Mohammad Rezwan Khan

_{Physics → models → evidence → energy systems. Illustrative; no live data.}

Portfolio · Publications · LinkedIn · ORCID

I turn battery thermal behavior, power-conversion constraints, and QA requirements into auditable engineering decisions. My work connects lithium-ion battery models, BESS delivery evidence, and smart-grid control from physical assumptions to reproducible tests.

Electrical R&D engineer · PhD in Energy Technology, Aalborg University
MATLAB/Simulink + Python · Battery systems · BESS QA/QC · Smart grids

Current mission: build inspectable energy-system tools that help researchers and engineers move from assumptions to defensible results.

The DOI anchors the scientific record; the linked repositories expose selected equations, assumptions, tests, and engineering limits.

Browse all 17 publications · Verify the ORCID record

Start here: the Energy Lab connects 1RC/2RC battery dynamics, thermal models, SOC estimation, power conversion, and grid-tied/grid-forming BESS control in one deterministically regression-tested reference repository with explicit validation limits.

These labels describe the evidence currently published; they are not claims of hardware qualification, grid-code compliance, or field validation.

Validation vocabulary: unit-tested means function-level expected behavior; regression-tested means deterministic outputs reproduced in CI; numerically verified means analytic, conservation, or convergence checks; and synthetic holdout evaluated means evaluation on generated unseen profiles. I reserve measured-data validated, independently replicated, HIL/field evaluated, and qualified/certified for projects that publish evidence meeting those stronger definitions.

• Traceable physical-cell datasets and held-out measured validation for battery parameter-identification models.
• Validation-based cycle-life model selection with uncertainty intervals for end-of-life and remaining-useful-life estimates.
• Reproducible BESS acceptance evidence spanning FAT, SAT, commissioning, energization, and handover.

Core tools: MATLAB · Simulink · Python · NumPy/SciPy · Jupyter · GitHub Actions
Domains: electro-thermal modeling · BMS/SOC estimation · power electronics · BESS assurance · grid-forming control

I welcome substantive collaboration around models, validation datasets, BESS engineering evidence, and reproducible energy-system research. Start with the Energy Lab roadmap or a focused open issue.

Follow an engineering question from the physical system to its source evidence.

The Energy Evidence Atlas brings my existing energy projects into a visual reading map, with a searchable project finder, research-to-code trails, commit-pinned sources and explicit limitations. The optional 3D assembly is conceptual; it does not represent selected equipment or establish hardware, facility or grid-code validation.

• Explore the cell-to-grid system tour.
• Find a model or toolkit.
• Read source records and engineering limits.
• Try the bounded reserve calculation, then open the linked Datacenter Twin Lab for its full documented scenarios.

Explore the self-paced learning workspace alongside the cell-to-grid 3D tour:

• Battery & thermal: MATLAB Simulink Energy Lab and Battery Cycle-Life Analyzer.
• Assurance & continuity: BESS QA/QC Toolkit and Datacenter Twin Lab.
• Dispatch & markets: VoltRL, HelioForge Energy Lab, and ZephyrTrade.

Progress is self-paced and stored locally in your browser. This workspace offers no certification; synthetic examples are for exploration and do not constitute field validation.

Explore 69 browser demos on MKLab (catalogue checked 2026-10-17) for energy, learning, civic information and everyday planning. Each app includes usage steps and visible limits.

The new source repositories remain private; the public website provides synthetic browser editions. Use fictional information only. These demos do not establish clinical, legal, financial or emergency-service readiness.

• MKLab: Explore 69 public browser demos for energy, learning, civic information and everyday planning. Fictional or synthetic demonstrations, not medical, financial, legal or emergency services. Some source repositories are private.
• Intelligent UI: Local, no API; deterministic approved-source routing, no connected AI, and not an engineering approval.
• HelioForge Energy Lab: Explore hybrid-energy architectures, linked lessons and bundled scenario results. The hosted preview shows bundled examples. New numerical runs require the separate local Python application; no calibrated plant control.
• ZephyrTrade: Compare wind-market strategies and calculate hypothetical offers from synthetic observations. Synthetic historical data, not live prices, order execution or investment recommendations.
• Datacenter Twin Lab: Learn electrical continuity through 24 lessons, including 12 continuity and 12 infrastructure lessons, with 18 synthetic scenarios and inspectable energy ledgers. Uncalibrated synthetic electrical model; not facility safety, certified uptime or live equipment control.

New MKLab additions include Shomonnoy, GridLoom Europe, Solmora Grid, Threadmark, Ionweave Control and JolPulse. Use their published synthetic-data boundaries; no operational forecast, charging-control or flood-warning claim is made.

• battery-power-models: 6 unit tests in the linked CI run. CI evidence.
• battery-cycle-life-analyzer: 94 passed and 6 skipped tests per Python matrix job. CI evidence.
• battery-thermal-modeling-notes: 84 unit tests in the linked CI run. CI evidence.
• voltrl: 67 benchmark tests; bundle audits are separate checks. CI evidence.
• smart-grid-storage-playbook: 151 unit tests in the linked CI run. CI evidence.
• zephyrtrade: 86 passed tests per Python matrix job. CI evidence.
• datacenter-twin-lab: 150 Python tests per matrix job; 88 browser journeys and 18 export checks are separate suites. CI evidence.

Source and CI records reviewed 7 October 2026. Counts refer to individual jobs at linked commits, not totals across matrix jobs. Current portfolio catalogue.

Snapshot checked 18 October 2026: 17 owned public repositories — 12 non-forks (including this profile), 5 forks. These counts exclude private and collaborator-owned repositories and are separate from the 69 MKLab demos. Browse the live GitHub repository list or the portfolio directory and dated source records. Workflow results apply to their linked commits; they are not independent validation.