Power transformer condition monitoring and fault diagnosis with multi-agent system based on ontology reasoning

Abstract

Power transformer is one of the key important and most expensive equipments in electrical power system. Building systems to monitor their real time behaviours and diagnose their faults autonomously with comprehensive knowledge-base are the key issue. This paper provides a new framework for power transformer monitoring and fault diagnosis based on ontology reasoner. The Gaia methodology is applied to clarify, simplify and standardize the design of the multi-agent system. The real time data is gathered from power transformer, saved into database and it is also available to user on request. Reasoning techniques such as rule-based reasoning and ontology-based reasoning can reduce the user's works. The built ontology provides the comprehensive knowledge-base for deducing and diagnosing its faults. The applied ontology reasoner for fault detection is based on description logic.