Muhammad Abdullah

Mr. Muhammad Abdullah

Department of Electrical Engineering, Information Technology University, Lahore, Punjab, Pakistan
orcid

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
4 May 2026
Verified
Conference Paper
Closed access
ASAbubakar Siddique
Abubakar Siddique
SBSonia Bibi
Sonia Bibi
MAMuhammad Abdullah
Muhammad Abdullah
UFUmar Fayyaz
Umar Fayyaz
Muhammad AbdullahMuhammad Abdullah
Muhammad Abdullah
This study proposes a physics-informed machine learning method for predicting the breakdown voltage (BDV) and estimating the insulation status of epoxy-based nanocomposite insulators. A finite element model using COMSOL Multiphysics simulated the electric field distribution and generated a systematic dataset of 300 samples, depending on nanoparticle processing conditions, filler material, curing temperature, and curing time. Based on the IEC 60243 standard, the BDV severity levels were categorized for the classification tasks. Using Google Colab, supervised learning models were developed for regression (BDV values) and classification (severity levels). Ensemble learning, particularly CatBoost and LightGBM, showed excellent performance, with high $\mathbf{R}^{\mathbf{2}}$ and low MAE. The experimental validation using a 90 kV high-voltage test rig confirmed that the BDV improvement was maximum for epoxy with SiO2 filler compared to unfilled epoxy. The excellent agreement between simulated, predicted, and experimental BDV values confirms the efficacy of the simulation-machine learning hybrid model. This study provides a reliable platform for fast optimization of dielectric materials and insulation performance assessment.

22 May 2026
Verified
Conference Paper
Closed access
ASAbubakar Siddique
Abubakar Siddique
Muhammad AbdullahMuhammad Abdullah
Muhammad Abdullah
SAShahid Atiq
Shahid Atiq
AAAli Ayyan
Ali Ayyan
FQFatima Qazi
Fatima Qazi