Bidirectional Long Short-Term Memory Networks for Rapid Fault Detection in Marine Hydrokinetic Turbines

Fault detection remains a key problem for reducing the operation and maintenance costs of marine hydrokinetic (MHK) turbines. With this in mind, a deep bidirectional long short-term memory (Bi-LSTM) network for rapid detection of MHK turbine faults is presented. The effectiveness of the proposed scheme is validated using simulated time-series sensor data gathered from a novel Fatigue, Aerodynamics, Structures, and Turbulence (FAST)-based MHK turbine simulation platform. Four-factor analysis of variance is performed along with post-hoc testing at the 95% confidence level to establish the model's capabilities. Operating conditions, input length, training pitch, and generalization pitch are used as factors in this experiment. Models are trained on data consisting of 500 baseline and 500 faulty examples of a single blade pitch imbalance. To evaluate generalization performance, models are applied, without fine-tuning, on the remaining levels of fault. Over 70% mean generalization accuracy in all cases, with an optimal accuracy over 95% given 1 second lengths of data, is observed. No significant difference is found due to varying operating conditions, indicating a robust model. Training on lower levels of fault demonstrated greater generalization capability. To the best of our knowledge, this is the first time Bi-LSTM has been applied as a building block, and that a deep learning-based method has been applied to only time-series sensor measurements for fault detection in MHK turbines.

[1]  Taghi M. Khoshgoftaar,et al.  Fusing Wavelet Features for Ocean Turbine Fault Detection , 2016 .

[2]  Diego Cabrera,et al.  Multimodal deep support vector classification with homologous features and its application to gearbox fault diagnosis , 2015, Neurocomputing.

[3]  Xin Zhou,et al.  Deep neural networks: A promising tool for fault characteristic mining and intelligent diagnosis of rotating machinery with massive data , 2016 .

[4]  Ruqiang Yan,et al.  Learning to Monitor Machine Health with Convolutional Bi-Directional LSTM Networks , 2017, Sensors.

[5]  Haibo He,et al.  Stacked Multilevel-Denoising Autoencoders: A New Representation Learning Approach for Wind Turbine Gearbox Fault Diagnosis , 2017, IEEE Transactions on Instrumentation and Measurement.

[6]  Li Jian Xun,et al.  State of health estimation combining robust deep feature learning with support vector regression , 2015, 2015 34th Chinese Control Conference (CCC).

[7]  Tianhao Tang,et al.  Imbalance fault detection of marine current turbine under condition of wave and turbulence , 2016, IECON 2016 - 42nd Annual Conference of the IEEE Industrial Electronics Society.

[8]  Wei Qiao,et al.  Bearing Fault Diagnosis for Direct-Drive Wind Turbines via Current-Demodulated Signals , 2013, IEEE Transactions on Industrial Electronics.

[9]  Eduard Muljadi,et al.  Review of Marine Hydrokinetic Power Generation and Power Plant , 2015 .

[10]  Randall David Wald.  Vibration analysis for ocean turbine reliability models , 2012 .

[11]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[12]  Halil Ceylan,et al.  A survey of health monitoring systems for wind turbines , 2015 .

[13]  Kevin A. Haas,et al.  Assessment of Energy Production Potential from Ocean Currents along the United States Coastline , 2013 .

[14]  Kuldip K. Paliwal,et al.  Bidirectional recurrent neural networks , 1997, IEEE Trans. Signal Process..

[15]  P. Jacobson.  Assessment and Mapping of the Riverine Hydrokinetic Resource in the Continental United States , 2012 .

[16]  Mohamed Benbouzid,et al.  An imbalance fault detection method based on data normalization and EMD for marine current turbines. , 2017, ISA transactions.

[17]  Yoshua Bengio,et al.  Learning long-term dependencies with gradient descent is difficult , 1994, IEEE Trans. Neural Networks.

[18]  Brennan T. Smith,et al.  Assessment of Energy Production Potential from Tidal Streams in the United States , 2011 .

[19]  J.A. Anderson,et al.  Neurocomputing: Foundations of Research@@@Neurocomputing 2: Directions for Research , 1992 .

[20]  Yide Wang,et al.  Blade imbalanced fault diagnosis for marine current turbine based on sparse autoencoder and softmax regression , 2018, 2018 33rd Youth Academic Annual Conference of Chinese Association of Automation (YAC).

[21]  F.R. Driscoll,et al.  A 20 KW open ocean current test turbine , 2008, OCEANS 2008.

[22]  Taghi M. Khoshgoftaar,et al.  Applying Feature Selection to Short Time Wavelet Transformed Vibration Data for Reliability Analysis of an Ocean Turbine , 2012, 2012 11th International Conference on Machine Learning and Applications.

[23]  Taghi M. Khoshgoftaar,et al.  Feature Selection for Optimization of Wavelet Packet Decomposition in Reliability Analysis of Systems , 2013, Int. J. Artif. Intell. Tools.

[24]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[25]  Antoine Picot,et al.  Current-Based Detection of Mechanical Unbalance in an Induction Machine Using Spectral Kurtosis With Reference , 2015, IEEE Transactions on Industrial Electronics.

[26]  Wei Qiao,et al.  A Survey on Wind Turbine Condition Monitoring and Fault Diagnosis—Part II: Signals and Signal Processing Methods , 2015, IEEE Transactions on Industrial Electronics.

[27]  Peter Tavner,et al.  Survey of commercially available condition monitoring systems for wind turbines. , 2014 .