New Matrix-weight IMM Algorithm Based on Coefficients Filters
Junchen Sha, Xing Jianping -, Gao Liang -, Liang Haozhe - · INTERNATIONAL JOURNAL ON Advances in Information Sciences and Service Sciences · 2012
In this paper, a fast interacting multiple model algorithm weighted by matrix (FMIMM) is presented based on the weighted least squares (WLS) method, which employs the steady state of the second and third order Kalman filters. And a CA model is introduced into the proposed algorithm to improve the performance when tracking a maneuvering target. Simulation results demonstrate that the proposed algorithm reduces the on-line computational burden about 20% while keeping high tracking accuracy both in position and velocity compared with the conventional MIMM algorithm, which makes the proposed algorithm a competitive alternative to conventional IMM and MIMM algorithms.