SDG multiple fault diagnosis by real-time inverse inference |
| |
Authors: | Zhao-qian Zhang Chong-guang Wu Bei-ke Zhang Tao Xia An-feng Li |
| |
Affiliation: | College Of Information Science and Technology, Beijing University of Chemical Technology, BeiJing 100029, China |
| |
Abstract: | In the past 20 years, one of the qualitative simulation technologies, signed directed graph (SDG) has been widely applied in the field of chemical fault diagnosis. However, the assumption of single fault origin was usually used by many former researchers. As a result, this will lead to the problem of combinatorial explosion and has limited SDG to the realistic application on the real process. This is mainly because that most of the former researchers used forward inference engine in the commercial expert system software to carry out the inverse diagnosis inference on the SDG model which violates the internal principle of diagnosis mechanism. In this paper, we present a new SDG multiple faults diagnosis method by real-time inverse inference. This is a method of multiple faults diagnosis from the genuine significance and the inference engine use inverse mechanism. At last, we give an example of 65t/h furnace diagnosis system to demonstrate its applicability and efficiency. |
| |
Keywords: | Qualitative simulation SDG Multiple fault diagnosis Real-time inverse inference Inference engine |
本文献已被 ScienceDirect 等数据库收录! |
|