Forecasting wind power generation patterns based on SOM clustering

Due to incontinent use of fossil fuels all over the world, it comes to be exhausted and also causes serious environmental pollutions and global warming. Therefore, people begin to find renewable energy which is clean, no limit and reproducible. Among several renewable energies, wind power is the most promising one which can be connected to the electric power system. However, it is very important to predict the wind power generation patterns in the electric power system to balance the load and generation. In this paper, we propose a framework to predict the wind power generation patterns with classification models. This framework consists of the following steps: (1) data preprocessing to handle noise data, missing values, (2) assignment of class labels to wind power generation patterns using SOM clustering, (3) classification model construction to predict the wind power generation patterns. The experiment result shows that the rules from decision tree are simple and easy to interpret. And it is possible to predict wind generation patterns.