Statistical modelling of wind energy resource

  • Supervisor: Dr Wolf-Gerrit Früh (Heriot-Watt University), Stan Zachary (Heriot-Watt University), Stuart Galloway (Strathclyde University)
  • Sponsoring company: SgurrEnergy 
  • PhD Student: Christina Skittides
  • This student has graduated and the project is complete.

The aim of the project is to develop a robust method to predict the wind resource at a proposed site from a brief measurement campaign at that site complemented with a much longer dataset from a nearby site.   Since we do not want to impose arbitrary constraints such as linear correlation between the two site, we explore techniques from Dynamical Systems Theory in which the common phase space attractor of the two sites is reconstructed using Takens' Time delay method and optimised by a singular value decomposition of that delay matrix.

This method will enable us to provide a nonparametric description of the two sites as a coupled system when the time-aligned parts of the wind measurements at the two sites are used to construct the attractor.   Applying the mapping found in the attractor construction then can be used to map the longer available data record from only one site to onto to the coupled system and thereby predict the resource at the other site.