Search

Frequently occurring and unavoidable perturbations in railway systems cause a reduction in performance and even the infeasibility of the original schedule. In addition, existing related studies assume that the maintenance routes are fixed in a deterministic situation. Therefore, it is necessary to focus on robust scheduling of electric multiple units first-level maintenance with flexible maintenance routes to defend against stochastic uncertainties. Firstly, a mixed-integer linear programming model with uncertain parameters in objective functions and constraints is built for the first time. Multiple constraints are formulated to consider flexible maintenance routes, train shunting conflicts, and track occupation conflicts in a stub-end depot. A robust optimization approach is adopted to obtain a deterministic robust counterpart model with uncertainties in processing/arrival/departure times. In this model, uncertainty and reliability levels are introduced to describe and quantify the disturbance degree of processing/arrival/departure times and the allowable violation degree of resource constraints respectively. Moreover, an adaptive iterative local search is proposed by embedding heuristic rules for the initialization. The proposed algorithm includes problem-specific neighborhood structures to improve the evolutionary ability, a variable neighborhood descent method to guide and drive the search toward promising areas of the search space, and an adaptive perturbation mechanism to enable the algorithm to jump out of local optimum. Numerical results from China's railway system validate the proposed model and quantitatively demonstrate the merit of the proposed algorithm. Further analysis shows that our approach can achieve an appropriate trade-off between robustness and efficiency for various uncertainty and reliability levels. • The SSEFM problem under stochastic uncertainties is addressed. • A multi-objective MILP model with deterministic robust equivalent counterpart is defined. • An adaptive iterative local search is proposed to solve this problem. • The algorithm is enhanced with a variable neighborhood descent and adaptive perturbation. [ABSTRACT FROM AUTHOR]

Pockmarks are circular or elliptical depressions in the seabed that are geomorphologically similar but otherwise distinct from volcanic craters. Pockmarks are found at various depths in the ocean including on the continental shelf, the continental slope and the deep-sea plain. This paper presents data covering a large-scale pockmark group in the middle part of the shallow sea sedimentary plain of the North Yellow Sea at a water depth of 51–58 m. The distribution of pockmarks within this group is very dense, and pockmark diameters range from several hundred meters to several kilometers. Based on multibeam echo sounder data, a high-resolution image of the pockmark area is obtained by efficient image processing methods. The textural features of the seabed surface layer are then described and analyzed. Multibeam echo sounder backscatter data are combined with seabed sediment sampling data to develop a statistical model linking seabed backscatter sonar strength and sediment grain-size characteristics. Additionally, an improved BP neural network is used to realize rapid and automatic classification and identification of seabed sediment types. Fine-scale processing analysis of sonar images and acoustic classification of seabed sediments as presented in this paper are useful for determining seabed surface textures and local sediment distribution laws. This study thus provides a basis for further study of the formation and evolution of seabed pockmarks. [ABSTRACT FROM AUTHOR]

High resolution multibeam bathymetry revealed the occurrence of numerous craterlike depressions, so-called pockmarks, in the sea floor of the North Yellow Sea Basin (NYSB). To investigate whether these pockmarks are related to any type of fluid flow or gas seepage from subsurface stratum, acoustic survey and box samplings were conducted in a selected area containing several typical pockmarks. This allowed for an integrated analysis of the morphology, distribution, and subsurface strata as well as a direct comparison of seafloor characteristics among different areas inside and outside of a pockmark. Additionally, surface sediments were analyzed for particle structure and grain size. Based on the combination of multibeam bathymetry and backscatter intensity data, 282 seafloor depressions were accurately identified. Moreover, the formation of pockmark can be attributed to existence of gas chimneys and other shallow gases revealed by shallow seismic profile records in the subsurface strata. Sediments inside and outside the pockmark showed remarkable discrepancies in terms of grain size, and benthic fauna. Additionally, strong internal reflection area was observed inside the depressions where benthic habitats are abundant and sediments contain strawberry pyrite and carbonate cementation. Side scan sonar also recorded sediment plumes in the water column, which indicates that the pockmarks are still active. Several "pearl-string-like" accumulations of pockmark were found in the study area, which well agree with the distribution of buried channels and lakes. These results thus provide new insights in interpreting shallow water fluid seepage, with implications for our current knowledge of overall sedimentation process in offshore continental shelves. • Active fluid seepage in shallow water continental shelf environment • Shallow-gas related acoustic features recorded on sub-bottom acoustic data • Benthic fauna anomalies, and strawberry pyrite associated with seafloor seeps • The large but shallow pockmarks can be ascribed to low fluid pressure. • String like distribution of pockmarks controlled by subsurface buried channels. [ABSTRACT FROM AUTHOR]