A STARMA Model for Wind Power Space-Time Series

This paper had analyzed and modeled the wind power characteristic from the perspective of space-time series. Firstly, measured data of wind power had been analyzed for the coupled spatial-temporal correlation. Then, a spatial relation matrix, which was used to describe the location of wind farms, had been embedded into the Space-Time Auto Regressive Moving Average (STARMA) model in order to reflect the spatial-temporal correlation of multi-dimensional wind power series. Simulation results showed that this model has restored not only temporal autocorrelation, but also time shifting characteristic of spatial correlation of the original wind power series, which essentially reflected the coupled spatial-temporal characteristic of real wind power series. This model can be used to produce huge amount of simulated wind power data, which have same characteristics with real wind power data, and can provide basics for the planning and operation of wind power integration systems.