关键词: 结构随机跳变系统, 自举滤波, 基础结构失真

Abstract:

Aimed at the question of “basic structure distortion” existing in nonlinear/non Gaussian estimation of stochastic systems, a new filtering method is presented that fits to systems with random changing structures. The method no longer restricts the structure nonlinearity and distribution style of the system state and noise. Simulation for the algorithm′s characteristic is conducted, which proved this method effective.

Key words: systems with random changing structures, bootstrap filtering, basic structure distortion