研究了模糊交互多模型算法（FIMM）和基于当前统计模型（CS）的自适应滤波算法，提出了一种基于当前统计模型的模糊交互多模型算法（CSFIMM）。该算法克服了模糊交互多模型算法精度较低和当前统计模型自适应滤波器方差调整有限的缺点。通过仿真，对所提出的算法和交互多模型（IMM）算法以及FIMM算法在估计精度和计算量两个方面进行了比较。

Based on the investigation of the fuzzy interacting multiple model algorithm and the adaptive filter based on the "current" statistical model, a extend algorithm called fuzzy interacting multiple model algorithm based on "current" statistical model(CSFIMM) is presented. This algorithm overcomes the weakness of the FIMM that it has low estimation accuracy and the limitation of the adaptive filter based on the "current" statistical model to adjust its system covariance. Comparison of its estimation accuracy and computation complexity with IMM and FIMM via a simulation scenario is given.

TN953