Abstract:
Operation and maintenance tasks on the wind turbines have an essen- tial role to ensure the correct condition of the system and to minimize losses and increase the productivity. The condition monitoring systems installed on the main components of the wind turbines provide information about the tasks that should be carried out over the time. A novel statistical methodology for multivariable analysis of big data from wind turbines is presented in this paper. The objective is to analyse the necessary information from the condition monitoring systems installed in wind farms. The novel approach filters the main parameters from the collected signals and uses advanced computational techniques for evaluating the data and giving mean- ing to them. The main advantage of the approach is the possibility of the big data analysis based on the main information available.
