论文部分内容阅读
为了解决在实际情况中噪声的统计特征信息经常残缺甚至完全未知的问题,在量测噪声的协方差未知的前提下,提出了一种方法解决含有马尔可夫参数线性离散系统的估计问题.在交互多模型算法的基础上融合H_∞滤波器构造估计器,得到交互过程中必需的噪声统计特征,进而对目标的状态进行估计.该方法不但解决了噪声信息缺失问题,而且保留了交互多模型算法的优异性能.通过与单模型H_∞滤波进行比较,试验仿真结果证实了该算法的有效性.
In order to solve the problem that the statistical information of noise is often incomplete or even unknown in practice, a method is proposed to solve the problem of linear discrete system with Markov parameters under the condition that the covariance of measurement noise is unknown. Based on the interaction and multi-model algorithm, the H_∞filter is constructed to construct the estimator to obtain the necessary noise statistical characteristics during the interaction and to estimate the state of the target. This method not only solves the problem of missing noise information, but also retains the interactive multi-model The performance of the proposed algorithm is compared with that of single-model H_∞ filter. The simulation results show the effectiveness of the proposed algorithm.