Abstract: Aiming at the location report gap of ACARS or ADS-B during global flight tracking, a flight tracking gap assessment model is established based on the ACARS and ADS-B messages. Firstly, the model reconstructs the missing points by cubic spline interpolation algorithm, then divides the point coverage area and the missing area in the evaluation area. SVM algorithm based on Gaussian kernel function is used to obtain the tracking gap location probability distribution inside the missing area. Finally, test data are used to evaluate the model performance. Results show that the assessment model is suitable for two kinds of multi-source heterogeneous data, and the accuracy is higher with shorter data period.

Key words: global flight tracking, multi-source heterogeneous data, tracking gap assessment model, tracking gap location
probability distribution