Fusion Tracking Algorithm Based on Stochastic Approximation

A practical fusion algorithm for tracking maneuvering target based on centralized structure of multi-sensor is proposed. This algorithm is implemented with two filters and state fusion, together with the current statistic model and adaptive filtering. Firstly, the fusion weighting coefficients are obtained using the stochastic approximation theory, a suitable method of estimation measurements noise variance is developed based on fuzzy inference. Two adaptive unscented Kalman filters with current statistical model are derived in parallel, and fuzzy rule is designed. For the target trajectories of maneuvering and non-maneuvering, computer simulation results show that the fusion algorithm tracks very well maneuvering target over a wide range of change of measurement noise and maneuvering, the algorithm has the robust performance of approach, and it is suitable for practical engineering system