In engineering, the wind excitations acting on the offshore wind turbine (OWT) structure cannot be obtained directly by the measured method. The traditional load simulation way may lead to poor accuracy because of the deviation between actual operational conditions and simulation environment parameters or load coefficients, which are always selected based on load spectrums, small scale model tests and engineering experience. To solve this project problem, an equivalent wind load optimization model and parameter identification process of OWT was established under the standstill conditions. Afterwards, two key load parameters in load model respectively called constant drag coefficient and unsteady lift coefficient can be obtained through recursive call on numerical model and feedback analysis using particle swarm optimization (PSO) algorithm based on measured vibration data. The applicability and accuracy of proposed calculation process was verified through an equivalent numerical model of OWT considering different particle position and wind speed. Further, the optimization identification on equivalent wind load and the drag and lift coefficients of one OWT was achieved based on measured data. It is illustrated that the identified drag coefficients show a drop trend from 0.629 to 0.154, consistent with the parameter range of 0.20–0.60 obtained from model experiments, before the wind speed reaches 6.0 m/s and then rise obviously with the increase of wind speed. On the contrary, the range of optimized lift coefficients form 0.133–0.480, which is as familiar as the experimental parameter range of 0.10–0.30, reflects a certain increase trend. Finally, the comparison between identified wind excitations and calculated loads based on the measured strain data explain that the proposed load optimization identification process have a good credibility on actual OWT.
Read full abstract