1. 基于小波重构 -Autoformer 的无人机融合空域饱和流量预测.
- Author
-
乔英聪, 马昕, 陈相佐, and 马熊
- Subjects
- *
TRAFFIC flow , *MODEL airplanes , *FORECASTING , *NOISE - Abstract
To achieve the prediction of UAV saturation flow in fusion airspace, this paper proposed a UAV saturation flow prediction method based on wavelet reconstruction-Auto former(WR-Auto former). Initially, traffic flow data was decomposed using wavelet transformation to mitigate the impact of noise and highlight data characteristics. Subsequently, it utilized the Auto former model's deep decomposition mechanism and autocorrelation mechanism for foundational prediction. Considering the key factors affecting saturation flow, it introduced three UAV traffic correction coefficients. Finally, it outputted the UAV saturation flow prediction for fusion airspace by combining the UAV saturation flow calculation method. Upon verification, the WR-Auto former model has reduced the average absolute error and mean squared error by 13% to 48% in 48 h and 96 h forecasts, and the predicted saturation flow has increased by 36% to 38% compared to the current state. The experimental results prove that the proposed model can achieve accurate predictions and enhance the saturation flow of UAV in fusion airspace, while the vertical separation of UAV meets the safety requirements for class A aircraft. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF