Skip to content

An SVM-based Fault Detector with a sliding window for rising stability on output. Electrical Engineering final B.Sc. project, Petroleum University of Technology.

Notifications You must be signed in to change notification settings

HabibNaeimi/Data-Driven-Fault-Detection

Repository files navigation


Data-Driven Fault Detection in Autonomous Drone

Designing and Implementing a Fault detector and a Sliding Window
Training and Testing on AirLab Failure and Anomaly(ALFA) Dataset
An Electrical Engineering final B.Sc. Project
Explore the docs »

Report Bug & Request Feature

About The Project

This repository contains the Python implementation of "Data-Driven Fault Detection in Autonomous Drones," an Electrical Engineering final B.Sc. project by Habibollah Naeimi at the Petroleum University of Technology.

This project has developed a fault detector based on the Support Vector Machine (SVM) classification method. The data are from the ALFA dataset by AirLab Failure and Anomaly at Carnegie Mellon University in Pittsburgh, Pennsylvania, USA. Three Fault detectors have been developed during this project for three datasets of engine faults, including GPS, Compass, and Global Location data.

Also, a sliding Window is developed for each detector in order to make the detector's output smoother and raise its stability.

Implementation of these classifiers is in Python programming language by using scikit-learn library. The sliding windows are also implemented in Python.

About

An SVM-based Fault Detector with a sliding window for rising stability on output. Electrical Engineering final B.Sc. project, Petroleum University of Technology.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published