摘要
在将误差谱度量推广应用到动态系统滤波器的评估时,整个时间轴上将绘出一个三维图,不直观且不利于分析.为了解决这一难题,提出一个新的度量—动态误差谱,并给出了三种不同应用背景下的计算形式.其中一种采用了几何平均的形式因而可以给出较为"公正"的评估结果,可直接应用于动态系统.接着从新的角度把这一度量应用于对四种非线性滤波算法的评估,仿真结果验证了动态误差谱这一新度量的有用性和有效性.
When applying error spectrum measure to performance evaluation for filtering of dynamic systems, it will plot a 3D figure over the total time span, which is not intuitive and difficult to be analyzed. In this study, to overcome its drawbacks, a new metric, dynamic error spectrum, is proposed to summarize the ES curve. Three forms under different application back-grounds are given, one of which is evenly taking into account both good and bad behaviors of an estimator and so can provide more impartial evaluation results. It can be applied to a variety of dynamic systems directly. Then from new perspectives, four nonlinear filters are chosen to illustrate the superiority of the metric. Simulation results validate its utility and effectiveness.
出处
《自动化学报》
EI
CSCD
北大核心
2014年第11期2650-2653,共4页
Acta Automatica Sinica
基金
国家自然科学基金创新研究群体科学基金(61221063)
国家自然科学基金(61074176)~~
关键词
非线性量测
目标跟踪
性能评估
动态误差谱
几何平均
Nonlinear measurements
target tracking
performance evaluation
dynamic error spectrum
geometric mean
