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 current source includes a bounded 24-hour browser energy-balance Terminal; advanced studies require the local Python application. Check the published build before assuming a hosted capability. Synthetic screening only; 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 30 October 2026: 12 owned public originals, including this profile. This active directory excludes forks and private repositories. It is separate from the browser-demo catalogue and historical upstream contributions above. Browse the live repository list. Source and workflow evidence apply to their linked revisions.