AFMM solution to the benchmark radar tracking problem
Mattias J. L. Carlsson, Mika V. Laaksonen · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 1999
The paper describes a new approach for an adaptive update- rate-tracking algorithm for phased array radars. Since phased array radars have the ability to perform adaptive sampling of the target by radar beam positioning, proper control of radar has the potential for significantly improving many aspects associated with tracking of multiple maneuvering targets. An adaptive multiple model filter for tracking maneuvering targets has been designed for applications to the Benchmark Tracking Problem (BTP). The first Benchmark Problem has previously been solved with (alpha) - (beta) -filters, standard Kalman filters and two and three model IMM filters. The tracking algorithm used in this paper is an Adaptive Forgetting through Multiple Model filter (AFMM). AFMM is especially suited for tracking of dynamic system with jumping and rapidly changing parameters. It can be viewed as a particular way of implementing adaptive gains or adaptive forgetting factors for tracking. Two motion models have been used: a constant velocity and a 3D turning rate. Preliminary result indicate promising capabilities of the filter in solving the BTP. This paper presents results for the BTP solved with an AFMM filter.