Real-Time Anomaly Detection for Signal Analysis

Project Overview

This project implements a real-time anomaly detection system using a deep learning autoencoder model. The system is designed to detect anomalies in time-series signals, specifically simulating vibration sensor data from a mechanical system.

Key Features

• Synthetic signal generation
• LSTM-based autoencoder for anomaly detection
• Real-time visualization of signals
• Interactive anomaly injection and detection

Project Structure

• training.py: Data generation, model training, and saving
• test_random.py: Real-time anomaly detection and visualization script

Prerequisites

Dependencies

• Python 3.8+
• Libraries:
  • numpy
  • pandas
  • tensorflow
  • scikit-learn
  • plotly
  • dash
  • keras

Installation

1. Clone the repository:

git clone https://github.com/yourusername/signal-anomaly-detector.git
cd signal-anomaly-detector

2. Create a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

3. Install required packages:

pip install numpy pandas tensorflow scikit-learn plotly dash keras

Usage

Training the Model

1. Run the training script to generate synthetic data and train the model:

python training.py

This will generate and save anomaly_detector_model.h5

Running Real-Time Anomaly Detection

1. Start the real-time anomaly detection:

python test_random.py

2. During runtime:

• The script will open a live dashboard in your default web browser
• Press 'a' + Enter to manually inject a random anomaly
• Anomalies will be highlighted in red on the graph

Model Details

• Model Type: LSTM Autoencoder
• Input: Time-series signals with a 10-step lookback
• Training Data: Synthetic sinusoidal signals with noise
• Anomaly Detection: Based on reconstruction error threshold

Contributing

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Preparing to Push to GitHub

1. Initialize Git repository (if not already done):

git init

2. Add files:

git add training.py test_random.py README.md

3. Commit changes:

git commit -m "Initial commit: Real-time Anomaly Detection project"

4. Create a new repository on GitHub

5. Link local repository to GitHub:

git remote add origin https://github.com/yourusername/signal-anomaly-detector.git

6. Push to GitHub:

git branch -M main
git push -u origin main

License

Distributed under the MIT License. See LICENSE for more information.

Contact

stanleyoz

Project Link: https://github.com/stanleyoz/signal_anomaly-detector