Three Lab members recently showcased their research at 2025 ICDS Poster Symposium.
1. Enhancing Predictive Maintenance in Power Distribution Networks
Presenter: Akintaro Akinwale, PhD Student
Title: Enhancing Predictive Maintenance Strategies in Power Distribution Network using Fault Pattern Analysis with Association Rule Mining
This research introduces a novel approach to predictive maintenance by leveraging fault pattern analysis combined with association rule mining. The methodology aims to improve reliability and reduce downtime in power distribution systems by identifying hidden fault correlations and optimizing maintenance schedules.
2. Day-Ahead Electricity Price Forecasting [ LinkedIn Post Link ]
Presenters: Pranav Balachander and Lasya Madhuri Gundapaneni, Undergraduate students
Title: Day Ahead Electricity Price Forecasting: A CNN-BiLSTM-Attention Hybrid Model with Hyperparameter Tuning
This project proposes a hybrid deep learning architecture integrating Convolutional Neural Networks (CNN), Bidirectional LSTM, and Attention Mechanisms to forecast electricity prices with high accuracy. By incorporating hyperparameter tuning, the model significantly improves predictive performance, offering valuable insights for energy market participants and grid operators.
Both posters attracted significant attention during the symposium, showcasing the lab’s interdisciplinary expertise in machine learning, energy systems, and predictive analytics.
