Abstract
Fault diagnosis technology is crucial to ensure the long-term reliability of the industrial process control system. With the increase of industrial data availability, conventional monitoring approaches may not function well under the assumption that the training and the application data come from the same distribution. Following the intuition that industrial data exhibit time dependency and inherent complex characteristics, this paper proposes an adaptive sequential fault diagnosis method based on a tensor factorization layer merged with deep neural network model (TF-DNN). Tensor representation is firstly applied to preserve the number of the raw data entries and their sequential dependence between observations. Multilinear mapping with tensor-to-tensor projection is then to transform the input and hidden tensor to the low-dimensional tensors, which makes the learned representations sparser with the grid-like structures. With the tensor factorization layers, the proposed deep network shares efficient knowledge across the spatiotemporal features of fault data. The outstanding performance of this method is demonstrated and compared to the existing models in the benchmark Tennessee Eastman process and a real industrial methanol plant.
Highlights
With the progress of computer technique, electronics and information technology, the modern industrial control systems tend to be large-scale, multivariable and complex
In this paper, a deep network based on tensor factorization layer has been proposed for fault detection and diagnosis in industrial process
(2) By stacking the multiple tensor factorization (TF) layers, TF-deep neural network (DNN) is trained with an end-to-end manner which synchronously updates the parameters of the feature extraction and classifier
Summary
With the progress of computer technique, electronics and information technology, the modern industrial control systems tend to be large-scale, multivariable and complex. For the emergence of ‘‘data rich but information poor’’ problem [7] in the process monitoring task, latent variable models are suitable to extract a low-dimensional representation from routine operating data. L. Luo et al.: Deep Learning With Tensor Factorization Layers for Sequential Fault Diagnosis and Industrial Process Monitoring which is implicitly defined by a kernel function. Zhang and Zhao [13] exploited the deep belief network (DBN) to build the FDD strategy for a chemical benchmark process They utilized the mutual information technology to overcome the curse of dimensionality and complexities in each sub-network. The main contributions of this paper are outlined as follows: 1) the proposed framework utilizes the multilinear discriminant analysis (MDA) on the layers to widen the neural network and share the parameters across different the temporal features. Let Xi,j represents the j-th sample from class ci, ni is the
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call