一种创新的生成式行车异常检测方法  

An Innovative Generative Method for Traffic Anomaly Detection

作　　者：汤文韬 TANG Wentao(College of Electronics and Information Engineering,Tongji University,Shanghai,China,201804)

机构地区：[1]同济大学电子信息与工程学院,上海201804

出　　处：《福建电脑》2021年第11期27-30,共4页Journal of Fujian Computer

摘　　要：据不完全统计,全世界每年因交通事故丧生的人口数量达到130万人。对于行车异常的检测可以提醒车主及时注意车况,对交通安全具有重要意义。本文借助XGBoost分类器,采用有监督的方式,结合打分函数,设计了一种生成式的行车异常监测算法。该算法通过对原始数据进行清洗和构建新特征、基于此设计异常检测方案,标记出异常车速,从而对于异常情况进行提醒,对于保证行车安全具有重要意义。According to incomplete statistics,the number of people dead in traffic accidents in the world reaches 1.3 million every year.The detection of driving abnormalities can remind owners to pay attention to the vehicle condition in time,which is of great significance to traffic safety.In this paper,with the help of xgboost classifier and the way of generating model,the scoring function is designed,and an unsupervised traffic anomaly monitoring algorithm is designed.The algorithm cleans the original data and constructs new features,designs an anomaly detection scheme based on this,and marks the abnormal speed,so as to remind the abnormal situation,which is of great significance to ensure driving safety.

关 键 词：XGBoost 行车数据 异常检测 交通安全 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]