Energy Control System

The system receives electricity prices, demand forecasts, asset constraints, and intraday updates, then creates a cost-minimizing energy schedule.

Tech Stack

• Backend: Python, FastAPI
• Optimization: Google OR-Tools
• Database: PostgreSQL
• Event streaming: Apache Kafka
• Frontend: React, Vite, Recharts
• Local deployment: Docker Compose

Main Components

Data Ingestion Service

The ingestion service accepts and validates input data:

• hourly electricity prices
• demand forecasts
• energy asset constraints
• intraday demand updates

Optimization Service

The optimization service builds a 24-hour day-ahead schedule. It decides how much energy should come from:

• grid electricity
• gas
• battery discharge

It also decides when the battery should be charged. The objective is to minimize total energy cost while satisfying demand and respecting operational constraints.

Main constraints:

• demand must be covered every hour
• battery state of charge must stay between 0 and capacity
• max charge and discharge limits must be respected
• gas output must not exceed its maximum capacity
• the battery cannot charge and discharge at the same time

Intraday Service

The intraday service simulates real-time correction during the operating day. It receives updated actual demand, compares it with the day-ahead schedule, calculates the deviation, and creates a corrective action.

The service stores intraday actions in PostgreSQL and publishes them to Kafka.

Visualization Layer

The React dashboard displays:

• input data
• day-ahead schedule
• battery state of charge
• intraday corrections
• test scenario results

Kafka Topics

• price_signals
• demand_forecasts
• ecs_schedule
• intraday_updates
• intraday_actions

API Endpoints

Ingestion Service

• POST /prices
• POST /forecast
• POST /assets
• POST /intraday
• GET /prices
• GET /forecast
• GET /assets
• GET /intraday
• GET /health

Optimization Service

• POST /optimize/day-ahead
• GET /schedule/{date}
• GET /health

Intraday Service

• POST /optimize/intraday
• GET /intraday/actions/{date}
• GET /health

Database Tables

• prices
• demand_forecasts
• energy_assets
• day_ahead_schedules
• intraday_actions
• simulation_runs
• test_results

Running Locally

Create a local environment file:

cp .env.example .env