Generative adversarial wavelet neural operator: Application to fault detection and isolation of multivariate time series data

Jyoti Rani,Tapas Tripura,Hariprasad Kodamana,Souvik Chakraborty

2024-01-09

Abstract:Fault detection and isolation in complex systems are critical to ensure reliable and efficient operation. However, traditional fault detection methods often struggle with issues such as nonlinearity and multivariate characteristics of the time series variables. This article proposes a generative adversarial wavelet neural operator (GAWNO) as a novel unsupervised deep learning approach for fault detection and isolation of multivariate time series processes.The GAWNO combines the strengths of wavelet neural operators and generative adversarial networks (GANs) to effectively capture both the temporal distributions and the spatial dependencies among different variables of an underlying system. The approach of fault detection and isolation using GAWNO consists of two main stages. In the first stage, the GAWNO is trained on a dataset of normal operating conditions to learn the underlying data distribution. In the second stage, a reconstruction error-based threshold approach using the trained GAWNO is employed to detect and isolate faults based on the discrepancy values. We validate the proposed approach using the Tennessee Eastman Process (TEP) dataset and Avedore wastewater treatment plant (WWTP) and N2O emissions named as WWTPN2O datasets. Overall, we showcase that the idea of harnessing the power of wavelet analysis, neural operators, and generative models in a single framework to detect and isolate faults has shown promising results compared to various well-established baselines in the literature.

Machine Learning,Signal Processing

What problem does this paper attempt to address?

This paper primarily addresses the problem of fault detection and isolation in complex systems. Traditional fault detection methods face challenges when dealing with nonlinear and multivariate time series data. To overcome this, the paper proposes a new unsupervised deep learning method called Generative Adversarial Wavelet Neural Operator (GAWNO) specifically for fault detection and isolation in multivariate time series processes. GAWNO combines the advantages of wavelet neural operators and generative adversarial networks (GANs) to effectively capture the distribution in the time series and the spatial dependencies among different variables. The workflow of GAWNO consists of two stages: first, training GAWNO with a dataset under normal operating conditions to learn its latent data distribution; then, based on the trained GAWNO, using the reconstruction error threshold method to detect and isolate faults. The paper validates the effectiveness of this method using datasets from the Tennessee Eastman Process and the Avedore wastewater treatment plant, and compares it with existing baseline methods, showing that GAWNO achieves better results in fault detection and isolation. In conclusion, this paper attempts to address how to utilize the innovative technology of GAWNO to improve the detection and isolation of faults in complex systems in industrial environments, aiming to enhance operational reliability and efficiency.

Generative adversarial wavelet neural operator: Application to fault detection and isolation of multivariate time series data

Data anomaly detection for structural health monitoring using a combination network of GANomaly and CNN

A Novel Unsupervised Graph Wavelet Autoencoder for Mechanical System Fault Detection

Wavelet neural operator: a neural operator for parametric partial differential equations

Effective time-series Data Augmentation with Analytic Wavelets for bearing fault diagnosis

Generalized Orthogonal Locality Preserving Projections for Nonlinear Fault Detection and Diagnosis

Multidimensional Non-Orthogonal Wavelet-Sigmoid Basis Function Neural Network For Dynamic Process Fault Diagnosis

Wavelet Neural Operator for solving parametric partial differential equations in computational mechanics problems

Improved Wavelet Neural Network Based on Hybrid Genetic Algorithm Applicationin on Fault Diagnosis of Railway Rolling Bearing.

Data-augmented wavelet capsule generative adversarial network for rolling bearing fault diagnosis

Wasserstein Generative Adversarial Network and Convolutional Neural Network (WG-CNN) for Bearing Fault Diagnosis

Fault Detection Based on Variational Autoencoders for Complex Nonlinear Processes

Non-Gaussian Process Fault Detection Based on Wavelet Kernel Clustering

Data Augmentation for Industrial Multivariate Time Series Via a Spatial and Frequency Domain Knowledge GAN

Fault Feature Recovery with Wasserstein Generative Adversarial Imputation Network with Gradient Penalty for Rotating Machine Health Monitoring under Signal Loss Condition

Explainable Graph Wavelet Denoising Network for Intelligent Fault Diagnosis

Filter-informed Spectral Graph Wavelet Networks for Multiscale Feature Extraction and Intelligent Fault Diagnosis

Denoising Graph Attention Wavelet Network: an interpretable multi-sensor information fusion diagnostic method for rotating systems

A novel fault localization with data refinement for hydroelectric units

Improved and Accurate Fault Diagnostic Model for Gas Turbine Based on 2D-Wavelet Transform and Generative Adversarial Network

Generative adversarial networks for data augmentation in machine fault diagnosis