APPLICATION OF INFORMATION THEORY'S CHANNEL CAPACITY AS AN INDUSTRY MACHINE HEALTH AND DIAGNOSTIC METRIC

摘要

A metric for degradation quantification, and assessment of machine health was constructed from an analogy between a machine and a communications channel. The metric is based on Shannon's channel capacity concept applied to the difference between measured signals (induced by faults) and baseline signals of a machine system response. Machine faults that hinder operation also alter signal flow through a machine system; the effects appear as a residual signal riding on a baseline signal. Similar to noise hampering data transmission, this residual signal can be viewed as noise within the context of information theory, permitting application of information theory to machines, resulting in a health measure for assessment of machine health. The technique successfully predicted failures in a closed loop chain robot. The technique is favorable for industry, because it applies to measurement data, and calculations are done in the time domain without need of expensive frequency transformations. The technique can be used in the semi-conducting industry as a tool to monitor system performance during production, and to lower fab operating cost by extending component use and scheduling maintenance as needed.