摘要
近年来,我国的隧道工程建设取得了巨大的成就,隧道施工过程中存在诸多不确定性的风险因素,造成施工事故频繁发生,对工程质量和风险控制提出巨大的挑战。针对工程监测数据进行异常分析是解决该问题的重要途径。应用新的智能算法进行异常诊断为工程运营提供了新的风险管理方法和思路。借鉴危险理论在异常诊断方面的优势,提出改进的DCA(dendritic cell algorithm)算法对时间序列数据进行异常检测,取得了良好的检测效果,对不断完善工程风险管理方法具有重要意义。
In recent years,the tunnel engineering construction has achieved great success in China.A lot of uncertain risk factors exist in the tunnel engineering construction and result in construction accidents frequently.It poses great challenges to project quality and risk control.To solve this problem,anomaly analysis on engineering monitoring data is an important way.The new intelligent algorithm used for anomaly detection provided new risk management methods and ideas for engineering operation.This paper proposed an improved dendritic cell algorithm to perform the anomaly detection for time series data in view of the fact that danger theory has great advantages in anomaly detection.And the algorithm can obtain good detection results on the anomaly detection for time series data.It is of great significance to improve engineering risk management.
作者
徐伟
孙向阳
邸泽雷坤
XU Wei;SUN Xiangyang;DIZE Leikun(SILC Business School,Shanghai University,Shanghai 201899,China)
出处
《四川职业技术学院学报》
2020年第4期154-160,共7页
Journal of Sichuan Vocational and Technical College
基金
国家自然科学基金资助项目(71871133)。
关键词
人工智能
时间序列
异常检测
危险理论
免疫算法
信息熵
artificial intelligence
time series
anomaly detection
danger theory
immune algorithm
information entrop
