结构随机跳变系统的自举滤波方法

BOOTSTRAP FILTERING OF SYSTEM WITH RANDOM CHANGING STRUCTURE

针对在结构随机跳变系统的滤波中存在的“基础结构失真”问题，提出了一种对系统结构的非线性、系统状态和噪声的分布形式不再限制的滤波新方法。并就该方法的性能同目前广泛应用的滤波方法做了仿真比较，仿真结果证实了该方法的有效性及其优良性能。

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.