This project models a modular, physics-informed smart grid system composed of decentralized household agents, each with local power generation, consumption, and battery storage. It simulates energy flows, system balancing, overload responses, and basic market dynamics across a neighborhood-scale power grid.
- Simulate real-time power distribution across household agents
- Model storage, transfer, and loss mechanisms within a local grid
- Implement overload handling (e.g. dump nodes / fail-safe agents)
- Capture system behavior under generation/load imbalances
- Provide a framework to test decentralized energy trading and pricing
- Household agents with solar generation, demand profiles, and battery systems
- Power balancing across agents and with grid connection
- Resistance-based line loss (Ohmic model)
- Dump load logic to absorb excess power in overload conditions
- Hour-by-hour time loop simulation over multiple days
- Optional pricing/market logic for peer-to-peer energy trades
- Modular structure for extensibility and experimentation
smart-grid-simulator/
│
├── agents/
│ ├── household.py # Generation, consumption, battery
│ ├── dumpnode.py # Absorbs overload / excess energy
│
├── network/
│ ├── transmission.py # Power flow and resistive losses
│
├── market/
│ ├── pricing.py # Optional pricing logic and P2P trades
│
├── planner/
│ ├── simulate_day.py # Time-loop simulation and event management
│
├── data/
│ └── sample_inputs/ # Example load/gen profiles
│
├── notebooks/
│ └── demo_run.ipynb # Example use and visualization
│
├── README.md
└── requirements.txt
- 10 households with hourly solar and consumption profiles
- Shared local battery capacity or individual storage
- Hour-by-hour simulation for a 7-day period
- Visualization of power flow, battery charge, and overload events
- Optional price signal simulation or peer-to-peer trading activation
- Integration of real ISO data (PJM, MISO, CAISO)
- Bayesian modeling of generation uncertainty (weather-based)
- Grid frequency & voltage fluctuation modeling
- Streamlit dashboard or REST API for grid planning
- Quantum-inspired optimization integration (AWS Braket)
- Python 3.x
pandas,numpy,matplotlib- Modular class structure
- Optional:
streamlit,plotly,boto3(AWS integration)
MIT License — open for educational and non-commercial adaptation.
Developed by Milad Alibabaie, with focus on modeling energy systems & designing infrastructure.