.png){kind=icon}
.png){kind=icon}
Return
Adaptive Interval Observer for Sensor Fault Detection with Minimum Detectable Fault Analysis under Unknown Time-Varying Parameters
.png){kind=icon}
PRE
AI
DOI：10.1109/TR.2026.3667860.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This article presents an adaptive interval fault detection method for autonomous mine haul trucks (AMHTs) with unknown time-varying parameters. Traditional observer-based fault detection approaches are susceptible to large-magnitude and unknown time-varying parameters, which often degrade state estimation accuracy and lead to the biased residuals and the increased false alarm rate. To address this issue, by time-varying learning gains and interval observer structures, an interval state estimation method is proposed in this article to improve the fault detection robustness. Moreover, although the conventional <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$H_-$</tex-math></inline-formula> performance index improves fault detection sensitivity, it does not provide the minimum detectable fault (MDF) quantification under unknown time-varying parameters. In addition, the MDF quantification is considered only as a postdesign evaluation in the existing schemes. By contrast, the MDF quantification is transformed into an optimization objective in this article. The objective is embedded within the design process of the adaptive fault detection interval observer to minimize the MDF index and enhance fault detection sensitivity. The effectiveness of the proposed method is validated through simulations on the AMHT suspension system. Comparative results demonstrate improved fault detection performance in the presence of unknown parameter variations and external disturbances.
Keywords:
Autonomous mine haul trucks (AMHTs)
fault detection
interval observers
minimum detectable faults (MDFs)
unknown time-varying parameters
Journal
IF:
5.7
Papers:
2.7K
Citations:
8.5K