Analysis of sequences of events for the characterisation of faults in power systems

A new formalism is described for the systematic analysis of sequences of events recorded in power systems. It aims to characterise sequential patterns, or episodes, formed by ordered sets of events associated with faults in the power system. The work introduces formal definitions to deal with sequences of events that are exemplified and evaluated in a case study. A sequence of events is a set of dated events recorded at a single point of the network during a specific period of time. However, only some subsets of them are of interest, either because they present similar features (shape, duration, etc.) or because they are related in a way that allows them to be considered together (timing, periodicity, etc.). These subsets of significant events in a sequence are called episodes and are instances of possible patterns that define how permanent and transient faults behave. With this goal, the authors evaluate the use of a priori patterns and a pattern discovery algorithm in sequences of electrical events recorded in different substations as part of the case study.

► Formal definitions to deal with sequences of events in power systems are introduced.
► Fault typologies are defined in terms of sequences of events.
► Events associated to single-phase faults collected in five substations are analysed.
► We find possible patterns about the behaviour of permanent and transient faults.
► Longer sequences of events are required to address the problem of fault prediction.