Modelling Out-Of-Sequence Measurements Using a Grey Relational Anaysis and Copulas Hybrid

One primary concern of researchers, particularly those dealing with multisensory target tracking and filtering, is effectively dealing with out-of-sequence measurements (OOSM). Recently, the use of Kalman filters has proven to be of great practical value in solving a variety of OOSM problems including multi-target tracking prediction. In this paper we argue that delayed and existing measurements are typically correlated and could be described by a joint distribution. We further arugue that the uncertainty of the measurements could be modelled using grey relational analysis (GRA). Thus, the proposed approach deals with the uncertainty of information and combines it with copula in order to model OOSM. Benchmarking results on simulated datasets show the use of GRA coupled with copulas as more robust to handling OOSM as compared to existing methods.