Data Driven Approach for System Disturbances in Digital Substation

Authors

  • A. Rathinam , V Kubendran , Priya Kumari , A. Ashwin Kumar , S. Srija , Divy Desai

Abstract

The substations are the essential part of the power grid as they are interconnected with physical devices such as
circuit breakers, transformers, transmission lines. Advancement in technologies has caused the transformation of
traditional grid into a digital grid. The upcoming of Digital Substations reduces the need of physical electric
connection among the high voltage equipment, growing a more secure environment. The data collected at field level
is used for energy management in Supervisory Control and Data Acquisition (SCADA) However, the abundance of
digital grid data is required to be utilized at its utmost potential thus providing a reliable and resilient grid
infrastructure. The increased volume of collected industrial data in power system demands more powerful and
intelligent machine learning tools for industrial analytics in power system field to strategizes and analyse the
historical data so to develop a predictive knowledge of disturbances in power system. The project tried to simulate
the grid disturbances affecting system parameters and developed an analytic model while selecting system parameter
as variables. It develops the tools used to conduct predictive machine learning analysis for large volumes of data.
The digital communication is replicated between field level and operator level with the help of OPC Matrikon server
and MATLAB simulation. The proposed model is evaluated on the standard IEEE 9 bus system and result is
obtained comparing data analytics algorithm. The result indicates the feasibility of forecasting digital substations
disturbances events using machine learning algorithm as support vector machine, KNN and decision tree, further the
most reliable one is taken out by comparing their accuracies.

Published

2020-03-31

Issue

Section

Articles