HybridSmart

Hybrid Energy System for Sustainable Smart Homes

Simple prototype that shows how a smart home can choose between solar, wind, battery and grid.

---

Project Overview

HybridSmart is a small local app built for a dissertation-style demo.

It decides which source to use based on:

• how much solar / wind is available
• house load
• battery level
• weather
• time of day

The logic is plain if/else. No machine learning.

---

Features

• Dashboard with sliders
• Short explanation after each run ("Why this decision was selected")
• Decision flow diagram
• Compare two scenarios
• Decision log + CSV download
• Gauges for renewable / grid / battery / efficiency
• 10 fixed test cases
• PDF report download
• Technical Information page
• About Project page
• Sample CSV with ~30 rows

---

Installation

cd HybridSmart
python3 -m venv venv
source venv/bin/activate        # Windows: venv\Scripts\activate
pip install -r requirements.txt
streamlit run app.py

If the flow diagram does not show, install Graphviz on your system too
(brew install graphviz on Mac). The app still works without it.

---

Folder Structure

HybridSmart/
├── app.py
├── decision_engine.py
├── utils.py
├── report_generator.py
├── data/
│   └── HESS_Dataset.csv
├── requirements.txt
└── README.md

---

Dataset

The app uses data/HESS_Dataset.csv (about 1000 hourly rows).

Original columns include:

• Timestamp
• Solar_Power_kW
• Wind_Power_kW
• Load_Demand_kW
• Battery_SoC_%
• Grid_Power_kW
• and related power fields

These are mapped for the decision engine / UI as:

• Solar (kW) / Wind (kW) / Load (kW) / Battery (%)
• Time of Day (from timestamp hour)
• Weather (estimated from solar level, because the CSV has no weather column)

Tables in the app show clean labels (no underscores), for example: Solar Used (kW), Selected Source, Grid in Dataset (kW).

---

Decision Logic

1. Adjust solar/wind using weather and time
2. Use solar first
3. Add wind if needed
4. Use battery if still short and battery > 30%
5. Use grid for leftover demand
6. Charge battery if there is spare renewable power

Rough estimate factors used in the demo:

• Grid cost ≈ £0.28 / kWh
• Grid CO₂ ≈ 0.4 kg / kWh

---

Pages

Page	What it does
Dashboard	Run a simulation
Scenario Comparison	Compare A vs B
Decision Log	See past runs + CSV
Sample Scenarios	Run rows from the CSV
Results Summary	Averages over all samples
Testing	10 preset tests
Technical Information	Architecture / tech list
About Project	Objective, limits, future work

---

Colour Key

• Green = renewable
• Orange = battery
• Red = grid

---

Screenshots

Add screenshots here later:

• Dashboard
• Explanation panel
• Comparison page
• Testing page
• PDF report

---

Future Improvements

• Hourly / daily simulation
• Better battery model
• Different household profiles
• Stronger test checks
• More detail in the PDF

---

Note

Cost and CO₂ figures are estimates for demonstration only.

License

All rights reserved.

• Commercial use is not allowed
• Use without written permission is not allowed

See the LICENSE file for full terms. Contact: mansisharmaseo@gmail.com