Abstract

ABSTRACT Information is key. Offshore wind farms are installed with supervisory control and data acquisition systems (SCADA) gathering valuable information. Determining the precise condition of an asset is essential on achieving the expected operational lifetime and efficiency. Equipment fault detection is necessary to achieve this. This paper presents a systematic literature review of machine learning methods applied to condition monitoring systems, using both vibration information and SCADA data together. Starting with conventional methods using vibration models, such as Fast-Fourier transforms to five prominent supervised learning regression models; Artificial neural network, support vector regression, Bayesian network, random forest and K-nearest neighbour. This review specifically looks at how conventional vibration data can be combined with SCADA data to determine the assets condition.

Highlights

  • The total capacity in Europe of installed wind power sat at 18,499 MW in 2018 which is an increase of 2649 MW from 2017 (WindEurope 2019)

  • The cost of maintenance relative to the levelised cost of energy (LCOE) is significantly increased compared to onshore. It is reported in the North-Sea that the operations and maintenance (O&M) cost between 20% and 25% of the LCOE compared to around 12% onshore (Röckmann, Lagerveld, and Stavenuiter 2017; Tavner 2012)

  • This article has given a systematic review of how monitoring for maintenance can be carried out, detailing specific measures that can be taken for condition-based maintenance

Read more

Summary

Introduction

The total capacity in Europe of installed wind power sat at 18,499 MW in 2018 which is an increase of 2649 MW from 2017 (WindEurope 2019). Companies monitor several parameters including; vibrations, oil quality and temperatures in some of the main assemblies (MartinezLuengo, Kolios, and Wang 2016) This information is used to infer the health of the assets to determine the remaining useful life or to determine if scheduled maintenance is required based on the monitored irregularities. This paper is a continuation of Martinez-Luengo, Kolios, and Wang (2016), which is carried out a detailed review of CMS, following the statistical pattern recognition paradigm Developing this idea, this report seeks to understand the types of maintenance procedures implemented in offshore wind engineering. One of the major drawbacks of vibration-based methods is that the results are difficult to interpret without the help of an expert Combining both sets of information compliment the analysis for easier insight and implementation on improving offshore-wind turbine operational management. Rounding the remaining sections with a discussion and conclusion separately

Method
Machine-learning
Supervised learning
Unsupervised learning
Semi-supervised learning
Reinforcement learning
Models
Artificial neural network
Support vector regression
K-nearest neighbour
Dynamic Bayesian network
Gaussian process regression
Data-driven decision making for wind turbine operational management
Acoustic condition modelling
SCADA modelling
Trending
Alarm assessment
Performance monitoring
Discussion
Findings
Limitations
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call