Floating wind turbines may encounter severe situations because of harsh environments. Higher cost of repair and maintenance of floating wind turbines have led researches to focus on damage detection methods that can prevent sudden failures. This paper presents an applicable method of damage detection and structural health monitoring for floating wind turbines based on the autoregressive moving average (ARMA) model and fuzzy classification. First, the dynamic model of a spar type floating wind turbine is constructed, by which the time responses of each degree of freedom of the system are acquired. With the system’s nonlinearity included, the intrinsic mode functions are obtained for the response signal. The Hilbert–Huang transform is applied and the appropriate measured signal for each degree of freedom is chosen for the ARMA modeling. In order to evaluate the proposed method, the ARMA parameters are first estimated for the undamaged condition then assumed damages are injected to the model and the ARMA parameters are once again estimated for the damaged condition. These parameters are considered as inputs for the fuzzy classification method. After training the system using the assumed damaged and undamaged conditions, the proposed method is simulated. Furthermore, the effect of measurement noise on the success rate is investigated. The results show that, in the presence of noise, the proposed method is able to identify the damage location and severity of mooring lines with acceptable success rate.
A Novel Parametric Modeling Method and Optimal Design for Savonius Wind Turbines
