Adaptive Variable Structure Multiple Model Filter for High Maneuvering Target Tracking

Abstract

A turn rate based adaptive variable structure multiple model algorithm is put forward to improve tracking accuracy for high maneuvering target. In the basis of switching grid interactive multiple model (SGIMM) scheme, supporting digraph is revised according to an improved estimated turn rate. This algorithm overcomes the model-set mismatch problem caused by inaccurate switching of supporting digraph. Simulation shows this algorithm improves tracking performance especially for high maneuvering targets.