一种基于阴性选择算法的飞行数据异常值检测方法

研究了基于阴性选择算法的飞行数据异常值检测问题。针对传统的基于数据趋势进行数据异常值检测方法难以处理连续异常值的问题,提出通过提取待测参数的相关参数并在检测器编码中加入相关参数信息,使得阴性选择算法能够用于连续多个异常值的检测,算法不需要关于数据异常的任何先验知识。实验表明,采用改进编码后的阴性选择算法识别飞行数据中的异常值具有较高的精度。

A Detection Method of Abnormal Flight Data Based on Negative Selection

The method of using negative selection algorithm to detect abnormal flight data was researched.Because original detection algorithm based on data tendency could hardly detect continuous abnormal values,the detector coding method was improved by inserting information of the pending parameter′s related parameter into the coding.Then the negative selection algorithm could be used to detect continuous abnormal values without any foregone knowledge about the abnormity.The experimental results showed that the precision was high in detecting abnormal flight data based on improved negative selection algorithm.