To solve the problem of accumulated navigation errors in underwater unmanned underwater vehicles during long-term operations,applying the Informer time series prediction model,deep mining and feature extraction of UUV navigation information and the inherent relationship between calculated navigation positions,predict the trend of heading error in the future time series of the target.Conduct comparative experiments and predict trajectory visualization,the experiment shows that the Informer model has achieved good results in predicting the problem of dead reckoning errors in complex environments,suitable for correcting UUV dead reckoning errors in complex environments.
关键词
航位推算/累积误差/Informer/航迹预测
Key words
dead reckoning/accumulated error/informer/track prediction
维普
万方数据