The real-time monitoring on the aircraft is realized by feeding back flight state data at several moments to Beidou land navigation station. The states of the aircraft at present and subsequent moments are needed to be predicted. Based on the studies of prediction and estimation, the flight state of the aircraft through Kalman and particle filter methods, the idea that dynamic weight average of multi-point state data adopted to obtain the flight state data of the aircraft at present and subsequent is presented. And it is displayed on digital earth in real-time to realize aircraft real-time monitoring and a better effect is obtained. The average relative error is 0.03% after predicting extracted aircraft real data and performing algorithm comparison experiment.