Intelligent Prediction Method for Updraft of UAV That Is Based on LSTM Network

Abstract

Listen

Translate article

Updrafts are widely exist in the nature and with the help of updrafts, a unmanned aerial vehicle (UAV) may improve its flight time performance. Some methods have been proposed for updraft prediction such as extended Kalman filter(EKF) or unscented Kalman filter (UKF). In this paper, an prediction method based on Long-short Term Memory (LSTM) network is proposed. Firstly, a flight simulation system is developed, which is used to generate the training data for LSTM network. Then an LSTM network is developed and trained for updraft prediction. Finally, some experiments are made to compare LSTM network with traditional methods that are based on EKF and UKF. The results show that LSTM network has a substantial advantage in terms of the prediction accuracy and convergence rate.