Distributed process monitoring framework based on decomposed modified partial least squares

Abstract

With the growing complexity of industrial processes, the scale of production processes tends to be large. The significant amount of measurement data in large‐scale processes poses challenges in data collection, management, and storage. In order to perform effective process monitoring in large‐scale processes, the distributed process monitoring strategy is widely applied. Meanwhile, product quality is an important indicator for industrial production. Therefore, a novel quality‐based distributed process monitoring scheme is proposed. Firstly, the Girvan‐Newman (GN) algorithm in complex network divides process variables into multiple sub‐blocks. Secondly, greedy algorithm‐based high‐dimensional mutual information (HDMI) is used to extract quality‐related variables in each sub‐block, through which the irrelevant and redundant variables are eliminated. Thirdly, the decomposed modified partial least squares (DMPLS) approach is used to detect whether a fault is quality‐related or not in each sub‐block. Finally, the Bayesian inference strategy is adopted to combine the detection results of all sub‐blocks. The effectiveness of the distributed DMPLS approach is illustrated through a numerical simulation and the Tennessee Eastman (TE) process. The results show the superiority of our proposed monitoring scheme.

Distributed temporal–spatial neighbourhood enhanced variational autoencoder for multiunit industrial plant‐wide process monitoring

“…The temporal information reconstructed sample is obtained by Equation (10), and the dynamic feature is obtained from the time series-related information. At the same time, the obtained time information reconstructed sample is used to expand the dynamic feature of the current sample, so the input sample at the current time can be expressed as x re-temporal b,t…”

“…[7,8] PLS establishes the linear relationships between process variables and quality variables, and is commonly used for quality-related variable modelling and monitoring. [9,10] Industrial processes are typically composed of several subprocess units, where the number of measured variables can be in the hundreds or thousands, and their interactions are complex. From the perspective of the modelling framework, traditional MSPMs mostly belong to centralized monitoring, which tends to establish a global model with all process variables as inputs.…”

Online prediction based on the Copula function for the intrinsic viscosity of polyester fibre

“…It is more suitable for industrial processes with complex or unclear mechanisms. [6] Commonly used data-driven modelling approaches include partial least squares (PLS), [7] support vector machines (SVM), [8] artificial neural networks (ANN), [9] etc. Cao et al composed a linear dynamic neural network and a nonlinear static neural network into a hybrid neural network and investigated serial and parallel combinations.…”

Nonparametric manifold learning approach for improved process monitoring

“…[ 4–10 ] Two representative methods, partial least squares (PLS) and principal component analysis (PCA), have been widely applied to fault detection. [ 11–14 ] To further improve the monitoring performance of different industrial processes, these techniques have been extended from different perspectives. [ 15–22 ] However, all methods based on the PLS and PCA aim to preserve the global structure information (variance information) of the process data, while ignoring the local structure information.…”