基于RNN的航空监视信息融合技术

he integration of aviation surveillance information is very important in the process of aviation surveillance processing. The main purpose of aviation surveillance information fusion is to use multi-source surveillance information to integrate more accurate flight target location information using various algorithms. The traditional Kalman filter-based aeronautical surveillance information fusion technology has the disadvantages of low fusion precision in the maneuvering state and repeated manual adjustment of the structural fusion model. Therefore, this paper proposes an aerial surveillance information fusion technology based on recurrent neural network (RNN), which uses the multi-source monitoring information fusion at the same time to obtain the location information of the flight target. Since the monitoring information of each monitoring source has its time stamp and the coordinate system in which it is located, it is necessary to align the monitoring information of different sources before and after the fusion. In this paper, the coordinate alignment is used for spatial alignment, and the least squares method is used to fit the function of each dimension information and time. The function is used to predict the monitoring information of each monitoring source, so as to achieve the purpose of time alignment. It is proved by experiments that the precision loss of monitoring information in space-time alignment process is small, and the fusion result error of RNN-based aeronautical surveillance information fusion algorithm is smaller than the average error of each monitoring source.