Design of an exponent implemented model-set adaptive IMM algorithm (EAIMM)

作者：Sha Junchen; Xing Jianping^*; Wu Hua; Liang Haozhe

来源：Journal of Computational Information Systems, 2012, 8(15): 6523-6535.

• EAIMM
• Interacting multiple model
• Model-set adaption
• Novel-IMM
• Selection probabilities

In this paper, an exponent implemented model-set adaptive IMM algorithm (EAIMM) is proposed, which realizes the model-set adaption without auxiliary information by designing exponential selection probability for each model set in EAIMM algorithm. The proposed algorithm not only resolves the underflow problem of standard Novel-IMM, but also outperforms the single model-set based IMM estimator that covers only a few system modes or parameters. Moreover, the exponential selection probabilities are more sensitive to system mode changes, which improves the convergence speed of EAIMM, thus the proposed algorithm is very promising for tracking maneuvering target.

• 出版日期2012