Torsional Vibration Characteristics of Wind Turbine Gear Systems Based on Inherent Randomness

Wind turbines are widely used to produce renewable energy to alleviate the energy crisis. However, failures often occur in such systems which cause efficiency decreases and economic loss. Robust analyses and designs as well as effective condition monitoring play an important role for the safe operations of wind turbine gear systems (WTGS). To this end, the inherent randomness of the WTGS should be considered and its effects on the vibration characteristics need to be properly evaluated. In this investigation, the dispersions in material properties and dynamic features in a WTGS are studied by using the regression-based Polynomial Chaos Expansion (PCE). For deterministic simulation, the ODE45 solver in Matlab is employed. Stochastic characteristics under different operation conditions with uncertainty will be discussed. It is found that randomness has obvious influence on the torsional vibrations of the WTGS and cumulative propagation of the uncertainty may lead to unexpected dynamic behaviors.