利用图论方法进行多不良数据检测与辨识

Detection and Identification of Multi-bad Data Using Graph Theory

本文研究了不良数据检测判据及可信量测判据，用图论的方法推导了对于给定量测系统中量测量之间的线性相关表达式（即回路矩阵），并充分利用了量测量问的相互核核作用，提出了一种新的多不良数据检测与辨识的方法，使检测相关多不良数据的可靠程度大大提高。

This paper discussed the criterion about bad data detection and convincible measurement. With the graph theoretic methed .derived the expressions (i. e. circuit matrix) of 1inear dependance between the measured quantities for a given measuring system. Fully utilized the condition to check between the measured quantities- A new method is put forward about multiple bad data detection and identification. There fore, the reliabi1ity to detect mu1tiple bad data of linear dePendance is even more increased.