Knowledge Extraction from Time Series of Electric Energy Demand using Temporal Data Mining

Planning activities are very important in the energy sector, where the utilities are seeking information that may assist in decisions regarding expansion needs and resource management, improving the quality of their services. This paper presents a methodology based on mining tools and representation of time series, in order to extract knowledge from series of electricity demand in various substations connected to an energy provider. To represent this knowledge, the language proposed by Morchen (2005) called Time Series Knowledge Representation (TSKR) is used. It was conducted a case study using time series of energy demand for 8 substations interconnected by a ring system, which feeds the metropolitan area of Goiania-GO (Brazil), provided by CELG (Companhia Energetica de Goias), responsible for the service of power distribution in the state of Goias (Brazil). Using the proposed methodology, three levels of knowledge that describe the behavior of the studied system were extracted, representing clearly the system dynamics, thus becoming a tool to assist planning activities.