Floating offshore wind turbine fault diagnosis via regularized dynamic canonical correlation and fisher discriminant analysis

Over the past decades, Floating Offshore Wind Turbine (FOWT) has gained increasing attention in wind engineering due to rapidly growing energy demands. However, difficulties turbine maintenance will increase harsh operational conditions. Fault diagnosis techniques play a crucial role enhance reliability of FOWTs and reduce cost offshore energy. In this paper, novel data-driven fault method using regularized dynamic canonical correlation analysis (RDCCA) Fisher discriminant (FDA) is proposed for FOWTs. Specifically, overcome collinearity problem that exists measured process data, with regularization scheme, developed exploit relationship between input output signals. Then, residual signals are generated from established RDCCA model detection. To further classify type, an FDA trained different training faulty data sets. Simulations on FOWT baseline based widely used National Renewable Energy Laboratory FAST simulator carried out demonstrate feasibility efficacy detection classification method. Results have shown many salient features potential applications