Data-driven modelling of a doubly fed induction generator wind turbine system based on neural networks

In a wind power system, the wind turbine captures wind energy and converts it into electric energy through a coupled rotating generator. This renewable energy conversion system usually consists of a wind turbine, rotor, gearbox and mostly a doubly fed induction generator (DFIG). It is a complex non-linear multi-input multi-output system with many uncertain factors. Meanwhile, the dynamics of the system is quite dependent on the wind velocity. Traditional analytical methods are quite difficult to model such a complex system. The recently developed data-driven method can be a suitable modelling technique for such system. Using a large amount of input-output on-line measurement data from the selected months, neural networks and neuro-fuzzy networks are fully utilised to model the DFIG. Detailed analysis and comparisons with the classical system identification techniques are addressed to show the advantages of the data-driven DFIG modelling approach.