Wind Power Resource Estimation with Deep Neural Networks

摘要: The measure-correlate-predict technique is state-of-the-art for assessing the quality of a wind power resource based on long term numerical weather prediction systems. On-site speed measurements are correlated to meteorological reanalysis data, which represent best historical estimate available atmospheric state. different variants MCP more or less correct statistical main attributes by making reanalyses bias and scaling free using on-site measurements. However, neglecting higher order correlations none utilize full potential We show that deep neural networks make use these correlations. Our implementation tailored requirements in context assessment. application this method set locations compare results simple linear fit frequency distribution as well standard regression MCP, represents industrial aerodynamics. network outperforms both other methods with respect correlation, root-mean-square error distance distribution. Site assessment can be considered one most important steps developing energy project. To end, approach described regarded novel, high-quality tool reducing uncertainties long-term reference problem