A Robust False Data Detection in Smart Grid Using Trees Decision Algorithm

Abstract

The smart grid has the objective of improving the control and the supervision of the electric grid with the aim to ensure a reliable and effective functioning; This is possible with a system based on a precise state estimate that accurately reflects the physical aspects of power grids. However, the progress of the grid element, basically the smart meter has considerably increased the number of access point to the electrical grid. As a result, a large flow of false data is generated in the electrical system that compromises the state estimation process. Unlike the conventional detection test, we propose a method capable of handling complex false data. Our analyzes are validated by experiments on a physical bus feeding system performed on PSS / E, then we use Matlab to build false data based on the Jacobian. The collected data is analyzed with a learning technique basing on decision tree method.