报告时间：11月21日（星期四）15:00

报告地点：信息楼自动化学院310报告厅

报 告 人：胡文凯，自动化学院特任教授

报告题目：Patten matching of alarm flood sequences in complex industrial facilities

内容简介：One most difficult issue in industrial alarm management is the presence of alarm floods, which is usually caused by propagation of abnormalities and has been revealed to be the leading cause of many industrial accidents. The presence of alarm floods makes the alarm systems partially or completely fail in assisting operators, and thus, deteriorate the situation and compromise process safety. This talk is concentrated on alarm flood analysis and presents how alarm floods can be solved with pattern matching techniques. An abnormal event usually causes a series of alarms in a chronological order, and the same type of abnormalities usually lead to similar alarm flood sequences. Thus, by comparing an incoming alarm sequence with potentially similar alarm floods in the historical database, it is possible to achieve an early warning of abnormalities, predict the cause of the incoming alarm flood, and take proactive operational actions to prevent the occurrence of an alarm flood and its negative consequences. The effectiveness of the methods is demonstrated by applications to real industrial facilities.

报告人简介：胡文凯，教授、博士生导师。于2010年和2012年分别在武汉大学获得学士和硕士学位，2016年在加拿大阿尔伯塔大学电气与计算机专业获得博士学位，2016年至2018年在阿尔伯塔大学从事博士后研究员工作。主持国家自然科学基金青年项目和湖北省自然科学基金青年项目各一项，参与加拿大自然与工程研究基金两项。发表论文三十余篇，其中在IEEE Trans. Industrial Electronics、IEEE Trans. Control Systems Technology和Control Engineering Practice等SCI期刊发表论文十余篇。获得地大百人、第30届CPCC张仲俊院士优秀论文奖等荣誉和奖励。主要研究方向为复杂工业系统数据挖掘和智能监控报警方法等。