Robust Just-in-time Learning Approach and Its Application on Fault Detection

Han, Yu; Yin, Shen; Luo, Hao

doi:10.1016/j.ifacol.2017.08.2437

Cited by 4 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, the process data can be sorted by operating mode and different models can be used, as in [7]. In just-in-time learning technique, the computation and relationship building is done online using similar data in 18 th IMEKO TC10 Conference "Measurement for Diagnostics, Optimisation and Control to Support Sustainability and Resilience" Warsaw, Poland, September 26-27, 2022 the database as in [8]. Recursive weighting can also be used to make well-known fault detection methods such as principal component analysis adaptive, as in [9].…”

Section: Related Results In the Literaturementioning

confidence: 99%

Adaptive Methods for Fault Detection on Research Engine Test Beds

Schadler¹,

Wohlthan²,

Wimmer³

2023

Proceedings of the 18th IMEKO TC10 Conference on Measurement for Diagnostics, Optimisation and Control

View full text Add to dashboard Cite

In the case of test beds for research engines, fault detection methods that use models based on historical data face a particular challenge. Due to the experimental design of the test bed, offline training of statistical models with a data set containing all possible variations is simply not possible. The methods must adapt to the current data situation directly on-site. But this involves risks. First, computational time and memory requirements can become extremely large with high data volumes. Second, the data may be faulty and thus negatively affecting the models. To avoid both, a selection of data is made before it is used to build the fault-free reference model. For this purpose, a new statistic is presented as the combination of the Mahalanobis distance and the forecast residual. With it, it is possible to reduce the update frequency and to increase the rate of detected faulty points, since the models are no longer manipulated by faulty data points and thus the residuals provide a better structure for fault detection.

show abstract

Section: Related Results In the Literaturementioning

confidence: 99%

Adaptive Methods for Fault Detection on Research Engine Test Beds

Schadler¹,

Wohlthan²,

Wimmer³

2023

Proceedings of the 18th IMEKO TC10 Conference on Measurement for Diagnostics, Optimisation and Control

View full text Add to dashboard Cite

show abstract

“…Compared with similar samples in historical databases, the signal status of online query could be possibly acquired in real time. Robust JITL strategies to leverage the weights of high leakage points of signals such as outliers had been successfully applied to the FD tasks [28]. A simulation study showed that the combined JITL-PCA models outperformed PCA in the analyzing of nonlinear signals [26].…”

Section: Introductionmentioning

confidence: 99%

Fault Detection for High-Speed Trains Using CCA and Just-in-Time Learning

et al. 2022

View full text Add to dashboard Cite

Online monitors of the running gears systems of high-speed trains play critical roles in ensuring operational safety and reliability. Status signals collected from high-speed train running gears are very complex regarding working environments, random noises and many other real-world constraints. This paper proposed fault detection (FD) models using canonical correlation analysis (CCA) and just-in-time learning (JITL) to process scalar signals of high-speed train gears, named as CCA-JITL. After data preprocessing and normalization, CCA transforms covariance matrices of high-dimension historical data into low-dimension subspaces and maximizes correlations between the most important latent dimensions. Then, JITL components formulate local FD models which utilize subsets of testing samples with larger Euclidean distances to training data. A case study introduced a novel system design of an online FD architecture and demonstrated that CCA-JITL FD models significantly outperformed traditional CCA models. The approach is applicable to other dimension reduction FD models such as PCA and PLS.

show abstract