Neural networks for wind turbine fault detection via current signature analysis
conference contributionposted on 2016-11-02, 10:11 authored by Raed Ibrahim, Jannis Weinert, Simon Watson
Cost-effective condition monitoring techniques are required to optimise wind turbine maintenance procedures. Current signature analysis investigates fault indications in the frequency spectrum of the electrical signal and is thereby able to detect mechanical faults without additional sensors. Due to the modern variable speed operation of wind turbines, fault frequencies are hidden in the non-stationary frequency spectra. In this work, artificial neural networks are applied to identify faults under transient conditions. The feasibility of the detection algorithm is demonstrated with a wind turbine SIMULINK model, which has been validated with experimental data. A framework is proposed for developing and training the algorithm for different rotational speeds. A simulation study demonstrates the ability of the algorithm not only to detect faults, but also to identify the strength of the faults as required for fault prognosis.
This project has partly received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement no 642108 (Advanced Wind Energy Systems Operation and Maintenance Expertise).
- Mechanical, Electrical and Manufacturing Engineering
Published inWindEurope Summit 2016
CitationIBRAHIM, R.K., TAUTZ-WEINERT, J. and WATSON, S.J., 2016. Neural networks for wind turbine fault detection via current signature analysis. Presented at the WindEurope Summit 2016, Hamburg, 27-29th Sept.
- VoR (Version of Record)
Publisher statementThis work is made available according to the conditions of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) licence. Full details of this licence are available at: https://creativecommons.org/licenses/by-nc-nd/4.0/
NotesThis is a conference paper.