将欧几里得平面或空间中各个方向上的点位方差的均值作为一种新的点位精度度量,不妨称之为点位均方向方差。从"方差"这一概念本身来看,点位均方向误差更具有"方差"蕴含的含义,能直观反映点位在各方向上平均离散状况;从可视化的角度来看,可以用点位均方向标准差为半径的圆,球或超球,近似描述出点位误差的大致分布;从概率的角度而言,比较接近误差椭圆,误差椭球或超椭球,这在扩展不确定度的描述时,不妨用相应的误差椭圆(椭球)所对应的概率值作为置信度进行描述,无须经复杂而繁琐的计算。
Multi-dimensiona sitional precision, which is error indicator always adopts Distance Mean Square Error (DMSE) when describing pcsummation of every two arbitrarilyorthogonal component error. Because the summation can not denote error value in every coordinate component, especially when the dimension is big enough. In this paper, mean of the coordinate component errors in Euclid plane or space is named as Positional Mean Direction Error (PMDE) to be another error indicator, from two-dimensional extended to three-, n-dimensional positional error. The new measure indicator has many advantages, firstly, from the concept variance angle,PMDE is much more meaningful to accord with the concept, and can reasonably represent average disperse degree of error distribution and intuitionisticly show even error in every direction; secondly, from the point of visualization view, the indicator can describe as a radius to visualize a circle, sphere and hyper-sphere, which is near to error ellipse, ellipsoid, or hyper-ellipsoid correspondingly; and thirdly from probability point of view, when describing expanded uncertainty, the probability of ellipse, ellipsoid and hyper-ellipsoid correspondingly can be used directly as confidence level, and have no need of intricate and fussy computation.