卡尔曼滤波器、H∞滤波器和鲁棒混合Kalman/H∞滤波器的比较

摘要

The Kalman filter, H∞ filter and robust mixed Kalman/H∞ filter are presented for discrete time-varying systems subject to model uncertainty over a finite horizon and a infinite horizon. In the simulations, it confirms that the robust mixed Kalman/H∞ filter performs better than Kalman filter and H∞ filter under the inaccurate model. Furthermore, the influence of the parameter to the filter is considered.%本文阐述了有限时间以及无限时间上基于离散不确定模型的卡尔曼滤波器，H∞滤波器以及鲁棒混合卡尔曼/H∞滤波器的算法。通过仿真，证实了在不确定系统中混合鲁棒滤波器性能优于卡尔曼滤波器和H∞滤波器，并且分析了混合鲁棒滤波器中参数设置对算法的影响。