HEALTH MONITORING FOR LOGIC CONTROL SYSTEM
This paper presents a real time health monitoring system for monitoring a machine and its components that are driven by a logic controller. The machine movements of a logic control system normally occur in a certain sequence and rhythm. By capturing the changes in the sequencing or timing of these movements, potential anomalies relevant to its components and actuators can be detected. The signals can be directly collected from controller I/Os and processed into a series of discrete events. A Timed Petri Net is used to model the events and therefore, any anomalies can be identified. In addition, a valuable rule is introduced to further improve diagnostic capability and eliminate false alarms that may be caused by normal variations. The system was first applied to monitor the health of an Automatic Tool Changer that is controlled by a Programmable Logic Controller (PLC). The results show that machine anomalies can be completely detected and the false alarms were significantly minimized.
Programmable Logic Controller Predictive Maintenance Timed Petri Net
Pu Liu Shixia Yan Donald Farquhar Roger Lindle David Smither
GE Global Research Center, 1800 Cailun Road, Shanghai, 201203, China. GE Global Research Center, One Research Circle, KWD-238, Niskayuna, NY 12309-1027, USA. Manufacturing & Quality Technology Department, GE Aviation, 1 Neumann Way, MD E74, Cincinnati, OH 45
国际会议
北京
英文
1024-1028
2008-10-27（万方平台首次上网日期，不代表论文的发表时间）