Symbolic reasoning and quantitative analysis for fault detection and isolation in process plants
| Autor: | Zohreh Fathi, Gerard Gilliland, W. Fred Ramirez, A.P. Tavares |
|---|---|
| Rok vydání: | 1993 |
| Předmět: |
Computer science
business.industry Kalman filter Work in process Machine learning computer.software_genre Fault detection and isolation Adaptive filter Electric power system Knowledge-based systems Symbolic reasoning Artificial Intelligence Control and Systems Engineering Redundancy (engineering) Artificial intelligence Electrical and Electronic Engineering business computer |
| Zdroj: | Engineering Applications of Artificial Intelligence. 6:203-218 |
| ISSN: | 0952-1976 |
| DOI: | 10.1016/0952-1976(93)90063-4 |
| Popis: | Fault diagnosis in the area of process operations is critical for modern production and is receiving increasing theoretical and practical attention. In spite of many research and practical attempts, process fault diagnosis remains a rather complex task. This work presents a diagnostic methodology in which the symbolic reasoning of knowledge-based systems techniques is integrated with quantitative analysis of analytical redundancy methods. The system first performs a diagnosis by means of a compiled knowledge structure, and then attempts to build a detailed explanation by using proper fault models. It also performs various statistical analyses to determine the process condition and check the validity of models (adaptive filters). This unified approach increases the completeness and reliability of the diagnostic system. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|