Thesis for the Degree of Master of Science


Year 2001
Sasa Haavisto

Monitoring of Rule based fault diagnosis system for industrial processesan Dearomatization Process

The number of processes and amount process equipment an operator has to monitor have increased simultaneously with the growth of automation. A fault diagnosis system can detect abnormalities in the process at an early stage and support the operator in decision making. With the help of a fault diagnosis system, the operator has more time for recovery action before a threatening fault situation becomes so bad that it cannot be controlled anymore.In the literature part, design process for user interfaces and different kinds of visualization methods are introduced. Also, fault diagnosis methods are studied with industrial application examples.In the experimental part, a rule-based fault diagnosis system was designed and implemented. Also, a process interface for Outokumpu Harjavalta Metals Damatic XD- automation system and user interfaces for maintaining and operating were implemented.The fault diagnosis system implemented was tested on-line in the copper flash smelting process. If abnormalities in the process, based on heuristic rules, are detected, the fault diagnosis system gives expert advice for recovery operation. The rules are based on limit value checking of measurements, formulas and Kohonen maps.Rule based fault diagnosis system for industrial processes

This info last modified 24 Sep 2017 - contact webmaster