K-NN Based Wind Speed Prediction for Energy Conversion and Management

-Electricity plays an important role in human civilization. Every aspect of human life and its well being depends on the energy conversion into electricity and utilize it for optimal performance in all kinds of industries. Wind energy is one such important energy being used in the world for producing electricity. However, wind speed is not predictable and thus the wind energy conversion is not straightaway schedulable. In this context, it is a challenging problem to predict wind speed accurately. It is very important for power generation and operations. Recently Yesilbudak et al. proposed a method that predicts wind speed using k-nearest neighbor classification algorithm. Various distance measures are used to achieve wind speed prediction. In this paper we implement a prototype application using Java programming language that will take pre-recorded datasets and performs the prediction of wind speed for a short term. The empirical results are encouraging. Index Terms – Energy conversion, wind speed, k-NN classification