Detecting faults in electrical power grids is of paramount importance, eitherfrom the electricity operator and consumer viewpoints. Modern electric powergrids (smart grids) are equipped with smart sensors that allow to gatherreal-time information regarding the physical status of all the componentelements belonging to the whole infrastructure (e.g., cables and relatedinsulation, transformers, breakers and so on). In real-world smart gridsystems, usually, additional information that are related to the operationalstatus of the grid itself are collected such as meteorological information.Designing a suitable recognition (discrimination) model of faults in areal-world smart grid system is hence a challenging task. This follows from theheterogeneity of the information that actually determine a typical faultcondition. The second point is that, for synthesizing a recognition model, inpractice only the conditions of observed faults are usually meaningful.Therefore, a suitable recognition model should be synthesized by making use ofthe observed fault conditions only. In this paper, we deal with the problem ofmodeling and recognizing faults in a real-world smart grid system, whichsupplies the entire city of Rome, Italy. Recognition of faults is addressed byfollowing a combined approach of multiple dissimilarity measures customizationand one-class classification techniques. We provide here an in-depth studyrelated to the available data and to the models synthesized by the proposedone-class classifier. We offer also a comprehensive analysis of the faultrecognition results by exploiting a fuzzy set based reliability decision rule.
展开▼
