Abstract:

The value of the sliding sampling interval length [N] of the fading Kalman filter is a single value, and it is difficult to achieve a balance between filtering precision and dynamics. Aiming at this problem, an adaptive fading Kalman filter based on IGGIII scheme is proposed. The filter performs the innovation chi-square detection firstly, and then constructs a value function of the three-segment sliding sampling interval length [N] similar to the IGGIII weight function to realize the adaptive optimization of the length [N] of the sliding sampling interval and improve the filtering precision. Matlab simulation results show that the adaptive fading Kalman filter based on IGGIII scheme has more accurate filtering results under stable system conditions, and it has better dynamics when system model parameters change, the filter enables a balance between filtering accuracy and dynamics.

Key words: Kalman filter, sliding sampling interval, Chi square test, IGGIII

摘要：

渐消卡尔曼滤波器滑动采样区间长度[N]的取值为单一定值，难以实现滤波精度与动态性之间的平衡。针对该问题，提出了一种基于IGGIII方案的自适应渐消卡尔曼滤波器。该滤波器进行新息异常卡方检测，通过构建一个类似于IGGIII权函数的三段式滑动采样区间长度[N]的取值函数，实现滑动采样区间长度[N]的自适应优化选取，提升滤波精度。Matlab仿真结果证明，基于IGGIII方案的自适应渐消卡尔曼滤波器在系统稳定情况下滤波结果更加精确，系统模型参数发生变化时具有更好的动态性，能够实现滤波精度与动态性之间的平衡。

关键词: 卡尔曼滤波器, 滑动采样区间, 卡方检测, IGGIII