Isolation and Localization of Unknown Faults Using Neural Network-Based Residuals

TL;DR

This paper introduces neural network-based residuals that incorporate physical insights for fault localization in industrial systems, enabling detection of unknown faults without extensive physical modeling.

Contribution

It presents a hybrid approach combining physical insights and neural networks to create residuals for fault isolation and localization of unknown faults.

Findings

Neural network residuals can be trained solely on fault-free data.

Residuals effectively localize unknown faults in simulation.

Hybrid residuals outperform traditional methods in fault localization.

Abstract

Localization of unknown faults in industrial systems is a difficult task for data-driven diagnosis methods. The classification performance of many machine learning methods relies on the quality of training data. Unknown faults, for example faults not represented in training data, can be detected using, for example, anomaly classifiers. However, mapping these unknown faults to an actual location in the real system is a non-trivial problem. In model-based diagnosis, physical-based models are used to create residuals that isolate faults by mapping model equations to faulty system components. Developing sufficiently accurate physical-based models can be a time-consuming process. Hybrid modeling methods combining physical-based methods and machine learning is one solution to design data-driven residuals for fault isolation. In this work, a set of neural network-based residuals are designed by incorporating physical insights about the system behavior in the residual model structure. The residuals are trained using only fault-free data and a simulation case study shows that they can be used to perform fault isolation and localization of unknown faults in the system.