Optimization of the FP-Growth Algorithm in Data Mining Techniques to Get the Electric Power Theft Pattern for the Development of Smart City

Abstract

In general, people's daily lives cannot be separated from energy resources, namely electricity. Where electricity is one of the basic needs of society for survival. The amount of electricity consumption is now increasing, because people almost every day and every home uses electronic devices. Not only the needs of homes that use electricity, even companies, schools, shopping centers also cause an increase in the use of electric current. The increasing use of electric current makes many people commit violations in the use of electricity. This makes the Electric Power Company in managing the distribution of electrical energy to the public to check for violations in the use of electricity. For this reason, data mining techniques are needed to get patterns in energy use violations by optimizing the FP-Growth algorithm in getting a good pattern where the 1300Kwh meter in the optimized pattern is recognized as having more frequent violations such as breaking electric current. So that the optimization pattern can be applied to the electricity meter to develop smart city concepts such as the smart grid. In this paper, we apply data mining with the FP-Growth algorithm in analyzing the pattern of electricity theft. The existing pattern describes the cause and effect of the theft of electric power. In the future smart electricity grid, also known as the "smart grid", network stability is expected to be achieved through different energy sources, among others. If the results of renewable energy that depend on weather and time are not adjusted to the latest electricity usage, there can be a network imbalance that ends in a blackout or blackout.