Meet Inspiring Speakers and Experts at our 3000+ Global Conference Series Events with over 1000+ Conferences, 1000+ Symposiums
and 1000+ Workshops on Medical, Pharma, Engineering, Science, Technology and Business.

Explore and learn more about Conference Series : World's leading Event Organizer

Back

Kshitija Bhasme

Kshitija Bhasme

SNDT Women’s University, India

Title: Machine learning approach to cyber security in smart grid

Biography

Biography: Kshitija Bhasme

Abstract

Nowadays, power grids are being integrated into the smart grid concept for efficiently managing and delivering electric power. A smart grid provides a two-way dialogue where electricity and the information can be exchanged between the utility and its customers by allowing them to sell and purchase electricity. Attacks on such systems may have catastrophic impact, especially on the customers which is a particular threat in the winter, when people can be left without heat. Hence, the mitigation solutions for these situations are necessary. It is impractical to test attacks and mitigation strategies on real-time networks. Therefore, we propose a predictive analysis to make the smart grid prepared to address power failure faults and attacks in the time of emergencies. We explain different power system scenarios and network designs by simulating IEEE 14 bus system models through Power System Analysis Toolbox (PSAT) in MATLAB for the assessment of power flow and to perform a complete and accurate power system analysis. A data sheet of power system analysis is created and fed into various machine learning algorithms in WEKA data mining tool. The goal is to make our machine learn different power system analysis scenarios in order to be able to detect normal, fault and attack instances in a power system by checking the accuracy of the instances being learnt by the machine and cluster formations of all the correctly classified instances. This can help save millions of dollars and long hours of investigation invested to study and detect unusual occurrences in a power system. We embark upon a journey towards a new era of reliability, availability and efficiency of the smart grid by bringing in the machine learning domain in the assessment process which is a faster and more reliable way of detecting adversities in a power system than the conventional way.