Sequential Monte Carlo methods for multi-sensor tracking with applications to radar systems

The aim of this thesis is to explore specific sequential Monte Carlo (SMC) methods and their application to the unique demands of radar and bearing only tracking systems. Asynchronous radar networks are of special interest and a novel algorithm, the multiple imputation particle filter (MIPF), is formulated to perform data fusion and estimation using asynchronous observations. Convergence analysis is carried out to show that the algorithm will converge to the optimal filter. Simulations are performed to demonstrate the effectiveness of this filter. Next, the problem of multi-sensor bearing only tracking is tackled. A particle based tracking algorithm is derived and a new filter initialization scheme is introduced for the specific task of multi-sensor bearing only tracking. Simulated data is used to study the efficiency and performance of the initialization scheme.

Sequential Monte Carlo methods for multi-sensor tracking with applications to radar systems

The aim of this thesis is to explore specific sequential Monte Carlo (SMC) methods and their application to the unique demands of radar and bearing only tracking systems. Asynchronous radar networks are of special interest and a novel algorithm, the multiple imputation particle filter (MIPF), is formulated to perform data fusion and estimation using asynchronous observations. Convergence analysis is carried out to show that the algorithm will converge to the optimal filter. Simulations are performed to demonstrate the effectiveness of this filter. Next, the problem of multi-sensor bearing only tracking is tackled. A particle based tracking algorithm is derived and a new filter initialization scheme is introduced for the specific task of multi-sensor bearing only tracking. Simulated data is used to study the efficiency and performance of the initialization scheme.