Abstract:Both of multimode and nonlinear characteristics have been common in modern industrial processes. In the present work, an improved monitoring strategy on the basis of a modified discriminant locality preserving projections (DLPP) algorithm and stationary test is proposed for nonlinear multimode process monitoring to analyze both within-mode and cross-mode information. First, the adjoined multiple local models are constructed in high-dimensional kernel space to extract within-mode information through kernel fuzzy c-means algorithm and kernel principal components. Afterwards, the weighted local representations are defined by integrating the local latent vectors within each mode and the corresponding posterior probabilities of data samples. Then, a modified DLPP algorithm is developed to preserve the local structure within each mode and separate different modes as far as possible by merging local models into a low-dimensional coordinated space. Augmented Dickey Fuller test (ADF test) strategy is utilized to decompose the final low-dimensional space into two subspaces including mode-specific subspace and mode-common subspace by checking the stability of the final global data. A Bayesian monitoring strategy is presented in mode-specific subspace and residual subspace to provide more reliable monitoring results. Finally, to illustrate the feasibility and effectiveness, the proposed algorithm is illustrated with multimode data generated from the Tennessee Eastman (TE) benchmark process and Three-phase flow process.

Multimode diagnosis for switched affine systems with noisy measurement

An Asynchronized Observer Based Fault Detection Approach for Uncertain Switching Systems with Mode Estimation.

Dissipativity Based Fault Detection for 2D Markov Jump Systems with Asynchronous Modes

Multi-mode data-based joint fault detection method

Lebesgue-Time–Space-Model-Based Diagnosis and Prognosis for Multiple Mode Systems

Dissipativity-Based Asynchronous Fuzzy Sliding Mode Control for T–S Fuzzy Hidden Markov Jump Systems

Robust Fault Detection Scheme for Asynchronous Switched Systems via Sliding Mode Observer

Multimode Process Monitoring Based on Switching Autoregressive Dynamic Latent Variable Model

H_/Hᄮ, Filtering Fault Detection for Asynchronous Switched Discrete-time Linear Systems With Mode-dependent Average Dwell Time

Fault Detection for a Class of Continuous-Time Switched Affine Systems

Networked Fault Detection for Markov Jump Nonlinear Systems

Active Fault Diagnosis for Stochastic Systems Within Bayesian Minimum Risk Decision Framework

Nonstationary Filtering for Fuzzy Markov Switching Affine Systems with Quantization Effects and Deception Attacks

Fault Detection of Switched Systems with Repeated Scalar Non-Linearities

Fault Detection for Linear Discrete Time-Varying Systems With Multiplicative Noise: The Finite-Horizon Case

Data-driven mode detection and stabilization of unknown switched linear systems

Stationarity Test and Bayesian Monitoring Strategy for Fault Detection in Nonlinear Multimode Processes

A Subspace Technique for The Identification of Switched Affine Models

Real-time Hybrid System Identification with Online Deterministic Annealing

Set-based Active Fault Diagnosis for Discrete-time Linear Descriptor Systems

An Optimal Fault Detection Approach for Piecewise Affine Systems Via Diagnostic Observers.