WindTurbineClassification
My master's dissertation on wind turbine fault prediction using machine learning
Science Score: 13.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
○codemeta.json file
-
○.zenodo.json file
-
✓DOI references
Found 2 DOI reference(s) in README -
○Academic publication links
-
○Committers with academic emails
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (11.7%) to scientific vocabulary
Keywords
Repository
My master's dissertation on wind turbine fault prediction using machine learning
Basic Info
Statistics
- Stars: 58
- Watchers: 3
- Forks: 11
- Open Issues: 0
- Releases: 2
Topics
Metadata Files
README.md
WindTurbineClassification
Specification of 'normal' wind turbine operating behaviour for rapid anomaly detection: through the use of machine learning algorithms
by Nithiya Streethran (nmstreethran@gmail.com)
This work is derived from my dissertation for the degree of Master of Science (MSc) in Renewable Energy Engineering at Heriot-Watt University, which was completed during a technical placement at Natural Power between May and August 2017.
Unfortunately, the datasets are proprietary industry data and I do not own the rights to distribute them to the public. Please do not contact me to request the datasets.
Abstract
Maximising the economic effectiveness of a wind farm is essential in making wind a more economic source of energy. This effectiveness can be increased through the reduction of operation and maintenance costs, which can be achieved through continuously monitoring the condition of wind turbines. An alternative to expensive condition monitoring systems, which can be uneconomical especially for older wind turbines, is to implement classification algorithms on supervisory control and data acquisition (SCADA) signals, which are collected in most wind turbines. Several publications were reviewed, which were all found to use separate algorithms to predict specific faults in advance. In reality, wind turbines tend to have multiple faults which may happen simultaneously and have correlations with one another. This project focusses on developing a methodology to predict multiple wind turbine faults in advance simultaneously by implementing classification algorithms on SCADA signals for a wind farm with 25 turbines rated at 2,500 kW, spanning a period of 30 months. The data, which included measurements of wind speed, active power and pitch angle, was labelled using corresponding downtime data to detect normal behaviour, faults and varying timescales before a fault occurs. Three different classification algorithms, namely decision trees, random forests and k nearest neighbours were tested using imbalanced and balanced training data, initially to optimise a number of hyperparameters. The random forest classifier produced the best results. Upon conducting a more detailed analysis on the performance of specific faults, it was found that the classifier was unable to detect the varying timescales before a fault with accuracy comparable to that of normal or faulty behaviour. This could have been due to the SCADA data, which are used as features, being unsuitable for detecting the faults, and there is potential to improve this by balancing only these classes.
Keywords: wind turbine, classification algorithm, SCADA, fault detection, condition monitoring
Scripts
| Source | Output
------ | -- | --
Process SCADA and downtime data | 
| 
Downtime categories | |
Merge SCADA and downtime data |
Power curves for all turbines | |
License
Unless otherwise stated:
- Code and scripts are licensed under the MIT License.
- Content, images, and documentation are licensed under a Creative Commons Attribution 4.0 International (CC-BY-4.0) License.
Project badges are generated using Shields.io and Simple Icons.
Owner
- Name: Nithiya Streethran
- Login: nmstreethran
- Kind: user
- Location: Ireland
- Website: https://nithiya.gitlab.io
- Repositories: 14
- Profile: https://github.com/nmstreethran
GitHub Events
Total
- Watch event: 6
Last Year
- Watch event: 6
Committers
Last synced: 6 months ago
Top Committers
| Name | Commits | |
|---|---|---|
| nmstreethran | n****n@g****m | 58 |
| Nithiya Streethran | n****t@g****m | 29 |
| Nithiya Streethran | 3****n@u****m | 9 |
Issues and Pull Requests
Last synced: 6 months ago
All Time
- Total issues: 0
- Total pull requests: 1
- Average time to close issues: N/A
- Average time to close pull requests: 10 minutes
- Total issue authors: 0
- Total pull request authors: 1
- Average comments per issue: 0
- Average comments per pull request: 0.0
- Merged pull requests: 1
- Bot issues: 0
- Bot pull requests: 0
Past Year
- Issues: 0
- Pull requests: 0
- Average time to close issues: N/A
- Average time to close pull requests: N/A
- Issue authors: 0
- Pull request authors: 0
- Average comments per issue: 0
- Average comments per pull request: 0
- Merged pull requests: 0
- Bot issues: 0
- Bot pull requests: 0
Top Authors
Issue Authors
Pull Request Authors
- nmstreethran (1)
Top Labels
Issue Labels
Pull Request Labels
Packages
- Total packages: 2
- Total downloads: unknown
-
Total dependent packages: 0
(may contain duplicates) -
Total dependent repositories: 0
(may contain duplicates) - Total versions: 4
proxy.golang.org: github.com/nmstreethran/WindTurbineClassification
- Documentation: https://pkg.go.dev/github.com/nmstreethran/WindTurbineClassification#section-documentation
- License: mit
-
Latest release: v1.1.0
published about 5 years ago
Rankings
proxy.golang.org: github.com/nmstreethran/windturbineclassification
- Documentation: https://pkg.go.dev/github.com/nmstreethran/windturbineclassification#section-documentation
- License: mit
-
Latest release: v1.1.0
published about 5 years ago
Rankings
Dependencies
- imbalanced-learn *
- jupyterlab *
- matplotlib *
- openpyxl *
- pandas *
- imbalanced-learn
- jupyterlab
- matplotlib
- openpyxl
- pandas
- python 3.*