Abstract
This paper describes a prototype expert system for analysing the impulse test results on transformers. The criteria by which test result interpretation and fault diagnosis can be made were determined in consultation with transformer test experts and from the literature published over the past five decades. This survey revealed considerable heuristic and judgemental knowledge involved in the analysis of test data. Precise terminology is defined to describe and classify the test data so that it can be consolidated into a single consistent body of knowledge. This knowledge base has been used to develop a set of guidelines for fault identification, in consultation with local high-voltage test experts. A decision tree is presented to summarise the available knowledge and its application to fault analysis. A rule-based expert system based on the above decision tree was developed. This expert system was validated using known faults and the corresponding test results with the help of local experts. Particular care was taken to create a friendly interface to ensure correct data entry by error checking and by making available feedback and on-line help to the user.
Original language | English |
---|---|
Pages (from-to) | 11-21 |
Number of pages | 11 |
Journal | Engineering Applications of Artificial Intelligence |
Volume | 4 |
Issue number | 1 |
DOIs | |
Publication status | Published - 1991 |
Externally published | Yes |
Keywords
- Artificial intelligence
- expert system
- power system
- transformer