AI Railway Energy Consumption Simulation
Budget: £10 – £15 GBP
I need a full-cycle simulation that predicts and optimises the energy consumption of a hybrid railway system. The core of the model must be driven by genetic algorithms—other AI methods can be auxiliary, but the evolutionary approach should remain the primary optimisation engine.
The simulator has to ingest weather and other environmental parameters (temperature, wind, gradient-related conditions, etc.) and translate them into accurate traction-power needs for each timetable segment. Rolling-stock characteristics and any operational constraints will be supplied once the project starts; please build the architecture so those inputs can be added or swapped easily.
Primary deliverables
• A documented Python (or MATLAB) codebase that runs the genetic-algorithm optimisation loop and outputs energy-use profiles.
• A concise technical note explaining model assumptions, fitness function design, and how environmental data flows through to the final consumption figures.
• A repeatable way (command line or lightweight GUI) to feed new weather datasets and produce revised forecasts without touching the core code.
Acceptance criteria
1. When provided a week-long environmental dataset, the simulation completes in under one hour on a standard desktop and reports total kilowatt-hours within ±3 % of the baseline values I will share for validation.
2. All key parameters (population size, mutation rate, convergence threshold) are exposed for tuning in a config file.
3. Code quality follows PEP 8 (or MATLAB style guide) and includes unit tests for each major function.
If you have prior railway or transport-energy modelling experience, mention it briefly with an example output graph or link. I’m ready to start as soon as I confirm your approach aligns with these needs.
The simulator has to ingest weather and other environmental parameters (temperature, wind, gradient-related conditions, etc.) and translate them into accurate traction-power needs for each timetable segment. Rolling-stock characteristics and any operational constraints will be supplied once the project starts; please build the architecture so those inputs can be added or swapped easily.
Primary deliverables
• A documented Python (or MATLAB) codebase that runs the genetic-algorithm optimisation loop and outputs energy-use profiles.
• A concise technical note explaining model assumptions, fitness function design, and how environmental data flows through to the final consumption figures.
• A repeatable way (command line or lightweight GUI) to feed new weather datasets and produce revised forecasts without touching the core code.
Acceptance criteria
1. When provided a week-long environmental dataset, the simulation completes in under one hour on a standard desktop and reports total kilowatt-hours within ±3 % of the baseline values I will share for validation.
2. All key parameters (population size, mutation rate, convergence threshold) are exposed for tuning in a config file.
3. Code quality follows PEP 8 (or MATLAB style guide) and includes unit tests for each major function.
If you have prior railway or transport-energy modelling experience, mention it briefly with an example output graph or link. I’m ready to start as soon as I confirm your approach aligns with these needs.