不确定动态系统的两种递推估计性能比较

对于噪声统计信息等不确定的动态系统的状态估计问题,人们提出了多种符合实际应用要求的递推算法,其中推广的遗忘因子递推最小二乘（EFRLS）和H∞滤波是两种有效估计的方法。对这两种估计的性能进行了比较,用计算机模拟实例说明了EFRLS方法在多种性能指标下较H∞滤波更优,为实际应用提供了有益的指导。

Performance Comparison of Two Recursive State Estimators for Uncertain Dynamic Systems

There are some recursive algorithms for state estimators in dynamic systems with uncertainty such as unknown noises, in which the extended forgetting factor recursive least squares （EFRLS） and H∞ filtering are two efficient approaches to estimating unknown state. The performance comparison of the above two methods is given in paper. Some examples prove that EFRLS has more advantages than H∞ filtering in the several aspects, and is suitable for the practical applications.