Orthonormal subspace analysis (OSA) is proposed for handling the decomposition issue and principal component selection in traditional key performance indicator (KPI)-related process monitoring methods such as partial least squares (PLS) canonical correlation (CCA). However, it not appropriate to apply static OSA algorithm a dynamic since pays no attention auto-correlation relationships variable...

The increased complexity of digitalized process systems requires advanced tools to detect and diagnose faults early maintain safe operations. This study proposed a hybrid model that consists Kernel Principal Component Analysis (kPCA) DNNs can be applied in various processes. complex data is processed by kPCA reduce its dimensionality; then, simplified used for two separate training (detection d...

In order to ensure the safe production of chemical handicrafts, it is necessary monitor process variables system in deal with failures. improve accuracy and monitoring process, whole modeled find out key that cause fault, at same time, impact analyzed propagation path variables. Based on complex network theory, fault node set obtained by combining measured data horizontal visibility map; At con...

A new fault-relevant KPCA algorithm is proposed. Then the fault detection approach is proposed based on the fault-relevant KPCA algorithm. The proposed method further decomposes both the KPCA principal space and residual space into two subspaces. Comparedwith traditional statistical techniques, the fault subspace is separated based on the fault-relevant influence. This method can find fault-rel...

An RNN-based forecasting approach is used to early detect anomalies in industrial multivariate time series data from a simulated Tennessee Eastman Process (TEP) with many cyberattacks. This work continues a previously proposed LSTM-based approach to the fault detection in simpler data. It is considered necessary to adapt the RNN network to deal with data containing stochastic, stationary, trans...

In this work, we apply the systematic approach to plant-wide control design presented in [1], based on the fundamentals of process networks, thermodynamics and systems theory, to the Tennessee Eastman (TE) Challenge Process, deriving robust decentralized controllers that will ensure the stability of the complete plant. We take one step further in the control design procedure by completing it wi...

The objective of this exercise is to demonstrate how a steady state gain matrix, K can be calculated from a linearized state space dynamic process model. The approach used has been published by Arkun and Downs [1], and it is aimed at handling cases where integrating elements, e.g. liquid levels, are present. The dynamic approach is applied to the Tennessee Eastman process, and the relative gain...

In this paper, a new sensor network design methodology has been proposed to improve the performance of ICA-based process monitoring approaches. Design procedure incorporates sensor cost, fault detectability and fault detection rate in the design formulation. The design problem has been transformed into an optimization problem. A genetic algorithm (GA) solver has been employed to yield optimal s...

In this paper, we present a Bayesian framework for fault detection. Principal component analysis (PCA) technique and quadratic test statistics are incorporated under a Conditional Gaussian Network (CGN). The proposed network, given an observation, is able to project it into an orthogonal space and to give a decision about the system state (faulty or not). The paper demonstrate, the probability ...