Integration of sequence learning and CBR for complex equipment failure prediction
ICCBR'11 Proceedings of the 19th international conference on Case-Based Reasoning Research and Development
Hi-index | 0.00 |
This work is focused on defining and implementing a new similarity criterion for sequences of symbolic representations. The proposed algorithm returns a normalized index related to the degree of matching between sequences of qualitative labels. Performance of this method has been tested in the classification of voltage sags (transient reduction of voltage magnitude) gathered at 25 kV distribution substations. The objective is to assist monitoring systems in locating the origin of such disturbances in the transmission (HV) or distribution (MV) system. The promising classification accuracy achieved when this method was used with test data suggests that the presented algorithm could be applied satisfactorily and confirms its utility in classification approaches.