How gauravsrc/ev-grid-intelligence works
EV Grid Intelligence is a synthetic-digital-twin decision framework for EV charging and battery swapping across five metro clusters. Its principal workflow generates or adapts data, forecasts demand and prices, plans energy and pack allocation, simulates baseline and AI operations, then evaluates KPIs and produces analyses and reports. The README documents strategic redeployment, bidding, pricing, and execution-facing capabilities; sampled code confirms the simulation and optimization core, but does not show live dispatch integrations. Links for unsampled subsystems rely on README statements or indexed references.
Main entry points: pipeline.py Pipeline orchestration · runner.py Cluster runner · engine.py Digital-twin engine · io.py Dataset cache
Components 23
Data and configuration
- Configuration
evgrid/config.py - Synthetic twin
evgrid/data/synth.py - Real-data adapter
evgrid/data/schema.py - Dataset cache
evgrid/data/io.py
Forecasting and planning
- Forecast features
evgrid/forecasting/features.py - Forecast models
evgrid/forecasting/models.py - Strategic redeployment
evgrid/optimization/strategic.py - Day-ahead scheduler
evgrid/optimization/day_ahead.py - LP builder
evgrid/optimization/lp_builder.py - Bid curves
evgrid/optimization/bidding.py - Dynamic pricing
evgrid/optimization/pricing.py
Operational simulation
- Policy definitions
evgrid/simulation/policies.py - Cluster runner
evgrid/simulation/runner.py - Digital-twin engine
evgrid/simulation/engine.py - Baseline calibration
evgrid/simulation/calibrate.py
Evaluation and analysis
- Pipeline orchestration
evgrid/pipeline.py - KPI computation
evgrid/simulation/kpis.py - Experiments
evgrid/analysis/experiments.py - Analysis figures
evgrid/analysis/figures.py - Figure styling
evgrid/analysis/style.py
Reporting
- Report builder
evgrid/reporting/build_report.py
Other components
- Framework user
- Grid and fleet systems
Connections 26
- Framework user → Pipeline orchestration: runs stages
README.md - Configuration → Pipeline orchestration: configures
evgrid/pipeline.py - Pipeline orchestration → Dataset cache: loads data
evgrid/pipeline.py - Dataset cache → Synthetic twin: generates on cache miss
evgrid/data/io.py - Real-data adapter → Dataset cache: adapts inputs
README.md - Pipeline orchestration → Forecast models: runs forecasting
evgrid/pipeline.py - Forecast models → Forecast features: builds features
evgrid/forecasting/models.py - Pipeline orchestration → Strategic redeployment: plans redeployment
evgrid/pipeline.py - Pipeline orchestration → Cluster runner: runs simulation
evgrid/pipeline.py - Cluster runner → Dataset cache: loads twin data
evgrid/simulation/runner.py - Cluster runner → Digital-twin engine: dispatches cluster tasks
evgrid/simulation/runner.py - Cluster runner → Forecast models: loads forecasts
evgrid/simulation/runner.py - Digital-twin engine → Day-ahead scheduler: solves day-ahead plan
evgrid/simulation/engine.py - Day-ahead scheduler → LP builder: builds and solves LP
evgrid/optimization/day_ahead.py - Digital-twin engine → Bid curves: forms bids
evgrid/simulation/engine.py - Digital-twin engine → Dynamic pricing: calculates prices
evgrid/simulation/engine.py - Policy definitions → Cluster runner: selects policy
evgrid/simulation/runner.py - Pipeline orchestration → Baseline calibration: calibrates baseline
evgrid/pipeline.py - Pipeline orchestration → KPI computation: computes results
evgrid/pipeline.py - Pipeline orchestration → Experiments: runs analyses
evgrid/pipeline.py - Experiments → Cluster runner: runs policy cases
evgrid/analysis/experiments.py - Pipeline orchestration → Analysis figures: builds figures
evgrid/pipeline.py - Analysis figures → Figure styling: applies styling
evgrid/analysis/figures.py - Framework user → Report builder: builds report
README.md - Grid and fleet systems → Real-data adapter: supplies extracts
README.md - Digital-twin engine → Grid and fleet systems: models execution
README.md