Improving tracking of maneuvering target in cluttered environment

The traditional algorithm-Interactive Multiple Models Probabilistic Data Association (IMMPDA)-for tracking maneuvering target in cluttered environment suffers from the shortcoming of low tracking precision. Going beyond the improvement by Pan et al^[6] of IMMPDA algorithm, we put forward a novel algorithm--Interactive Multiple Models Adaptive Probabilistic Data Association (IMMAPDA). Our novel IMMAPDA algorithm, whose four subtopics are the calculation of interactive/mixed probability, adaptive Kalman filtering, the updating of model probability, and estimation of state and covariance; under the four subtopics we give respectively appropriate mathematical equations, which can all be found in the open literature but, when we put them together, can form the mathematical foundation of our algorithm. Finally we give a numerical example that gives the actual trajectory of the maneuvering target. These simulation results indicate preliminarily that our novel IMMAPDA algorithm has a better performance than IMMPDA with decreased computational burden for tracking maneuvering target in cluttered environment. The RMSEs (root mean square errors) of position and velocity are decreased by about 32% and 25% respectively compared with the traditional algorithm.