New Progress in Wind Prediction Based on Nonlinear Amendment

Abstract

In recent years, large-scale wind power integration on electric system gradually becomes to be a major trend to the development of wind power industry. Thus, high precision wind speed and power prediction technology is urgently needed. Being different from traditional wind prediction models that largely rely on various numerical methods, this paper considers the dynamical essence features of atmospheric motion. A brand new Lorenz disturbance prediction model, which is based on wavelet neural networks (WNN), is proposed and called LSWNN short-term wind speed prediction model. Compared with the results of WNN model, LSWNN model is more accurate for the actual wind speed distribution forecasting. In this article, the research not only has important theoretical value on analyzing atmospheric nonlinear motion process, but also has profound engineering guidance in wind speed prediction and wind energy resource exploitation.