Application Of Concept Lattices In Distributed Power Networks Fault Diagnosis

Formal concept analysis is a powerful tool for data analysis which making rapid developing in recent years. considering complexity of value-reducting the decision table when Rough Sets used in fault diagnosis of power system, and the limitations of traditional artificial intelligence technology in the fault diagnosis of large scale power networks, a new method of distributed fault diagnosis based on formal concept analysis is proposed. First, dividing the large scale power network into desired number of connected sub-networks, change the decision table of every sub-network into formal context after attribute-reducting, then proposes a new method to construct concept lattice by forming knowledge element of every single decision attribute based on the theory of formal concept analysis, gets the final decision concept lattice through reducting the redundant concept, then forms the pool of fault diagnosis rules of sub-networks and the tie lines, at the same time forms the Hasse diagram which shows the reasoning process. Research shows that this method is more simple and intuitive than Rough Sets, and also it can show out the type of fault of the tie lines and that may occur when lost some of the core attributes, it shows the method is pratical.