29 / 2025-03-26 14:44:07

Structured Augmented Sparse Dictionary Learning for Incipient Fault Detection and Isolation

Statistical properties embedding,Manifold structure preservation,Structured sparse coding,Dictionary learning-based monitoring

To address the common challenges encountered by dictionary learning-based monitoring, this paper presents a novel structured modeling method for detection and isolation of incipient faults that involves structured sparse coding and sequential dictionary augmentations. Through learning a basic dictionary for normal patterns and augmenting the low-dimensional sparse dictionaries for analyzing different fault patterns, the process signals can be decomposed into the fault-free and fault-related components. To guarantee the in-statistical-control status of the fault-free part and improve detection sensitivity, a L2-penalty is imposed on the sum of coefficient vectors to ensure that the monitoring statistic related to the fault-free part will not exceed the control limit. It is also suggested to maintain the similarity among sparse atoms for robust fault information extraction by combining two Frobenius norm penalties. Instead of imposing L1-sparsity constraint on the atoms, a hard sparsity constraint is used to correctly select fault-related feature variables, so that fault patterns can be better revealed. The informative dictionaries are, then, incorporated into the moving window-based monitoring strategy, yielding the friendly non-interfering statistics and effective isolation models for enhanced detectability and isolatability. The superior performance of our proposed approach is validated by a genuine industrial glass melter process.