Research on Performance Prediction Model of Wind Turbine Gearbox Lubricating Oil Based on Deep Learning
DOI:
https://doi.org/10.65455/2s6npy06Keywords:
Gearbox Lubricating Oil, Antioxidant Performance Prediction, Data Preprocessing, Feature Extraction, Machine Learning, SNV-PCA-BP Model, Rapid DetectionAbstract
Gearbox lubricating oil oxidation degradation severely impairs the operational stability of wind turbines and increases maintenance costs. Conventional detection methods (e.g., Rotating Pressure Vessel Oxidation Test (RPVOT), Pressure Differential Scanning Calorimetry (PDSC)) have high instrument dependence and long testing cycles, and may not meet on-site rapid detection demands. This study took in-service industrial gear oil samples as the research object, selected five physicochemical indices as input variables, and systematically optimizes data preprocessing (Standard Normal Variate (SNV)), feature extraction (Principal Component Analysis (PCA)), and machine learning algorithms (Back Propagation Neural Network (BPNN)/Support Vector Machine (SVM)/Random Forest (RF)). The proposed SNV-PCA-BP hybrid model achieved excellent predictive performance with a R2 of 0.9960 and Residual Prediction Deviation (RPD) of 6.1247, which is 360 times more efficient (defined as single-sample detection time) than traditional RPVOT/PDSC methods based on parallel tests of 62 samples. This model provides a low-cost and reliable technical support for the predictive maintenance of wind turbine gearboxes.
References
[1]DÍAZ-DÍAZ A M, LÓPEZ-BECEIRÓ J J, DÍAZ R P A, et al. Oxidative stability of edible oils: linking rancimat to PDSC results. Journal of Food Measurement and Characterization, 2026, 20:1139-1150.
[2]JIANG S, XIAO Y, LI Q, et al. SERS technology in virus Detection: Advances, challenges, and future perspectives. Biosensors and Bioelectronics, 2025, 290: 117902.
[3]ABDELFATTAH W, ABOSAODA M K, DOSHI H, et al. Development of data driven models to accurately estimate density of fatty acid ethyl esters. Scientific Reports, 2025, 15: 30961.
[4]SAGRALOFF N, DOBLER A, TOBIE T, et al. Development of an Oil Free Water-Based Lubricant for Gear Applications. Lubricants, 2019, 7(4): 33.
[5]GRIBOK A, HINES J W, URMANOV A M, et al. Heuristic, systematic, and informational regularization for process monitoring. International Journal of Intelligent Systems, 2002, 17.
[6]ROYLANCE B J, POCOCK G. Wear studies through particle size distribution I: Application of the Weibull distribution to ferrography. Wear, 1983, 90(1): 113–136.
[7]GONCALVES I M, MURILLO M, GONZÁLEZ A M. Determination of metals in used lubricating oils by AAS using emulsified samples. Talanta, 1998, 47(4): 1033–1042.
[8]KROGSØE K, ERIKSEN R L, HENNEBERG M. Performance of a light extinction based wear particle counter under various contamination levels. Sensors and Actuators A: Physical, 2021, 331: 112956.
[9]ABBASI S, JANSSON A, OLANDER L, et al. A pin-on-disc study of the rate of airborne wear particle emissions from railway braking materials. Wear, 2012, 284–285: 18–29.
[10]AHMED H E, HASSAN M, NOUR M, et al. Lubricant Oils as a Certified Reference Material for Cleveland Open Cup Flash Point Testers. MAPAN, 2017, 32: 215–222.
[11]MACIÁN V, TORMOS B, GÓMEZ Y, et al. Proposal of an FTIR Methodology to Monitor Oxidation Level in Used Engine Oils: Effects of Thermal Degradation and Fuel Dilution. Tribology Transactions, 2012, 69.
[12]ESQUIVEL M M, RIBEIRO M A, BERNARDO-GIL M G. Relations between Oxidative Stability and Antioxidant Content in Vegetable Oils Using an Accelerated Oxidation Test - Rancimat. 2009, 4(4).
[13]GAO T, HU L, JIA Z, et al. SPXYE: an improved method for partitioning training and validation sets. Cluster Computing, 2019, 22: 3069–3078.
[14]RAWAT S, CUI H, XIE Y, et al. An improved framework for multi-objective optimization of cementitious composites using Taguchi-TOPSIS approach. Expert Systems with Applications, 2025, 272: 126732.
[15]DHANOA M, LÓPEZ S, SANDERSON R, et al. Methodology adjusting for least squares regression slope in the application of multiplicative scatter correction to near‐infrared spectra of forage feed samples. Journal of Chemometrics, 2023, 37(11):e3511.
[16]SHRESTHA B, POSOM J, SIRISOMBOON P, et al. Comprehensive Assessment of Biomass Properties for Energy Usage Using Near-Infrared Spectroscopy and Spectral Multi-Preprocessing Techniques. Energies, 2023, 16(14): 5351.
[17]TALEBANPOUR BAYAT E, HEMMATEENEJAD B, AKHOND M, et al. On the dependency between principal components: Application to determine the rank of a matrix in an evolutionary process. Journal of Chemometrics, 2018, 33: e3102.
[18]SOARES A, GALVÃO FILHO A, GALVÃO R. Improving the computational efficiency of the successive projections algorithm by using a sequential regression implementation: a case study involving nir spectrometric analysis of wheat samples. Journal of the Brazilian Chemical Society, 2010, 21(4): 760–763.
[19]YARAHMADI M N, MIRHASSANI S A, HOOSHMAND F. Handling the significance of regression coefficients via optimization. Expert Systems with Applications, 2024, 238: 121910.
[20]BAGHOLIZADEH M, NASAJPOUR-ESFAHANI N, PIRMORADIAN M, et al. Using different machine learning algorithms to predict the rheological behavior of oil SAE40-based nano-lubricant in the presence of MWCNT and MgO nanoparticles. Tribology International, 2023, 187: 108759.
[21]BARBERO JIMÉNEZ Á, LÓPEZ LÁZARO J, DORRONSORO J R. Finding optimal model parameters by deterministic and annealed focused grid search. Neurocomputing, 2009, 72(13–15): 2824–2832.
[22]MA Y, HUANG M, WAN J, et al. Prediction model of DnBP degradation based on BP neural network in AAO system. Bioresource Technology, 2011, 102(5): 4410–4415.
[23]PANG H X, DONG W D, XU Z H, et al. Novel Linear Search for Support Vector Machine Parameter Selection. Journal of Zhejiang University-SCIENCE C (Computers & Electronics), 2011, 12: 885–896.
[24]WAINER J, FONSECA P. How to tune the RBF SVM hyperparameters? An empirical evaluation of 18 search algorithms. Artificial Intelligence Review, 2021, 54: 4771–4797.
[25]LIU Y, ZHOU Y, WEN S, et al. A Strategy on Selecting Performance Metrics for Classifier Evaluation. International Journal of Mobile Computing and Multimedia Communications, 2014, 6(4): 20–35.
[26]CHICCO D, WARRENS M J, JURMAN G. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation. PeerJ Computer Science, 2021, 7: e623.
[27]MUKHIN A A, SKRYABINA A E, FADEEEV V K, et al. Express method for determining acid number of lubricating oils for gas-pumping units. Chemistry and Technology of Fuels and Oils, 2013, 49: 359–361.
[28]LACROIX-ANDRIVET O, HUBERT M, LOUTELIER-BOURHIS C, et al. Characterization of Base Oil and Additive Oxidation Products from Formulated Lubricant by Ultra-High Resolution Mass Spectrometry. Lubricants, 2023, 11(8): 345.
[29]SALEM N, HUSSEIN S. Data dimensional reduction and principal components analysis. Procedia Computer Science, 2019, 163: 292–299.
[30]DOS SANTOS E M, SABOURIN R, MAUPIN P. Overfitting cautious selection of classifier ensembles with genetic algorithms. Information Fusion, 2009, 10(2): 150–162.
[31]GRISANTI E, TOTSKA M, HUBER S, et al. Dynamic localized SNV, Peak SNV, and partial peak SNV: Novel standardization methods for preprocessing of spectroscopic data used in predictive modeling. Journal of Spectroscopy, 2018, 2018: 5037572.
[32]UMPRECHT A, FONSECA DIAZ V, HÜPFL B, et al. Unsupervised optimization of spectral pre-processing selection to achieve transfer of Raman calibration models. Measurement, 2025, 255: 117906.
[33]HATHOUT R M. Using principal component analysis in studying the transdermal delivery of a lipophilic drug from soft nano-colloidal carriers to develop a quantitative composition effect permeability relationship. Pharmaceutical Development and Technology, 2014, 19(5): 598–604.
[34]ZHANG J, CAO J, WANG L. Robust Bayesian functional principal component analysis. Statistics and Computing, 2025, 35: 46.
[35]LONG J, LI T, YANG M, et al. Hybrid Strategy Integrating Variable Selection and a Neural Network for Fluid Catalytic Cracking Modeling. Industrial & Engineering Chemistry Research, 2019, 58(1): 247–258.
[36]BURKHART C, JOHANSSON J, UKONSAARI J, et al. Performance of lubricating oils for wind turbine gear boxes and bearings. Proceedings of the Institution of Mechanical Engineers, Part J: Journal of Engineering Tribology, 2017, 232.
Downloads
Published
License
Copyright (c) 2026 Applied Artificial Intelligence Research
This work is licensed under a Creative Commons Attribution 4.0 International License.