Locations: PIMENT laboratory, University of La Reunion, Reunion Island
School of Mathematics and Statistics, University of South Australia, Adelaide - Australia
The large-scale integration of intermittent renewables (solar and wind) inside electricity networks presents a problem for the supply-demand balance and for the stability of the grid. The forecasting of the energy production from these renewables is investigated in order to solve these problems and to increase the rate of renewables in the grids. The challenge lays both on technical and economical aspects.
Réuniwatt (http://www.reuniwatt.com) has developed a unique and patented solar power production forecasting named Soleka.
In the project SOLFIN, funded by the French Agency of Energy (ADEME), REUNIWATT company and PIMENT laboratory plan to explore financial and economic mathematics in order to improve the forecast of power production of intermittent renewables. Indeed, variations of the solar irradiance seem similar to fluctuations of shares and securities (large amplitude, clusters with high or low volatility).
The aim of the position is to test these mathematical models in order to develop a model of forecast of the short-term variability of the solar irradiance. A preliminary work should be the bibliography of methods and models used in the financial and economic fields.
• Time series models ARMA, ARIMA, ARMAX, etc.
• ARCH/GARCH models that take into account the heteroscedasticity of the process in order to forecast the volatility,
• Chaos models in order to assess the foreseeable nature of irradiance times series and the number of associated dimensions,
• Stochastic volatility models and models based on stochastic differential equations,
• The CBOE volatility index (VIX),
Required expertise: PhD degree or postdoc position in the field of applied mathematics to financial and/or economic studies.