Wavelet-Like Transform to Optimize the Order of an Autoregressive Neural Network Model to Predict the Dissolved Gas Concentration in Power Transformer Oil from Sensor Data

Dissolved gas analysis (DGA) is one of the most important methods to analyze fault in power transformers. In general, DGA is applied in monitoring systems based upon an autoregressive model; the current value of a time series is regressed on past values of the same series, as well as present and past values of some exogenous variables. The main difficulty is to decide the order of the autoregressive model; this means determining the number of past values to be used. This study proposes a wavelet-like transform to optimize the order of the variables in a nonlinear autoregressive neural network to predict the in oil dissolved gas concentration (DGC) from sensor data. Daubechies wavelets of different lengths are used to create representations with different time delays of ten DGC, which are then subjected to a procedure based on principal components analysis (PCA) and Pearson’s correlation to find out the order of an autoregressive model. The representations with optimal time delays for each DGC are applied as input in a multi-layer perceptron (MLP) network with backpropagation algorithm to predict the gas at the present and future times. This approach produces better results than choosing the same time delay for all inputs, as usual. The forecasts reached an average mean absolute percentage error (MAPE) of 5.763%, 1.525%, 1.831%, 2.869%, and 5.069% for C2H2, C2H6, C2H4, CH4, and H2, respectively.

[1]  Xiangning Lin,et al.  Combined Forecasting Method of Dissolved Gases Concentration and Its Application in Condition-Based Maintenance , 2019, IEEE Transactions on Power Delivery.

[2]  Chia-Hung Lin,et al.  Dissolved gases forecast to enhance oil‐immersed transformer fault diagnosis with grey prediction–clustering analysis , 2011, Expert Syst. J. Knowl. Eng..

[3]  Huakun Que,et al.  Grey relational analysis using Gaussian process regression method for dissolved gas concentration prediction , 2019, Int. J. Mach. Learn. Cybern..

[4]  Ke Wang,et al.  Optimal dissolved gas ratios selected by genetic algorithm for power transformer fault diagnosis based on support vector machine , 2016, IEEE Transactions on Dielectrics and Electrical Insulation.

[5]  Peter Tiño,et al.  Learning long-term dependencies in NARX recurrent neural networks , 1996, IEEE Trans. Neural Networks.

[6]  Xiaohui Yang,et al.  Bio-Inspired PHM Model for Diagnostics of Faults in Power Transformers Using Dissolved Gas-in-Oil Data , 2019, Sensors.

[7]  Meng Joo Er,et al.  NARMAX time series model prediction: feedforward and recurrent fuzzy neural network approaches , 2005, Fuzzy Sets Syst..

[8]  Q. H. Wu,et al.  Condition Monitoring and Assessment of Power Transformers Using Computational Intelligence , 2011 .

[9]  Eugen Diaconescu,et al.  The use of NARX neural networks to predict chaotic time series , 2008 .

[10]  H. Kaiser,et al.  A Study Of A Measure Of Sampling Adequacy For Factor-Analytic Correlation Matrices. , 1977, Multivariate behavioral research.

[11]  Imed Riadh Farah,et al.  Wavelet Transform Application for/in Non-Stationary Time-Series Analysis: A Review , 2019, Applied Sciences.

[12]  Hanbo Zheng,et al.  A novel model based on wavelet LS-SVM integrated improved PSO algorithm for forecasting of dissolved gas contents in power transformers , 2018 .

[13]  Gehao Sheng,et al.  Prediction Method for Power Transformer Running State Based on LSTM_DBN Network , 2018, Energies.

[14]  K. Gnana Sheela,et al.  Review on Methods to Fix Number of Hidden Neurons in Neural Networks , 2013 .

[15]  Haroldo de Faria,et al.  A review of monitoring methods for predictive maintenance of electric power transformers based on dissolved gas analysis , 2015 .

[16]  Fábio Henrique Pereira,et al.  Disease spreading in complex networks: A numerical study with Principal Component Analysis , 2017, Expert Systems with Applications.

[17]  Yuan Tian,et al.  Dissolved Gas Analysis in Transformer Oil Using Pt-Doped WSe2 Monolayer Based on First Principles Method , 2019, IEEE Access.

[18]  Rui Yang,et al.  Diagnosis of solid insulation deterioration for power transformers with dissolved gas analysis-based time series correlation , 2015 .

[19]  Suwarno,et al.  Transformer Paper Expected Life Estimation Using ANFIS Based on Oil Characteristics and Dissolved Gases (Case Study: Indonesian Transformers) , 2017 .

[20]  D. Gogolewski.  Influence of the edge effect on the wavelet analysis process , 2020 .

[21]  J. Fuhr.  Condition assessment of power transformers , 2012, 2012 IEEE International Conference on Condition Monitoring and Diagnosis.

[22]  Khmais Bacha,et al.  Power transformer fault diagnosis based on dissolved gas analysis by support vector machine , 2012 .

[23]  Hava T. Siegelmann,et al.  Computational capabilities of recurrent NARX neural networks , 1997, IEEE Trans. Syst. Man Cybern. Part B.

[24]  Reza Effatnejad,et al.  Using Dissolved Gas Analysis Results to Detect and Isolate the Internal Faults of Power Transformers by Applying a Fuzzy Logic Method , 2017 .

[25]  A. M. Guimarães,et al.  A PCA and SPCA based procedure to variable selection in agriculture , 2014 .

[26]  Junhui Zhao,et al.  Dissolved Gases Forecasting Based on Wavelet Least Squares Support Vector Regression and Imperialist Competition Algorithm for Assessing Incipient Faults of Transformer Polymer Insulation , 2019, Polymers.

[27]  Tao Yu,et al.  Dissolved Gas Analysis Principle-Based Intelligent Approaches to Fault Diagnosis and Decision Making for Large Oil-Immersed Power Transformers: A Survey , 2018 .

[28]  Gilberto Francisco Martha de Souza,et al.  Nonlinear Autoregressive Neural Network Models for Prediction of Transformer Oil-Dissolved Gas Concentrations , 2018 .

[29]  I. Daubechies.  Orthonormal bases of compactly supported wavelets , 1988 .

[30]  M. Bartlett.  Properties of Sufficiency and Statistical Tests , 1992 .

[31]  Mario Manana,et al.  Dissolved Gas Analysis Equipment for Online Monitoring of Transformer Oil: A Review , 2019, Sensors.

[32]  Osama E. Gouda,et al.  Condition assessment of power transformers based on dissolved gas analysis , 2019 .

[33]  Qingguo Chen,et al.  Experimental Study on Trap Characteristics of Nano-Montmorillonite Composite Pressboards , 2018, Energies.

[34]  Debangshu Dey,et al.  Recent Trends in the Condition Monitoring of Transformers , 2013 .

[35]  Sun-Yuan Kung,et al.  A delay damage model selection algorithm for NARX neural networks , 1997, IEEE Trans. Signal Process..

[36]  I. Daubechies,et al.  Factoring wavelet transforms into lifting steps , 1998 .

[37]  Hao Yu,et al.  Selection of Proper Neural Network Sizes and Architectures—A Comparative Study , 2012, IEEE Transactions on Industrial Informatics.

[38]  A. Abu-Siada,et al.  A review of dissolved gas analysis measurement and interpretation techniques , 2014, IEEE Electrical Insulation Magazine.