Graphical analysis of wind turbine dynamics trough SCADA data mining

The widespread diffusion of SCADA control systems in modern wind turbine technology has revolutionized the management of machine portfolios. Potentially, every wind farm owner can develop its own tools for performance evaluation, fault diagnosis and prevention, optimization. Wind technology in the age of information therefore faces the challenge of elaborating vast amounts of data, processing them into knowledge, visualizing them intuitively. Due to the stochastic nature of the source, this ambition lies at the crossroad of engineering, physics, statistics and computer science and has therefore stimulated considerable debate in the scientific literature and fruitful collaborations between academia and industry. In this study, a mathematical and graphical method is proposed for elaborating turbine state dynamics and SCADA measurements in a combined way. Its strength lies in intuitiveness and versatility: it can be be used for a bird's eye view on a portfolio of machines, for inquiring whether maintenance programs are been planned judiciously, for a first investigation on the reasons of faults, especially for SCADA systems which are poor in providing information to the farm owner. Further, the graphical output does not depend on the input language of the single SCADA supplier and is potentially universal. Actually, this method is tested on some wind farms owned by Renvico srl having different SCADA systems suppliers.