Open AccessProbabilistic forecast for aggregated wind power outputs based on regional NWP data
Author(s) -
Wang Zhao,
Wang Weisheng,
Liu Chun,
Wang Bo,
Feng Shuanglei
Publication year - 2017
Publication title -
the journal of engineering
Language(s) - English
Resource type - Journals
ISSN - 2051-3305
DOI - 10.1049/joe.2017.0587
Subject(s) - wind power forecasting , probabilistic logic , wind power , numerical weather prediction , kernel density estimation , probabilistic forecasting , euclidean distance , meteorology , wind speed , computer science , principal component analysis , kernel (algebra) , electric power system , power (physics) , mathematics , statistics , geography , artificial intelligence , engineering , physics , quantum mechanics , electrical engineering , combinatorics , estimator
With the high level of wind power penetration, the big users in power systems, like system operators and market traders, demand the aggregated wind power forecasts for a specific region. A method searching for the similar wind characteristics in the numerical weather predictions (NWP) data is proposed here. The similarity is defined by a weighted Euclidean distance, and the dimensionality is reduced by means of principal component analysis. The probabilistic forecasting results are constructed by the selected samples with the proposed distance‐weighted kernel density estimation method. A case study of 28 wind farms in the East China is provided to evaluate the performance. The proposed method is easy to apply and performs well in practice.