Estimation and characteristic analysis of aggregated generation of geographically distributed wind farms

A good understanding of the aggregated stochastic characteristics of available wind generation is very important for secured and economic system operation in market environment. This paper focus on estimating and analyzing the aggregated generation of geographically distributed wind farms. A dynamic system model is formulated to describe the relationship between atmospheric and near-surface wind fields of the individual wind farms. A recursive algorithm based on Kalman filter theory is developed to update the parameters from NWP and to estimate the generation of individual wind farms based on their geographical locations, the wind dynamics in atmosphere, and the initial temporal and spatial conditions of the wind fields. The stochastic characteristics of the total available wind generation can then be estimated as the aggregation of the individual generation estimation. The actual data validates the assertion that the aggregated wind generation of distributed wind farms is less volatile than that of a single wind farm.