Distribution Grid Optimization Engine

Executive Summary

This project implements a hybrid physics-software engine for analyzing low-voltage distribution grids. It was engineered to address specific DSO challenges such as reverse power flows and voltage violations caused by Distributed Energy Resources (DERs).

Key Features

• Physics Core: Forward-Backward Sweep (FBS) solver optimized for high R/X ratio radial feeders.
• Domain Modeling: IEC 61970 (CIM) compliant data models using Pydantic for robust type validation.
• Advanced Controls: Fuzzy Logic Controller for Volt-VAR optimization in PV inverters.
• Analytics: Hosting Capacity Analysis using binary search ($O(\log N)$) for rapid connection assessment.

Architecture

• Layer 1: Domain Data (Pydantic): Enforces physical constraints (e.g., non-negative resistance) at the I/O boundary.
• Layer 2: Topology (NetworkX): Validates radiality and determines feeder ordering.
• Layer 3: Solvers (SciPy/NumPy): Implements numerical methods for non-linear power flow.
• Layer 4: Application: Orchestrates solvers to answer business questions (e.g., "Can I connect this EV charger?").

Engineering Assumptions & Future Roadmap

Physical Assumptions

• Three-Phase Balance: The solver currently assumes a perfectly balanced three-phase system and models the positive sequence network.
• Topology: The Forward-Backward Sweep (FBS) implementation is optimized for radial distribution feeders (trees). Meshed topologies would require a Newton-Raphson solver upgrade.

Production Considerations (TODOs)

• Scaling: The current BFS parent lookup is $O(N^2)$ in the worst case. For production grids (>10k nodes), this would be optimized to $O(N)$ by caching the tree structure.
• Concurrency: The hosting capacity algorithm currently mutates the grid state in-place. A production version would use immutable data structures to ensure thread safety in a REST API context.
• Grid Codes: The Fuzzy Logic Controller is a conceptual demonstration. Real-world implementation would require adherence to VDE-AR-N 4105 static voltage support curves.

Quick Start

Option 1: Using Python Directly (Development)

1. Install dependencies:

   pip install -r requirements.txt
   
   Run the simulation and validation suite:
   

Bash

python main.py

rem to update 0. CONFIGURATION & CONSTANTS before runing

Option 2: Using Docker (Recommended for Production)

Docker provides an isolated, reproducible environment for running the grid engine:

Build and run with docker-compose (easiest):

# Build and start the container
docker-compose up -d

# View logs
docker-compose logs -f

# Stop the container
docker-compose down

Or using Docker directly:

# Build the image
docker build -t cim-grid-engine .

# Run a simulation
docker run --rm cim-grid-engine

# Run with output volume mounted
mkdir -p output
docker run --rm -v $(pwd)/output:/app/output cim-grid-engine

Benefits of Docker deployment:

• ✅ Consistent environment across systems (Windows/Mac/Linux)
• ✅ No dependency conflicts
• ✅ Production-ready containerization
• ✅ Easy integration with CI/CD pipelines
• ✅ Scalable for distributed grid analysis