Particle filter based attitude matching algorithm for in-flight transfer alignment

Attitude plus velocity matching transfer alignment (TA) (Rapid Alignment Prototype RAP) algorithm has the advantage of faster convergence and it does not require pre - planned lengthy manoeuvre like velocity matching algorithm, which makes it best suited for tactical missions. For large initial misalignment angles, TA problem becomes nonlinear, for which, a conventional particle filter (CPF) based TA algorithm can be used to estimate misalignment. Time varying nature as well as dependencies on external parameters like sensor measurements of state transition matrix of TA problem, makes the system behavior unpredictable and hard to model. A CPF fails in this situation, due to its inability to capture complex nonlinearity associated with the system through system dynamics, due to sample impoverishment problem. Current work addresses this scenario and proposes an evolutionary strategy based algorithm, which performs effectively in such condition, by simulating varied system dynamics through generation of multiple support points. These algorithms are designed and tested for both perfectly modeled and perturbed systems. Simulation results are presented which shows the effectiveness of the proposed algorithm.