Abdullah Sarim Khan

Implementing an embedded Extended Kalman Filter (EKF) in C/C++ on a 32-bit microcontroller for real-time SoC, SoH and RUL estimation using equivalent-circuit battery models. Designing a custom PCB-based BMS integrating multi-cell voltage/current/temperature sensing, active cell balancing and battery protection for second-life LiFe4O battery systems.

Building an explainable deep learning framework for LFP battery RUL prediction covering capacity degradation, knee-point proximity, and regime-aware modelling. Benchmarking ML/DL models on public battery aging datasets and validating on in-house NEWARE/BettaTeQ cycling data using PyTorch and SHAP.

Fabricated Li-ion battery packs (10S4P, 4P) with overvoltage/undervoltage/overcurrent protection circuits. Developing a hardware BMS monitor for real-time SoC/SoH analysis; schematic and PCB layout (KiCAD) complete.

Optimized decision-tree and random-forest models with Artificial Bee Colony (ABC) to estimate SoC for LG HG2 18650 cells across 0–40 C. Analyzed temperature-driven accuracy variation to improve estimator reliability for battery energy storage applications.

Modelled Li-ion cells with a second-order (2-RC) equivalent circuit model and simulated a passive cell-balancing circuit in MATLAB/Simulink. Integrated an ADAM-optimized feed-forward neural network for SoC prediction and resistor switching; validated on an OPAL-RT OP-4510 hardware-in-the-loop testbed.

Ongoing Master of Science in Computer Science (Research) with a thesis on explainable deep learning for LFP battery RUL prediction.

Bachelor of Engineering in Electrical from Aligarh Muslim University, completed in 2025.