Solar Microgrid Cyberattack Detection System

A machine learning-based intrusion detection system (IDS) for solar microgrid infrastructure, developed as a capstone project for Pinellas County, Florida. The system detects six classes of cyberattacks on photovoltaic (PV) energy systems using a feedforward neural network achieving 96.23% classification accuracy.

---

Demo

📽️ Demo

---

Project Overview

Critical infrastructure such as solar microgrids is increasingly vulnerable to cyberattacks that can disrupt energy generation, damage equipment, and destabilize the grid. This project addresses that threat by building a machine learning powered detection system trained on simulated attack data from two real Pinellas County solar sites.

Location: Pinellas County, Florida
Sites: Market Parking Lot PV System, Public Park PV System
Simulation Tool: HelioScope
Academic Institution: James Madison University

---

Attack Classes Detected

The model classifies normal operation and six cyberattack scenarios:

Class	Attack Type
0	Normal Operation
1	False Data Injection
2	Inverter Manipulation
3	Output Drop
4	Sensor Spoofing
5	Clipping Attack

---

Model Performance

Metric	Score
Accuracy	96.23%
Architecture	Feedforward Neural Network
Framework	TensorFlow / Keras
Training Data	HelioScope simulation + synthetic attack injection

---

System Architecture

HelioScope Simulation Data ↓ Synthetic Attack Injection ↓ Feature Engineering + Preprocessing ↓ Feedforward Neural Network (TensorFlow/Keras) ↓ Flask Operator Dashboard ├── Role-Based Authentication ├── Real-Time Risk Scoring ├── Chart.js Visualizations └── Attack Classification Display

Tech Stack

Component	Technology
Machine Learning	Python, TensorFlow, Keras
Web Framework	Flask
Visualizations	Chart.js
Authentication	Role-based access control (RBAC)
Data Simulation	HelioScope, HOMER Pro
Data Processing	Pandas, NumPy, Scikit-learn

---

Running the Project

Prerequisites

pip install -r requirements.txt

Run the Flask Dashboard

python app.py

Navigate to http://localhost:5000 in your browser.

Run in Google Colab

Open UpdatedCapstone_3_2.ipynb in Google Colab and run all cells.

---

Security & Compliance

This project incorporates relevant regulatory standards for solar energy systems:

• NEC Articles — National Electrical Code compliance for PV systems
• Florida Statutes — State-level energy infrastructure regulations
• ICS/OT Security — Industrial Control System threat modeling