Prediction of three-dimensional ocean temperature in the South China Sea based on time series gridded data and a dynamic spatiotemporal graph neural network

Ocean temperature is an important physical variable in marine ecosystems,and ocean temperature prediction is an important research objective in ocean-related fields.Currently,one of the commonly used methods for ocean temperature prediction is based on data-driven,but research on this method is mostly limited to the sea surface,with few studies on the prediction of internal ocean temperature.Existing graph neural network-based methods usually use predefined graphs or learned static graphs,which cannot capture the dynamic associations among data.In this study,we propose a novel dynamic spatiotemporal graph neural network(DSTGN)to predict threedimensional ocean temperature(3D-OT),which combines static graph learning and dynamic graph learning to automatically mine two unknown dependencies between sequences based on the original 3D-OT data without prior knowledge.Temporal and spatial dependencies in the time series were then captured using temporal and graph convolutions.We also integrated dynamic graph learning,static graph learning,graph convolution,and temporal convolution into an end-to-end framework for 3D-OT prediction using time-series grid data.In this study,we conducted prediction experiments using high-resolution 3D-OT from the Copernicus global ocean physical reanalysis,with data covering the vertical variation of temperature from the sea surface to 1000 m below the sea surface.We compared five mainstream models that are commonly used for ocean temperature prediction,and the results showed that the method achieved the best prediction results at all prediction scales.

Exploring the molecular mechanism of action of curcumin for the treatment of diabetic retinopathy,using network pharmacology,molecular docking,and molecular dynamics simulation

Background:Based on network pharmacology and molecular docking,the present study investigated the mechanism of curcumin(CUR)in diabetic retinopathy treatment.Methods:Based on the DisGeNET,Swiss TargetPrediction,GeneCards,Online Mendelian Inheritance in Man,Gene Expression Omnibus,and Comparative Toxicogenomics Database,the intersection core targets of CUR and diabetic retinopathy were identified.The intersection target was imported into the STRING database to obtain the protein-protein interaction map.According to the Database for Annotation,Visualization and Integrated Discovery database,the intersected targets were enriched in Gene Ontology(GO)and Kyoto Encyclopedia of Genes and Genomes pathways.Then Cytoscape 3.9.1 is used to make the drug-target-disease-pathway network.The mechanism of CUR and diabetic retinopathy was further verified by molecular docking and molecular dynamics simulation.Results:There were 203 intersecting targets of CUR and diabetic retinopathy identified.1320 GO entries were enriched for GO functions,which were primarily involved in the composition of cells such as identical protein binding,protein binding,enzyme binding,etc.It was found that 175 pathways were enriched using Kyoto Encyclopedia of Genes and Genomes pathway enrichment methods,which were mainly included in the lipid and atherosclerosis,AGE-RAGE signaling pathway in diabetic complications,pathways in cancer,etc.In the molecular docking analysis,CUR was found to have a good ability to bind to the core targets of albumin,IL-1B,and IL-6.The binding of albumin to CUR was further verified by molecular dynamics simulation.Conclusion:As a result of this study,CUR may exert a role in the treatment of diabetic retinopathy through multi-target and multi-pathway regulation,which indicates a possible direction of future research.

Numerical simulation of three-dimensional fracturing fracture propagation in radial wells

A fracture propagation model of radial well fracturing is established based on the finite element-meshless method.The model considers the coupling effect of fracturing fluid flow and rock matrix deformation.The fracture geometries of radial well fracturing are simulated,the induction effect of radial well on the fracture is quantitatively characterized,and the influences of azimuth,horizontal principle stress difference,and reservoir matrix permeability on the fracture geometries are revealed.The radial wells can induce the fractures to extend parallel to their axes when two radial wells in the same layer are fractured.When the radial wells are symmetrically distributed along the direction of the minimum horizontal principle stress with the azimuth greater than 15,the extrusion effect reduces the fracture length of radial wells.When the radial wells are symmetrically distributed along the direction of the maximum horizontal principal stress,the extrusion increases the fracture length of the radial wells.The fracture geometries are controlled by the rectification of radial borehole,the extrusion between radial wells in the same layer,and the deflection of the maximum horizontal principal stress.When the radial wells are distributed along the minimum horizontal principal stress symmetrically,the fracture length induced by the radial well decreases with the increase of azimuth;in contrast,when the radial wells are distributed along the maximum horizontal principal stress symmetrically,the fracture length induced by the radial well first decreases and then increases with the increase of azimuth.The fracture length induced by the radial well decreases with the increase of horizontal principal stress difference.The increase of rock matrix permeability and pore pressure of the matrix around radial wells makes the inducing effect of the radial well on fractures increase.

Three-Dimensional Simulation of Hydrodynamic Mechanism of Fluidized Bed Methanation

Organic solid waste(OSW)contains many renewable materials.The pyrolysis and gasification of OSW can realize resource utilization,and its products can be used for methanation reaction to produce synthetic natural gas in the specific reactor.In order to understand the dynamic characteristics of the reactor,a three-dimensional numerical model has been established by the method of Computational Fluid Dynamics(CFD).Along the height of the reactor,the particle distribution in the bed becomes thinner and the mean solid volume fraction decreases from 4.18%to 0.37%.Meanwhile,the pressure fluctuation range decreased from 398.76 Pa at the entrance to a much lower value of 74.47 Pa at the exit.In this simulation,three parameters of gas inlet velocity,operating temperature and solid particle diameter are changed to explore their influences on gas-solid multiphase flow.The results show that gas velocity has a great influence on particle distribution.When the gas inlet velocity decreases from 6.51 to 1.98 m/s,the minimum height that particles can reach decreases from 169 to 100 mm.Additionally,as the operating temperature increases,the particle holdup inside the reactor changes from 0.843%to 0.700%.This indicates that the particle residence time reduces,which is not conducive to the follow-up reaction.Moreover,with the increase of particle size,the fluctuation range of the pressure at the bottom of the reactor increases,and its standard deviation increases from 55.34 to 1266.37 Pa.

Dynamic simulation insights into friction weakening effect on rapid long-runout landslides:A case study of the Yigong landslide in the Tibetan Plateau,China

This study proposed a novel friction law dependent on velocity,displacement and normal stress for kinematic analysis of runout process of rapid landslides.The well-known Yigong landslide occurring in the Tibetan Plateau of China was employed as the case,and the derived dynamic friction formula was included into the numerical simulation based on Particle Flow Code.Results showed that the friction decreased quickly from 0.64(the peak)to 0.1(the stead value)during the 5s-period after the sliding initiation,which explained the behavior of rapid movement of the landslide.The monitored balls set at different sections of the mass showed similar variation characteritics regarding the velocity,namely evident increase at the initial phase of the movement,followed by a fluctuation phase and then a stopping one.The peak velocity was more than 100 m/s and most particles had low velocities at 300s after the landslide initiation.The spreading distance of the landslide was calculated at the two-dimension(profile)and three-dimension scale,respectively.Compared with the simulation result without considering friction weakening effect,our results indicated a max distance of about 10 km from the initial unstable position,which fit better with the actual situation.

Image Representations of Numerical Simulations for Training Neural Networks

A large amount of data can partly assure good fitting quality for the trained neural networks.When the quantity of experimental or on-site monitoring data is commonly insufficient and the quality is difficult to control in engineering practice,numerical simulations can provide a large amount of controlled high quality data.Once the neural networks are trained by such data,they can be used for predicting the properties/responses of the engineering objects instantly,saving the further computing efforts of simulation tools.Correspondingly,a strategy for efficiently transferring the input and output data used and obtained in numerical simulations to neural networks is desirable for engineers and programmers.In this work,we proposed a simple image representation strategy of numerical simulations,where the input and output data are all represented by images.The temporal and spatial information is kept and the data are greatly compressed.In addition,the results are readable for not only computers but also human resources.Some examples are given,indicating the effectiveness of the proposed strategy.

Full Dynamic Model for Liquid Sloshing Simulation in Cylindrical Tank Shape

This study presents a comprehensive full dynamic model designed for simulating liquid sloshing behavior within cylindrical tank structures. The model employs a discretization approach, representing the liquid as a network of interconnected spring-damper-mass systems. Key aspects include the adaptation of liquid discretization techniques to cylindrical lateral cross-sections and the calculation of stiffness and damping coefficients. External forces, simulating various vehicle maneuvers, are also integrated into the model. The resulting system of equations is solved using Maple Software with the Runge-Kutta-Fehlberg method. This model enables accurate prediction of liquid displacement and pressure forces, offering valuable insights for tank design and fluid dynamics applications. Ongoing refinement aims to broaden its applicability across different liquid types and tank geometries.

Physics-informed neural network-based petroleum reservoir simulation with sparse data using domain decomposition

Recent advances in deep learning have expanded new possibilities for fluid flow simulation in petroleum reservoirs.However,the predominant approach in existing research is to train neural networks using high-fidelity numerical simulation data.This presents a significant challenge because the sole source of authentic wellbore production data for training is sparse.In response to this challenge,this work introduces a novel architecture called physics-informed neural network based on domain decomposition(PINN-DD),aiming to effectively utilize the sparse production data of wells for reservoir simulation with large-scale systems.To harness the capabilities of physics-informed neural networks(PINNs)in handling small-scale spatial-temporal domain while addressing the challenges of large-scale systems with sparse labeled data,the computational domain is divided into two distinct sub-domains:the well-containing and the well-free sub-domain.Moreover,the two sub-domains and the interface are rigorously constrained by the governing equations,data matching,and boundary conditions.The accuracy of the proposed method is evaluated on two problems,and its performance is compared against state-of-the-art PINNs through numerical analysis as a benchmark.The results demonstrate the superiority of PINN-DD in handling large-scale reservoir simulation with limited data and show its potential to outperform conventional PINNs in such scenarios.

Elucidating the molecular targets of Curcuma longa for breast cancer treatment using network pharmacology,molecular docking and molecular dynamics simulation

Background:To elucidate the molecular mechanisms of Curcuma longa(C.longa)in breast cancer treatment.Methods:Phytocompounds of C.longa were obtained from Dr.Duke’s Phytochemical and Ethnobotanical Database.Potential active targets were retrieved from Bindingdb,SEA and Swiss Target Prediction databases.Breast cancer targets were retrieved from the Therapeutic Target Database.Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses were done using DAVID and KOBAS3.0 databases respectively.The Cytoscape software was used to construct the phytocompound-target-pathway network.The PyRx and Desmond software were utilized for molecular docking and molecular dynamics simulation respectively.Results:Out of one hundred and fifty-six phytocompounds,fifty-four modulated proteins involved in breast cancer.Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway analysis indicated C.longa exerts its therapeutic effect through regulating several key pathways.Molecular docking analysis revealed that most phytocompounds of C.longa had a good affinity with the key targets.Molecular dynamics simulation showed that ethinylestradiol formed stable ligand-protein complexes.Conclusion:The results of this study will enhance our understanding of the potential molecular mechanisms by which C.longa inhibits breast cancer and lay a foundation for future experimental studies.

Three-Dimensional and Cross-sectional Characteristics of Normal Grain Growth Based on Monte Carlo Simulation

An appropriate Monte Carlo method was developed to simulate the three-dimensional normal grain growth more completely. Comparative investigation on the three-dimensional and the cross-sectional characteristics of normal grain growth was done. It was found that the time exponent of grain growth determined from cross-section exhibits the same rule of increasing slowly with time and approaching the theoretical value n = 0.5 of steadygrain growth as the three-dimensional (3-D) system. From change of the number of grains per unit area with timemeasured in cross-section, the state of 3-D normal grain growth may be predicted. The gtain size distribution incross-section is different from that in 3-D system and can not express the evolution characteristic of the 3-D distribution. Furthermore, there exists statistical connection between the topological parameters in cross-section and thosein three-dimensions.

Three-dimensional visualization and virtual reality simulation role in hepatic surgery:Further research warranted

Artificial intelligence(AI)is the study of algorithms that enable machines to analyze and execute cognitive activities including problem solving,object and word recognition,reduce the inevitable errors to improve the diagnostic accuracy,and decision-making.Hepatobiliary procedures are technically complex and the use of AI in perioperative management can improve patient outcomes as discussed below.Three-dimensional(3D)reconstruction of images obtained via ultrasound,computed tomography scan or magnetic resonance imaging,can help surgeons better visualize the surgical sites with added depth perception.Preoperative 3D planning is associated with lesser operative time and intraoperative complications.Also,a more accurate assessment is noted,which leads to fewer operative complications.Images can be converted into physical models with 3D printing technology,which can be of educational value to students and trainees.3D images can be combined to provide 3D visualization,which is used for preoperative navigation,allowing for more precise localization of tumors and vessels.Nevertheless,AI enables surgeons to provide better,personalized care for each patient.

Three-dimensional spatial structure of the macro-pores and flow simulation in anthracite coal based on X-ray μ-CT scanning data

The three-dimensional(3 D) structures of pores directly affect the CH4 flow.Therefore,it is very important to analyze the3 D spatial structure of pores and to simulate the CH4 flow with the connected pores as the carrier.The result shows that the equivalent radius of pores and throats are 1-16 μm and 1.03-8.9 μm,respectively,and the throat length is 3.28-231.25 μm.The coordination number of pores concentrates around three,and the intersection point between the connectivity function and the X-axis is 3-4 μm,which indicate the macro-pores have good connectivity.During the single-channel flow,the pressure decreases along the direction of CH4 flow,and the flow velocity of CH4 decreases from the pore center to the wall.Under the dual-channel and the multi-channel flows,the pressure also decreases along the CH4 flow direction,while the velocity increases.The mean flow pressure gradually decreases with the increase of the distance from the inlet slice.The change of mean flow pressure is relatively stable in the direction horizontal to the bedding plane,while it is relatively large in the direction perpendicular to the bedding plane.The mean flow velocity in the direction horizontal to the bedding plane(Y-axis) is the largest,followed by that in the direction horizontal to the bedding plane(X-axis),and the mean flow velocity in the direction perpendicular to the bedding plane is the smallest.

Three-Dimensional Finite Element Numerical Simulation and Physical Experiment for Magnetism-Stress Detecting in Oil Casing

The casing damage has been a big problem in oilfield production. The current detection methods mostly are used after casing damage, which is not very effective. With the rapid development of China's offshore oil industry, the number of offshore oil wells is becoming larger and larger. Because the cost of offshore oil well is very high, the casing damage will cause huge economic losses. What's more, it can also bring serious pollution to marine environment. So the effective methods of detecting casing damage are required badly. The accumulation of stress is the main reason for the casing damage. Magnetic anisotropy technique based on counter magnetostriction effect can detect the stress of casing in real time and help us to find out the hidden dangers in time. It is essential for us to prevent the casing damage from occurring. However, such technique is still in the development stage. Previous studies mostly got the relationship between stress and magnetic signals by physical experiment, and the study of physical mechanism in relative magnetic permeability connecting the stress and magnetic signals is rarely reported. The present paper uses the ANSYS to do the three-dimensional finite element numerical simulation to study how the relative magnetic permeability works for the oil casing model. We find that the quantitative relationship between the stress' s variation and magnetic induction intensity's variation is: Δδ =K* ΔB, K = 8.04×109, which is proved correct by physical experiment.

THE SYSTEM SIMULATION OF THREE-DIMENSIONAL RADAR

To provide a test platform for Electronic Warfare (EW) system, it is needed to simulate the radar received Intermediate Frequency (IF) signals and radar system functions.This letter gives a description of a radar system simulation software developed for frequencyphase scanning three-dimensional (3-D) radar. Experimental results prove that the software could be used for system evaluation and for training purposes as an attractive alternative to real EW system.

Three-Dimensional Simulations of RESET Operation in Phase-Change Random Access Memory with Blade-Type Like Phase Change Layer by Finite Element Modeling

An optimized device structure for reducing the RESET current of phase-change random access memory （PCRAM） with blade-type like （BTL） phase change layer is proposed. The electrical thermal analysis of the BTL cell and the blade heater contactor structure by three-dimensional finite element modeling are compared with each other during RESET operation. The simulation results show that the programming region of the phase change layer in the BTL cell is much smaller, and thermal electrical distributions of the BTL cell are more concentrated on the TiN/GST interface. The results indicate that the BTL cell has the superiorities of increasing the heating efficiency, decreasing the power consumption and reducing the RESET current from 0.67mA to 0.32mA. Therefore, the BTL cell will be appropriate for high performance PCRAM device with lower power consumption and lower RESET current.

Three-Dimensional Rigid-Plastic FEM Simulation of Bulk Forming Processes with New Contact and Remeshing Techniques

Some techniques such as die surface description, contact judgement algorithm and remeshing are proposed to improve the robustness of the numerical solution. Based on these techniques, a three-dimensional rigid-plastic FEM code has been developed. Isothermal forging process of a cylindrical housing has been simulated. The simulation results show that the given techniques and the FEM code are reasonable and feasible for three-dimensional bulk forming processes.

Three-dimensional discrete element numerical simulation of Paleogene salt structures in the western Kuqa foreland thrust belt

Taking the Paleogene salt strata in the west of Kuqa foreland thrust belt as study object, the deformation features of salt structure in the compression direction and perpendicular to the compression direction were examined to find out the control factors and formation mechanisms of the salt structures. By using the three-dimensional discrete element numerical simulation method, the formation mechanisms of typical salt structures of western Kuqa foreland thrust belt in Keshen and Dabei work areas were comprehensively analyzed. The simulation results show that the salt deformation in Keshen and Dabei work areas is of forward spread type, with deformation concentrated in the piedmont zone;the salt deformation is affected by the early uplift near the compression end, pre-existing basement faults, synsedimentary process and the initial salt depocenter;in the direction perpendicular to the compression direction, salt rocks near the compression end have strong lateral mobility with the velocity component moving towards the middle part, and the closer to the middle, the larger the velocity will be, so that salt rocks will aggregate towards the middle and deform intensely, forming complex folds and separation of salt structures from salt source, and local outcrop with thrust faults. Compared with 2 D simulation, 3 D simulation can analyze salt structures in the principal stress direction and direction perpendicular to the principal stress, give us a full view of the formation mechanisms of salt structures, and guide the exploration of oil and gas reservoirs related to salt structures.

Numerical simulations and comparative analysis of two- and three-dimensional circulating fluidized bed reactors for CO2 capture

Carbon dioxide(CO2),the main gas emitted from fossil burning,is the primary contributor to global warming.Circulating fluidized bed reactor(CFBR)is proved as an energy-efficient method for post-combustion CO2 capture.The numerical simulation by computational fluid dynamics(CFD)is believed as a promising tool to study CO2 adsorption process in CFBR.Although three-dimensional(3D)simulations were proved to have better predicting performance with the experimental results,two-dimensional(2D)simulations have been widely reported for qualitative and quantitative studies on gas-solid behavior in CFBR for its higher computational efficiency recently.However,the discrepancies between 2D and 3D simulations have rarely been evaluated by detailed study.Considering that the differences between the 2D and 3D simulations will vary substantially with the changes of independent operating conditions,it is beneficial to lower computational costs to clarify the effects of dimensionality on the numerical CO2 adsorption runs under various operating conditions.In this work,the comparative analysis for CO2 adsorption in 2D and 3D simulations was conducted to enlighten the effects of dimensionality on the hydrodynamics and reaction behaviors,in which the separation rate,species distribution and hydrodynamic characteristics were comparatively studied for both model frames.With both accuracy and computational costs considered,the viable suggestions were provided in selecting appropriate model frame for the studies on optimization of operating conditions,which directly affect the capture and energy efficiencies of cyclic CO2 capture process in CFBR.

Simulation-Based Construction of Three-Dimensional Process Model for Punching Cartridge Cases

A method of constructing three-dimensional process model for the punching cartridge cases is presented based on DEFORM simulation analysis. Using DEFORM software,the finite element simulation models for the punching and forming process of cartridge cases are established,and the corresponding simulation result model of each intermediate procedure is obtained by continuously performing the forming process simulation. The simulation model cannot annotate size and process information due to poor interface between DEFORM software and CAD software. Thus,a 3D annotation module is developed with secondary development technology of UG NX software. Consequently,the final process model with dimension and process information is obtained. Then,with the current 3D process management system,the 3D punching and forming process design of cartridge cases can be completed further. An example is also provided to illustrate that the relative error between the simulation process model and the physical model is less than 2%,which proves the validity and reliability of the proposed method in this study.

Application of Three-Dimensional Visual Simulation in Tidal Defense Engineering

Any tidal defense engineering involves the collection and analysis of massive information about engineering structures and their surrounding environment. Traditional method, which is carried out mainly by means of twodimensional drawings and textures, is not efficient and intuitive enough to analyze the whole project and reflect its spatial relationship. Three-dimensional visual simulation provides an advanced technical means of solving this problem. In this paper, triangular irregular network (TIN) model simplified by non-uniform rational B-splines (NURBS) technique was used to establish the digital terrain model (DTM) of a super large region. Simulation of dynamic water surface was realized by combining noise function with sine wave superposition method. Models of different objects were established with different modeling techniques according to their characteristics. Application of texture mapping technology remarkably improved the authenticity of the models. Taking the tidal defense engineering in the new coastal region of Tianjin as a case study, three-dimensional visual simulation and dynamic roaming of the study area were realized, providing visual analysis and visible demonstration method for the management and emergency decision-making associated with construction.