Md Ittehadul Islam

• Data analysis and literature review of VLE, LLE, and cloud-point data for polymer and copolymer solutions for thermodynamic modeling

• Pursuing an M.Sc. in Simulation Sciences at RWTH Aachen University (GPA: 1.9, 113 ECTS completed) • Coursework spans numerical methods, machine learning, control engineering, and computational fluid dynamics

• Designed a blackbox autoregressive feedforward neural network to model cabin thermal dynamics under varying operating conditions • Integrated the cabin model with a graybox powertrain model to build a holistic BEV thermal architecture covering cabin, battery, and powertrain domains • Developed and evaluated Economic Model Predictive Control (EMPC) strategies targeting energy efficiency while maintaining cabin comfort • Validated the control framework under realistic driving scenarios

• Supported the training of a Reinforcement Learning (RL) agent to minimize NOx and soot emissions in a dual-action manual transmission system • Validated results across MiL, HiL, test bench, and real vehicle environments • Benchmarked RL performance against an automatic transmission counterpart • Conducted literature review on RL-based emission reduction strategies

• Developed and evaluated three Physics-Informed Neural Network (PINN) models with varying levels of physical knowledge for state estimation in a Van de Vusse Reactor • Compared model performance across full-physics, partial-physics, and data-driven configurations to assess the impact of embedded domain knowledge

• Developed and optimized process simulation code for superstructure optimization in cement production • Applied computational modeling to evaluate and compare production process configurations for energy efficiency

• Contributed to the development of the in-house CFD tool MAIA by implementing newer functions and algorithms

• Investigated the feasibility and performance of flexible PV panels under the climatic conditions of the Khulna region

• IELTS instructor for higher education enthusiasts