Projects
As an electrical engineer turned software engineer, I am passionate about building solutions for various engineering problems. Here are some of the projects I've worked on that reflect my skills and interests.
Development of Monitoring System for Fault Classification in Solar PV Plants using ML
​Undergraduate Thesis, SSN College of Engineering
-
Designed ML-based monitoring system to dynamically detect and classify electrical faults in large-scale PV arrays, and developed a web app using Flask.
-
Generated data using MATLAB Simulink and trained SVM, Decision-Tree, and Random Forest, models. Developed a CNN model which generated a high training accuracy of 84% and low testing loss of 0.048 for fault detection through thermal imagery.
-
Communicated research paper to ICEES-2023 Conference.
Neonatal Seizure Detection using Machine Learning
Personal Research Project
• Proposed an ML-based architecture using ProtoNN that can be deployed in ultra-edge devices for Neonatal Seizure Detection by processing EEG signals
• Our architecture achieved a best sensitivity of 87% and the model size of the ML classifier is optimized to just 4.84 KB with a minimum prediction time of 182.61 milliseconds
• Communicated the research work as a paper in AISP’22 Conference
Dynamic Global Maximum Power Point Tracking of PV Systems under Variant Partial Shading
using GWO-FLC
Student Funded Project, SSN Institutions
• Developed a high-efficiency global maximum point tracker for PV systems using Grey Wolf Optimisation
along with Fuzzy Logic Controller for partially-shaded conditions
• Proposed solution carries out a re-initialisation technique ensuring no oscillations around MPPs
• Published the research work as a paper at ICPEDC-2021
