Multiparameter estimation and LSTM-based prediction method for health state of single-layer reticulated shells

Guo, Xiaonong; Guo, Xiaonong; Zong, Shaohan; Xu, Hongjun

doi:10.1016/j.jobe.2023.107128

Cited by 1 publication

(1 citation statement)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Increasing experience in the use of up-to-date measurement and monitoring technologies and rapid developments in computer processing power and storage capacity have caused a trend towards data-driven approaches for condition monitoring. The data-driven determination of the structural condition via the application of monitoring systems has a very long history, and it is also applied to the predictive maintenance of engineering structures [9][10][11][12][13][14][15][16][17][18]. Identifying and implementing key performance indicators play major roles here [8,19].…”

Section: Introductionmentioning

confidence: 99%

Data-Driven Condition Assessment and Life Cycle Analysis Methods for Dynamically and Fatigue-Loaded Railway Infrastructure Components

Granzner,

Strauss,

Reiterer

et al. 2023

Infrastructures

View full text Add to dashboard Cite

Railway noise barrier constructions are subjected to high aerodynamic loads during the train passages, and the knowledge of their actual structural condition is relevant to assure safety for railway users and to create a basis for forecasting. This paper deals with deterministic and probabilistic approaches for the condition assessment and prediction of the remaining lifetime of railway noise barriers that are embedded in a safety concept that takes into account the damage consequence classes. These approaches are combined into a holistic assessment concept, in other words, a progressive four-stage model in which the information content increases with each model stage and thus successively increases the accuracy of the determined structural conditions at the time of observation and the forecast of the remaining service life of the structure. The analytical methods used in the first stage of the developed holistic framework are based on common static calculations used in engineering practice and, together with expert knowledge and large-scale fatigue test results of noise barrier constructions, form the basis for the subsequent stages. In the second stage of the data-driven condition assessment and life cycle analysis approach, linking routines are implemented that combine the condition assessments from the visual inspections with the additional information from temporary or permanent monitoring systems with the analytical methods. With the application of numerical finite element methods for the development of a digital twin of the noise barrier in the third stage and the probabilistic approaches in the fourth stage, a maximum determination accuracy of the noise barrier condition at the time of observation and prediction accuracy of the remaining service life is achieved. The data-driven condition assessment and life cycle analysis approach enables infrastructure operators to plan their future investments more economically regarding the maintenance, retrofitting, or new construction of railway noise barriers. Ultimately, the aim is to integrate the presented four-stage holistic assessment concept into the specific maintenance and repair planning of infrastructure operators for aerodynamically loaded railway noise barrier constructions.

show abstract