一种复合型自适应Kalman滤波器的仿真研究

Study on the Simulation of a Compound Adaptive Kalman Filter

本文探讨带模型误差和噪声统计特性误差背景下的自适应Kalman滤波问题。通过引入虚拟噪声补偿技术和实际不发散新算法相结合的方法,提出了一种复合型自适应Kalman滤波器的新方案。不同类型的数值仿真例子表明了本方案的有效性。为便于进一步分析,研究新方案的性能,我们开发了用True—BASIC语言编制的计算机通用仿真程序。

This paper deals with the problem of adaptive Kalman filtering against the background with model errors and errors of noise statistical characteristics. In accordance with the adoption of fictious noise compensation technique combined with actual nondivergent computation, a new project of compound adaptive Kalman filter is presented. Several examples of different kinds of numberical simulations show that the new project given in this paper is effective.For the purpose of further analysis and study of the performances of the new project, we have exploited a common simulation software with True BASIC.