论文部分内容阅读
本文提出了未知噪声统计的非线性系统中新的自适应推广的Kalman滤波算法。作者提出了用虚拟时变噪声统计,补偿线性化模型误差的新思想。在本文中,作者指出了文献[3]中,用Sage和Husa的常值噪声统计估值器来估计虚拟噪声是不合理的。另外,即使原非线性系统的噪声统计是零均值的,但线性化的模型的噪声统计一般是非零均值的。两个数值模拟例子说明了本文方法的有效性。
In this paper, a new adaptive Kalman filtering algorithm is proposed for nonlinear systems with unknown noise statistics. The author proposes a new idea to compensate the error of the linearized model by using the statistical time-varying noise statistics. In this paper, the author points out that in [3], it is not reasonable to estimate the virtual noise using Sage and Husa’s statistical estimators of constant noise. In addition, even if the noise statistics of the original nonlinear system are zero mean, the noise statistics of the linearized model are generally non-zero mean. Two numerical examples illustrate the effectiveness of the proposed method.