Deep learning has attracted intense interest in Prognostics and Health Management (PHM), because of its enormous representing power, automated feature learning capability and best-in-class performance in solving complex problems. This paper surveys recent advancements in PHM methodologies using deep learning with the aim of identifying research gaps and suggesting further improvements. After a brief introduction to several deep learning models, we review and analyze applications of fault detection, diagnosis and prognosis using deep learning. The survey validates the universal applicability of deep learning to various types of input in PHM, including vibration, imagery, time-series and structured data. It also reveals that deep learning provides a one-fits-all framework for the primary PHM subfields: fault detection uses either reconstruction error or stacks a binary classifier on top of the network to detect anomalies; fault diagnosis typically adds a soft-max layer to perform multi-class classification; prognosis adds a continuous regression layer to predict remaining useful life. The general framework suggests the possibility of transfer learning across PHM applications. The survey reveals some common properties and identifies the research gaps in each PHM subfield. It concludes by summarizing some major challenges and potential opportunities in the domain.
This paper presents certain new approaches to reliability modeling of systems subject to shared loads. It is assumed that components in the system degrade continuously through additive impact under load. Reliability assessment of such systems is often complicated by the fact that both the arriving load and the failure of components influence the degradation of the surviving components in a complex manner. The proposed approaches seek to ease this problem by first deriving the time to prior failures and the arrival of random loads and then determining the number of failed components. Two separate models capable of analyzing system reliability as well as arriving at system maintenance and design decisions are proposed. The first considers constant load and the other cumulative load. A numerical example is presented to illustrate the effectiveness of the proposed models.
Malignant melanoma is aggressive and has a high mortality rate. Toll-like receptor 4 (TLR4) has been linked to melanoma growth, angiogenesis and metastasis. However, signal transduction mediated by TLR4 for driving melanoma progression is not fully understood. Signal transducer and activator of transcription 3 (STAT3) has been identified as a major oncogene in melanoma progression. We found: that TLR4 expression positively correlates with activation/phosphorylation of STAT3 in human melanoma samples; that TLR4 ligands activate STAT3 through MYD88 and TRIF in melanoma cells; and that intratumoral activation of TLR4 increases STAT3 activation in the tumor and promotes tumor growth, angiogenesis, epithelial-mesenchymal transition (EMT) and the formation of an immunosuppressive tumor microenvironment in mice. Further, we found that the effects mediated by activating TLR4 are weakened by suppressing STAT3 function with a dominant negative STAT3 variant in melanoma. Collectively, our work identifies STAT3 activation as a key event in TLR4 signaling-mediated melanoma progression, shedding new light on the pathophysiology of melanoma.
This paper develops a maintenance model for mission-oriented systems subject to natural degradation and external shocks. For mission-oriented systems which are used to perform safetycritical tasks, maintenance actions need to satisfy a range of constraints such as availability/reliability, maintenance duration and the opportunity of maintenance. Additionally, in developing maintenance policy, one needs to consider the natural degradation due to aging and wearing along with the external shocks due to variations of the operating environment. In this paper, the natural degradation is modeled as a Wiener process and the arrival of random shock as a homogeneous Poisson process. The damage caused by shocks is integrated into the degradation process, according to the cumulative shock model. Improvement factor model is used to characterize the impact of maintenance actions on system restoration. Optimal maintenance policy is obtained by minimizing the long-run cost rate. Finally, an example of subsea blowout preventer system is presented to illustrate the effectiveness of the proposed model.
The version in the Kent Academic Repository may differ from the final published version. Users are advised to check http://kar.kent.ac.uk for the status of the paper. Users should always cite the published version of record.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.