Traffic flow forecasting is challenging for us to analyze intricate spatial–temporal dependencies and obtain incomplete information of spatial–temporal connection. Existing frameworks mostly construct spatial and temporal modeling based on a fixed graph structure and given time series. However, a fixed adjacency matrix is limited to learn effective spatial–temporal correlations of the network because it represents incomplete information for missing genuine relation. To solve the difficulty, we design a novel spatial–temporal adaptive fusion graph network (STFAGN) for traffic prediction. First, our model combines fusion convolution layers with a novel adaptive dependency matrix by end-to-end training to capture the hidden spatial-temporal dependency on the data to complete incomplete information. Second, STFAGN could, in parallel, acquire hidden spatial–temporal dependencies by a fusion operation and temporal trend by fast-DTW. Meanwhile, we use ReZero connection as a simple change of deep residual networks to facilitate deep signal propagation and faster converge. Lastly, we conduct comparative experiments on two public traffic network datasets, whose results demonstrate the superiority of our algorithm compared to state-of-the-art baseline types. Ablation experiments also prove the rationality of the framework of STFAGN.

The subsurface flow in fractured reservoirs is strongly affected by the distribution of fracture networks. Discrete fracture models, which represent all fractures individually by unstructured grid systems, are thus developed and act as a more accurate way for fractured reservoir simulation. However, it is usually not realistic to directly apply discrete fracture models to simulate field scale models for efficiency reasons. There is a need for upscaling techniques to coarsen the high resolution fracture descriptions to sizes that can be accommodated by reservoir simulators. In this paper, we extended the adaptive local-global upscaling technique to construct a transmissibility-based dual-porosity dual-permeability model from discrete fracture characterizations. An underlying unstructured fine-scale grid is firstly generated as a base grid. A global coarse-scale simulation is performed to provide boundary conditions for local regions and local upscaling procedures are carried out in every local region for transmissibility calculations. Iterations are performed until the consistency between the global and local properties is achieved. The procedure is applied to provide dual-porosity dual-permeability (DPDK) parameters including coarse-scale matrix-matrix, fracture-fracture and matrix-fracture flux transmissibilities. The methodology is applied to several cases. The simulation results demonstrate the accuracy, efficiency and robustness of the proposed method.

Purpose: In recent years, China's marine industry has maintained rapid growth in general, and marine-related economic activities have continued to improve. The purpose of this research is to analyze the basic situation of China's marine economy development, identify the problems therein, forecast development trends and propose policy recommendations accordingly.Design/methodology/approach: This research conducts a comprehensive and detailed analysis of the development of China's marine economy with rich data in diversified aspects. The current situation of China's marine economy development is analyzed from the perspective of scale and structure, and the external and internal development environment of China's marine economy is discussed. With the application of measurement and prediction method such as trend extrapolation, exponential smoothing, grey forecasting and neural network method, the future situation of China's marine economy development is forecasted.Findings: In a complex environment where uncertainties at home and abroad have increased significantly, China's marine economy development suffers tremendous downward pressure in recent years. As China has achieved major achievements in the prevention and control of the COVID-19 epidemic, the marine economy development will gradually return to normal. It is estimated that the gross marine production value in 2022 will exceed 10 trillion yuan. China's marine economy will continue to maintain a steady growth trend in the future, and its development prospects will remain promising.Originality/value: This research explores the current situation and trends of China's marine economy development and puts forward policy recommendations to promote the steady and health development of China's marine economy accordingly.

In this study, a 3D multicomponent multiphase simulator with a new fracture characterization technique is developed to simulate the enhanced recovery of coalbed methane. In this new model, the diffusion source from the matrix is calculated using the traditional dual-continuum approach, while in the Darcy flow scale, the Discrete Fracture Model (DFM) is introduced to explicitly represent the flow interaction between cleats and large-scale fractures. For this purpose, a general formulation is proposed to model the multicomponent multiphase flow through the fractured coal media. The S&D model and a revised P&M model are incorporated to represent the geomechanical effects. Then a finite volume based discretization and solution strategies are constructed to solve the general ECBM equations. The prismatic meshing algorism is used to construct the grids for 3D reservoirs with complex fracture geometry. The simulator is validated with a benchmark case in which the results show close agreement with GEM. Finally, simulation of a synthetic heterogeneous 3D coal reservoir modified from a published literature is performed to evaluate the production performance and the effects of injected gas composition, well pattern and gas buoyancy.

Summary: Multiphase flow behavior in the complex porous media of lacustrine shale oil is critical to the oil production performance. Shale oil reservoir processes spatial spaces of multiscale porous media. In view of the mesoscopic scale, the fluids flow behaviors of shale oil reservoirs are significantly different from that of tight oil reservoirs and the multimedia flow fluxes in shale have to be reevaluated. Furthermore, upscaling methods from the mesoscopic scale to the macroscopic level as well as coupling methods of multimedia mass transfer have not been adequately established. Based on the multiple media model, such as organic-inorganic matter system and embedded discrete fracture model (DFM), this work proposes a multiscale porous media flow simulation method that is applicable for lacustrine shale oil reservoirs. In the model, various reservoir matrix and flow spaces, such as organic pores, intragranular pores, intergranular pores, and fracture networks composed of bedding fractures and hydraulic fractures, are included. Methods of mass flux conductivity estimations among multiscale media and the relevant upgrading methods are also proposed. Validation of the model is first conducted by the comparison of the oil production performance estimated by the proposed model and a theory solution, and the model is further compared to Gulong shale oil reservoir data to determine its availability in field application. The results show that the proposed simulation model is capable of accurately characterizing the multiphase flow characteristics in multiscale media in shale. It is further demonstrated that the proposed model significantly improves the simulation accuracy over the current nonupscaling models. Field study shows that, based on the accurate characterization of the complex flows in shale oil reservoirs, the research output can provide support for future development of the Gulong shale oil reservoir.