Wind resource estimation using a MCP method based on deep neural networks

A novel implementation of neural networks for a measure-correlate-predict (MCP) technique is being introduced. This approach represents a state-of-the-art machine learning procedure making use of sophisticated methods to limit generalisation errors. Moreover, the implementation is fully tailored to the requirements of MCP methods in the context of wind resource assessment. We show the application of this method to a set of different locations and compare the results to a simple linear fit to the wind speed frequency distribution as well as to a standard linear regression MCP. The neural network based MCP outperforms both other methods with respect to correlation, root-mean-square error and the error in the wind speed frequency distribution. Hence, this approach can be regarded as a novel, high-quality tool reducing uncertainties in the long-term reference problem using siteassessment products and tools. It therefore tackles one of the most important steps developing a wind energy project.