Towards deep probabilistic graph neural network for natural gas leak detection and localization without labeled anomaly data

Deep learning has been widely applied to automated leakage detection and location of natural gas pipe networks. Prevalent deep learning approaches do not consider the spatial dependency of sensors, which limits leakage detection performance. Graph deep learning is a promising alternative to prevailing approaches as it can model spatial...

Influence of unilateral low adhesion on transient wheel-rail rolling contact and wheel damages

A time-domain finite element model is developed to study the transient rolling contact of a driving wheelset over a curved track with Low Adhesion Zones (LAZs) shorter than 1.0 m. LAZs on one rail, i.e., unilateral LAZs occurring more likely, is treated for a speed up to 500 km/h. Structural vibrations of wheelset are analyzed to explain the...

Groundwater–surface water interactions, vegetation dependencies and implications for water resources management in the semi-arid Hailiutu River catchment, China – a synthesis

During the last decades, large-scale land use changes took place in the Hailiutu River catchment, a semi-arid area in northwest China. These changes had significant impacts on the water resources in the area. Insights into groundwater and surface water interactions and vegetation-water dependencies help to understand these impacts and formulate...