中国空间科学技术2016，Vol.36Issue(5)：18-24,7.DOI:10.16708/j.cnki.1000-758X.2016.0051
遥测伪周期时间序列子序列异常检测算法
Detection algorithm of abnormal sub-sequence in pseudo periodic time series of telemetry
摘要
Abstract
In order to detect the abnormal sub-sequence in the pseudo periodic time series of spacecraft telemetry,a sub-sequence segmentation algorithm on the wide range of maximum cycle window was proposed based on the two conventional segmentation methods.Then,an anomaly sub-sequence detection method based on an optimized sequence model was built by using the double quantile ranges criterion to set the threshold of the distance detection.The satellite experimental results show that the method can effectively detect the abnormal sub-sequence of satellite telemetry data.关键词
异常检测/卫星遥测/伪周期/时间序列/异常子序列/2次四分位距准则Key words
anomaly detection/satellite telemetry/pseudo periodic/time series/abnormal sub-sequence/double quantile ranges criterion分类
信息技术与安全科学引用本文复制引用
杜莹,李四虎,胡绍林,章乐,王飞..遥测伪周期时间序列子序列异常检测算法[J].中国空间科学技术,2016,36(5):18-24,7.基金项目
国家自然科学基金(61473222) （61473222）
