A platform that analyzes energy consumption data from IoT devices and manual uploads, providing anomaly detection, personalized savings recommendations, and carbon footprint calculations, supported by a modern visual dashboard.
- IoT data simulator and manual CSV upload
- AI-based analysis: consumption patterns, anomaly detection, scenario generation
- Personalized recommendation engine (% savings, USD and carbon equivalent)
- Live and visual dashboard (Streamlit)
- PDF/CSV report download
- Real IoT device integration (Tuya, Shelly, MQTT)
- Open data sources integration (UCI, EPIAS, etc.)
- Bill-based estimation
- Device-based estimation
- Start the application with Streamlit:
streamlit run app.py
- Load or simulate data through the dashboard.
- View analyses and recommendations, download reports.
- Python 3.9+
- Required packages: pandas, numpy, scikit-learn, streamlit, plotly, fpdf, requests, paho-mqtt
pip install -r requirements.txt- IoT Simulation: Generate synthetic data for testing
- CSV Upload: Upload your own consumption data
- Real IoT Devices: Connect Tuya, Shelly, or MQTT devices
- Bill Estimation: Estimate hourly consumption from monthly bills
- Device Estimation: Calculate consumption based on device specifications
- Open Data: Access UCI Household, EPIAS Turkey, and sample datasets
open_data.py modülü, çeşitli açık veri kaynaklarından enerji tüketim verilerini almak için aşağıdaki API'leri kullanır:
-
EPIAS Transparency Platform (Türkiye Gerçek Zamanlı Tüketim)
- API:
https://seffaflik.epias.com.tr/transparency/service/consumption/real-time-consumption - Amaç: Türkiye'nin saatlik toplam elektrik tüketim verilerini almak ve analizlerde kullanmak.
- API:
-
Kaggle UCI Household Power Consumption Dataset
- API:
https://archive.ics.uci.edu/ml/machine-learning-databases/00235/household_power_consumption.zip - Amaç: Ev tipi elektrik tüketim örnek verisiyle analiz ve test yapmak.
- API:
-
Open Power System Data
- API:
https://data.open-power-system-data.org/time_series/latest/time_series_60min_singleindex.csv - Amaç: Farklı ülkelerin şebeke bazlı saatlik elektrik tüketim verilerini almak ve karşılaştırmalı analizler yapmak.
- API:
Her bir API, platformun analiz ve raporlama fonksiyonlarını desteklemek için veri sağlar. Bu modül sayesinde kullanıcılar, gerçek ve örnek veri kaynaklarından kolayca veri çekebilir ve analiz edebilir.
timestamp,device_id,consumption_kWh
2024-01-15 00:00:00,Refrigerator,0.45
2024-01-15 01:00:00,Refrigerator,0.51
2024-01-15 00:00:00,AC,1.20- Anomaly Detection: Machine learning-based anomaly identification
- Pattern Analysis: Hourly and daily consumption patterns
- Savings Recommendations: Personalized tips with kWh, CO2, and USD savings
- Carbon Footprint: Calculate environmental impact
- Interactive Charts: Plotly-powered visualizations
- Export Reports: PDF and CSV formats
Solo hackathon project. For demo and presentation purposes.
