Radar data trajectory processing based on multi-model filtering

Radiolocation and radio navigation


Sychev M. I.

Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia

e-mail: sychev@mai.ru


The aircraft coordinate and movement parameters radar estimation is discussed. Multi-model filters application is proved. The emphasis is on horizontal movement models selection. Three civil aircraft basic horizontal movement types and their parameters are defined. The first type is uniform motion, the second is standard turn, and the third is uniform speed change.

Three types of multi-model filters are described. They differ in the merge point of hypothesis for the model state vector changes, and its result implementation. The combined information from all the hypothesises is used in the first algorithm to proceed the transition to the next step. The first algorithm can be viewed as Nth Kalman filter operating in parallel. Here N is the number of models the state vector changes. The second algorithm describes state vector extrapolation for each hypothesis individually, and then their interrelation. As a result, the second algorithm has N2 extrapolation blocks and N update blocks. Hypotheses combined after the update is performed exclusively to provide information to the filter output, but is not used on the next step. The third algorithm is a N2 Kalman filter operating in parallel with the combined information for the individual hypotheses after their upgrade.

Various variants of estimation of motion models parameters, such as speed of the aircraft turn and longitudinal acceleration are considered. The different types of algorithms for these parameters estimation are presented. One of them is independent filters for these parameters estimation. The other is extended state vector describing aircraft coordinates and motion parameters.

The approach to the simplified multi-model filters design is offered based on approximation of a correlation matrix of a state vector for the specified motion model by a constant matrix.

Results of comparative simulation of the considered multi-model algorithms are presented.


  1. Eurocontrol Standard Document for Radar Surveillance in En-Route Airspace and Major Terminal Areas. SUR.ET1.ST01.1000-STD-01-01, Edition 1.0, March 1997.

  2. Yaakov Bar-Shalom, X.-Rong Li, Thiagalingam Kirubarajan. Estimation with Applications To Tracking and Navigation. John Wiley&Sons, Inc., New York, 2001.

  3. Bakulev P.A., Sychev M.I., Nguen Chong Lu. Radiotekhnika, 2004, no.1, pp.26-32.

  4. Sychev M.I., Fesenko S.V. Informatsionno-izmeritel’nye i upravlyayushchie sistemy, 2016, no.2, pp.10-18.

  5. Sychev M.I., Fesenko S.V. Trudy MAI 2015, no.83: http://www.mai.ru/science/trudy/published.php?ID=62280


mai.ru — informational site MAI

Copyright © 2000-2020 by MAI