Blade Damage Forecast Research Based on Support Vector Machine
By doing modal analysis of the whole wind turbine system, we can get the first twenty orders natural frequencies and corresponding mode shapes, By analyzing the dynamic characteristics of the blade, the weakness points of blade were fund. Under rotor rotating excitation in the normal state, mass eccentricity states and stiffness damage states of blade, the strain energy change rates (SECR) of nacelle are obtained. While based on the SECR of nacelle, the methods of strain energy change rate and support vector machine are introduced to indentify locate the mass eccentricity and stiffness damages of blade. The research show that mass eccentricity and stiffness damages at different location and in different degree can be efficiency identified and forecasted by means of support vector machine classification method.
Z. W. Ma et al., "Blade Damage Forecast Research Based on Support Vector Machine", Advanced Materials Research, Vols. 588-589, pp. 278-282, 2012