A Sugeno-type neuro-fuzzy adaptive filter for online maneuvering target tracking

A neuro-fuzzy adaptive filter employing Sugeno-type If-Then rules with online structure and parameter learning capability is developed for the online maneuvering target tracking problem. The maneuver is considered as an inherent part of the target dynamics; this makes the system nonstationary. To show the performance of the proposed filter, we use the same dynamic model of the MA-2d radar for comparison with the interacting multiple model techniques. The performance of the designed filter was evaluated by Monte Carlo simulation over a test trajectory. The results show the effectiveness of the proposed filter.