Multiple model dynamic estimation approach in target tracking and localization

Abstract

The objective of this thesis is to conduct further research in Multiple Model (MM) estimation approach. More specifically, we conduct research on model and algorithm enhancement and new application exploration. In model and algorithm enhancement, we focus on model design for accelerated target tracking and Sojourn-Time-Dependent semi-Markov (STDM) switching for dynamic MM estimation. As for application exploration, the MM approach is introduced to solve some difficult problems, which include Radio Frequency (RF) localization using distorted antenna beam patterns and indoor contaminant source localization with limited number of sensors.