensation ability, it enhances the stability of the system effectively. An efficient energy management system combined with supervisory control unit (SCU) for FESS and wind speed prediction has been developed to improve the smoothing of the wind farm output effectively. Wind speed prediction model is developed by artificial neural network (ANN) which has advantages over the conventional prediction scheme including data error tolerance and ease in adaptability. The model for prediction with ANN is developed in MATLAB/Simulink and interfaced with PSCAD/EMTDC. Effectiveness of the proposed control system is illustrated using real wind speed data in various operating conditions.