Prediction of offshore wind power monopile scour depth under waves and current conditions
The scouring issue of monopile foundations in offshore wind farms remains a pivotal challenge impeding development of the industry.Precise prediction of the scour depth for monopiles is of paramount importance.In this paper,a predictive model was established for the equilibrium scour depth under various wave-current conditions based on the M5'model tree algorithm.Initially,a dataset from physical model experiments on monopile scour was acquired,and key dimensionless parameters influencing scour prediction were identified.Subsequently,several predictive models were established based on the M5'model tree algorithm,correlating various input parameters with output parameters.The models were evaluated using statistical metrics such as the correlation coefficient,consistency index,dispersion index,and bias.These statistical parameters served as benchmarks to compare and assess the performance of each predictive model against existing prediction formulas.The research findings indicate that due to current,the optimal combination of input parameters for predicting outcomes includes relative water depth,Froude number,relative median grain size,and Reynolds number.These parameters yield the best predictive performance.Conversely,the Keulegan-Carpenter number and the scattering parameter emerge as the most critical control parameters for predicting wave-induced scour depth.The ideal input parameter combination comprises the Froude number,Reynolds number,and relative median grain size in combined waves and current.These research outcomes can guide the prediction of scour depth around monopiles,enhancing the accuracy of such predictions.
monopilelocal scourM5'model tree algorithmequilibrium scour depth
万方数据
维普
