Multi Agent System Based Fault Location in a Smart MicroGrid System

Abstract

Multi Agent System Based Fault Location in a Smart MicroGrid System Seetaram Alwala The electric power grid is inherently a huge network that has been going through a lot of transformation because of its restructuring and increase in the deployment of Distributed Generation. The steep increase in electricity demand and the large financial losses caused by the blackouts has initiated research towards increasing the security and robustness of the electric grid and making it a Smart Grid. Smart Grids might incorporate MicroGrids, which can be defined as low to medium voltage networks that contain aggregation of certain loads and Distribution Generation units that can operate in either grid-connected or islanded mode. One of the leading technologies to deploy intelligence in the electric grid is the agent technology. Multi Agent systems are basically a congregation of individual agents that can sense changes in their environment and autonomously make decisions to meet their design objectives. The technological advancement in the agent technology can provide solutions for a wide range of Smart Grid activities. In a MicroGrid, a single fault can lead to catastrophic, cascading events that could result in a blackout. This makes fault location an important problem to solve. The penetration of Distributed Generation units causes a two way power flow in the traditional radial distribution system. This, coupled with the inherent unbalanced nature of the distribution system, makes the fault location a difficult problem to solve. In this thesis, a Multi Agent system has been formulated to determine the precise fault location and identify faulty zones. The algorithm for the fault detection is based on the magnitude of the sequence components of currents and current directions during a fault in the system. v

Comparing Different Fault Identification Algorithms in Distributed Power System

“…In [2] the symmetrical components of the current will be calculated (Positive sequence, Negative Sequence and Zero Sequence) and the symmetrical components will be used to identify the location of the fault in PDS.…”

“…The proposed method that will be used in this thesis is the change of current, which was proposed in [1] and the symmetrical components proposed in [2]. The identifying algorithm will be executed as MATLAB M-File function, and the simulation will display the location of the fault for each different scenarios.…”