Trajectory reconstruction algorithm based on fixed-lag filter-smoother

To reconstruct the practical trajectory via effectively using flight data measured by radar for projectiles, reconstruction model with state equations and observation equations is established. Taking account of the inherent model nonlinearity and potential non-Gaussian distribution of reconstruction states, a one-step fixed-lag filter-smoother algorithm is proposed as the estimation method combined with Bootstrap particle filtering. Simulation results show that the proposed Monte Carlo smoothing algorithm can achieve much more accurate estimates than the Bootstrap particle filtering and unscented Kalman filter. Consequently, the proposed algorithm provides a novel effective estimation approach to trajectory reconstruction.