Multitier Dynamic Dual Kullback–Leibler Divergence Based on Manifold and Sensitive Features for Incipient Fault Detection
- 29 July 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-10
Abstract
To solve the problem of incipient fault detection in an effective and timely manner, a multitier dynamic dual Kullback–Leibler Divergence (MDD-KLD) based on manifold and sensitive features is proposed. Firstly, the subspace highlighting the dynamic trend of the data is constructed using an exponentially weighted moving average window on the basis of manifold learning. Then, two sensitive detection indicators are constructed by leveraging the sensitivity of the KLD to minor changes in the probability distribution of the data, which is called dynamic dual KL divergence model. In addition, by calculating the KLD between normal data and priori fault data to determine the optimal sensitive variable subblock for a specific fault, the fault sensitive features are extracted to construct a complementary detection model, which supports the primary detection model, resulting in the multitier dynamic dual-KL divergence (MDD-KLD) model. Finally, testing on an industrial case showed that the primary model attains 97% average accuracy for routine faults, while the complementary detection model further enhances both accuracy and fault detection time for incipient faults relative to the primary model.Keywords
Funding Information
- National Natural Science Foundation of China (62473155, 62473154, 62473156)
- Shanghai Oriental Talents Program Youth Project (QNKJ2024034)
- Shanghai Chenguang project (21CGA37)
This publication has 32 references indexed in Scilit:
- Incipient fault detection and isolation for dynamic processes with slow feature statistics analysisChemical Engineering Science, 2024
- Incipient fault detection with probability transformation and statistical feature analysisAutomatica, 2024
- Incipient fault diagnosis and trend prediction in nonlinear closed-loop systems with Gaussian and non-Gaussian noiseComputers & Chemical Engineering, 2023
- Hierarchical hybrid distributed PCA for plant-wide monitoring of chemical processesControl Engineering Practice, 2021
- A modified neighborhood preserving embedding-based incipient fault detection with applications to small-scale cyber–physical systemsISA Transactions, 2020
- A novel diagnostic and prognostic framework for incipient fault detection and remaining service life prediction with application to industrial rotating machinesApplied Soft Computing, 2019
- An improved incipient fault detection method based on Kullback-Leibler divergenceISA Transactions, 2018
- Multimode process monitoring using improved dynamic neighborhood preserving embeddingChemometrics and Intelligent Laboratory Systems, 2014
- Incipient fault diagnosis using support vector machines based on monitoring continuous decision functionsEngineering Applications of Artificial Intelligence, 2014
- Nonlinear process monitoring based on linear subspace and Bayesian inferenceJournal of Process Control, 2010