Wind speed forecast for the stratospheric airship by incremental extreme learning machine
Guangyu Xu, Shaoping Shen, Sun Jietao · 2017
This paper investigates the wind speed forecast for the stratospheric airship fixed over a geo-location because the wind speed forecast is a key challenge for the airship station-keeping control. In view of the wind speed series which changes with the time and space and shows the non-linear and non-stationary characteristics, this paper put forward a kind of adaptive model based on Incremental extreme learning machine(I-ELM). Firstly, in order to avoid the random selection of input dimensionality of the I-ELM, the phase space of the wind speed is reconstructed. Secondly, search and optimize the model parameter of the I-ELM to improve the generalization performance of the I-ELM. Finally, the prediction model is established by using the optimized I-ELM, and the results of the online prediction of short-term wind speed are obtained. The simulation result verifies that the proposed forecasting model is able to excavate the wind speed series features effectively and has relatively high prediction accuracy.