A variable structure-autonomous-interacting multiple model ground target tracking algorithm in dense clutter

---

Tezin Türü: Doktora

Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Elektrik ve Elektronik Mühendisliği Bölümü, Türkiye

Tezin Onay Tarihi: 2013

Öğrenci: GÖKÇEN ALAT

Danışman: MEHMET KEMAL LEBLEBİCİOĞLU

Özet:

Tracking of a single ground target using GMTI radar detections is considered. A Variable Structure- Autonomous- Interactive Multiple Model (VS-A-IMM) structure is developed to address challenges of ground target tracking, while maintaining an acceptable level computational complexity at the same time. The following approach is used in this thesis: Use simple tracker structures; incorporate a priori information such as topographic constraints, road maps as much as possible; use enhanced gating techniques to minimize the e ect of clutter; develop methods against stop-move motion and hide motion of the target; tackle on-road/o -road transitions and junction crossings; establish measures against non-detections caused by environment. The tracker structure is derived using a composite state estimation set-up that incorporate multi models and MAP and MMSE estimations. The root mean square position and velocity error performances of the VS-A-IMM algorithm are compared with respect to the baseline IMM and the VS-IMM methods found in the literature. It is observed that the newly developed VS-A-IMM algorithm performs better than the baseline methods in realistic conditions such as on-road/o -road transitions, tunnels, stops, junction crossings, non-detections.