A signed directed graph and qualitative trend analysis-based framework for incipient fault diagnosis
A signed directed graph and qualitative trend analysis-based framework for incipient fault diagnosis
复制标题
DOI：
10.1016/s0263-8762(07)73181-7
复制
发表时间：
2007-10-01
影响因子：
3.9
通讯作者：
Venkatasubramanian, V.
中科院分区：
文献类型：
--
作者：
Maurya, M. R.;Rengaswamy, R.;Venkatasubramanian, V.
本文提出了一种结合符号有向图（SDG）和定性趋势分析（QTA）的早期故障诊断框架。SDG是该框架中的第一级，并基于过程的初始响应提供可能的候选故障集。基于QTA（2级）执行对实际故障的搜索，该QTA使用传感器的时间演变来进一步解决问题。因此，该框架结合了SDG的完整性属性和QTA的高诊断分辨率属性。还提出了解决由于初始反应的不正确测量而引起的不正确诊断的问题的方法。田纳西州伊士曼（TE）的案例研究所提出的方法进行了测试。在所有可能的单一故障情况下执行正确的故障诊断。结果表明，该框架提供了快速，可靠和准确的早期故障诊断。
In this article a combined signed directed graph (SDG) and qualitative trend analysis (QTA) framework for incipient fault diagnosis has been proposed. The SDG is the first level in this framework and provides a possible candidate set of faults based on the incipient response of the process. The search for the actual fault is performed based on a QTA (level 2), which uses the temporal evolution of the sensors for further resolution. Thus, this framework combines the completeness property of SDG with the high diagnostic resolution property of QTA. Methods to address the problem of incorrect diagnosis arising due to incorrect measurement of initial response have also been presented. The proposed approach is tested on the Tennessee East-man (TE) case study. Correct fault diagnosis is performed in all possible single fault scenarios. It is shown that this framework provides fast, reliable and accurate incipient fault diagnosis.