Search

Accurate wind speed forecasting plays a crucial role in the efficient and economical management of power supply systems. In this study, a novel framework combining variational mode decomposition (VMD), graph neural network (GNN) and temporal forecasting component is proposed for wind speed forecasting using multiple atmospheric variables. VMD is employed to decompose atmospheric variables into distinct subsequences at various frequencies, and GNN is utilized to effectively pass, aggregate, and update variable features, thus enabling the extraction of pairwise dependencies among the different variables. Subsequently, the transformer model is used as the temporal forecasting component in the proposed framework. Compared with several state-of-the-art transformer-based models and baseline models in AI field, the superior performance of our hybrid framework is observed. Lastly, several other deep learning models, including multi-layer perceptrons (MLP), long short-term memory (LSTM), and gated recurrent unit (GRU), are selected as forecasting component for assessing the applicability of the transformer. Interestingly, the transformer exhibits the lowest performance, while the GRU demonstrates the most promising results, with the mean absolute error (MAE) of 0.1356 m/s and 0.1085 m/s for one-step ahead forecasting, respectively. Overall, this research provides valuable insights into a novel framework and the applicability of the transformer for wind speed forecasting. © 2023 Elsevier Ltd

Tornado-induced railway accidents occurred globally and the risk analysis of high-speed train in tornado-like wind fields is required, particularly in the context of the ongoing development of high-speed rail systems. To study the applicability of Burgers-Rott and Sullivan tornado models, CFD simulations are firstly carried out. It reveals that the Burgers-Rott model can effectively simulate the tangential velocity characteristics of tornado-like vortices up to the two-celled turbulent vortex stage while the Sullivan model is a suitable choice for vortices in the multi-vortex stage. Subsequent analyses delve into the behavior of the dynamic amplification factor (DAF) in tornado-like vortices. The Burgers-Rott model exhibits a higher DAF value when the passing time exceeds 1.5s whereas the dominance of the Sullivan model is evident when the passing time is less than 1.5s. Finally, a linear relationship between characteristic wind curve (CWC) and train speed is noted. This relationship manifests as the characteristic wind speed reaches approximately 30 m/s as the train speed is 150 km/h. It is also noteworthy that the CWC exhibits an upward trajectory as the translational speed of the tornado intensifies. In conclusion, this study serves as a valuable reference for assessing the operational safety of high-speed trains in tornado-like vortices.

The development of wind power is crucial in addressing the global energy and environmental crises. To enhance wind farm efficiency, a novel bilevel non-uniform optimization approach is proposed to conduct the wind farm layout optimization (WFLO) and electrical cable routing (ECR) simultaneously. Additionally, 2D and 3D wake modelling strategies are investigated and validated to accurately evaluate the wind power generation from multiple wind turbines. The refined levelized cost of energy (LCOE), considered as the objective function, is proposed, and validated through two realistic offshore wind farms. Finally, a comparison of the uniform turbine layout optimized using both independent and bilevel optimization strategies, as well as the original layout, reveals that the LCOE decreases by 0.9% and 1.0%, respectively. The results shows that the bilevel optimization approach exhibits a slight advantage. Compared to the uniform optimization, the non-uniform optimization of turbine layout and cable routing shows better performance, with the LCOE reduction of 4.75%, 7.68%, and 9.11% for 2 MW, 3.6 MW and 8 MW, respectively. It concludes that the effectiveness of implementing a non-uniform and uniform turbine layout optimization for turbines with smaller and larger power capacities. The proposed method has the potential to improve offshore wind farm efficiency.

A new implementation of many-body calculations is of paramount importance in the field of computational physics. In this study, we leverage the capabilities of Field Programmable Gate Arrays (FPGAs) for conducting quantum many-body calculations. Through the design of appropriate schemes for Monte Carlo and tensor network methods, we effectively utilize the parallel processing capabilities provided by FPGAs. This has resulted in a remarkable tenfold speedup compared to CPU-based computation for a Monte Carlo algorithm. We also demonstrate, for the first time, the utilization of FPGA to accelerate a typical tensor network algorithm. Our findings unambiguously highlight the significant advantages of hardware implementation and pave the way for novel approaches to many-body calculations., Comment: 9 pages, 8 figures

Limei Zhang,1,2,* Shuzhao Chen,1,2,* Weida Wang,1,2 Yun Wang,1,2 Yang Liang1,2 1Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, People’s Republic of China; 2Department of Hematologic Oncology, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Guangzhou, People’s Republic of China*These authors contributed equally to this workCorrespondence: Yang Liang; Yun Wang, Department of Hematologic Oncology, Sun Yat-sen University Cancer, State Key Laboratory of Oncology in South China, 651 Dongfeng Road East, Guangzhou, Guangdong, 510060, People’s Republic of China, Email liangyang@sysucc.org.cn; wangyun@sysucc.org.cnPurpose: We aimed to assess the prognostic value of pretreatment inflammatory and nutritional parameters for predicting overall survival (OS) in patients with newly diagnosed multiple myeloma (NDMM), and to build a new scoring system using the most important variables.Methods: We retrospectively analyzed baseline clinical and laboratory data for patients with NDMM, who were randomly grouped into training and validation cohorts at a ratio of 8:2. The Inflammatory Nutritional Score (INS) was developed based on the least absolute shrinkage and selection operator (LASSO) Cox regression. The INS and other independent prognostic factors were entered into a multivariate Cox model and merged to generate a nomogram model for predictive optimization. Performance and predictive accuracy were assessed using the concordance index (C-index), calibration plots, and time-dependent receiver operating characteristic (ROC) curves.Results: In total, 442 eligible patients were enrolled. Six inflammatory/nutritional variables, including the Nutritional Risk Index (NRI), body mass index (BMI), neutrophil-lymphocyte ratio (NLR), monocyte-lymphocyte ratio (MLR), platelet-lymphocyte ratio (PLR), and albumin-alkaline phosp

The reliable wind speed forecasting is critical for wind farms as it enables them to improve cost-effectiveness and optimize energy efficiency. In this study, a novel hybrid deep learning model is proposed for multi-step wind speed forecasting, which simultaneously captures pairwise dependencies and temporal features of multiple atmosphere variables. Initially, to extract intrinsic mode functions (IMFs) of nonstationary atmospheric variables, the Variational Mode Decomposition (VMD) is used. The decomposed atmospheric variables are then grouped according to similarity of mode and the pairwise dependencies are captured using the Graph Neural Network (GNN). Subsequently, the Temporal Convolutional Network (TCN) is applied for wind speed forecasting. The proposed model notably outperforms existing models with a Mean Absolute Error (MAE) of less than 0.1 m/s in one-hour ahead forecasting (1-h). Additionally, it improves wind speed forecasting accuracy by 79.22% at 1-h, 79.13% at 2-h, 80.00% at 3-h, 78.70% at 6-h, and 58.41% at 12-h, respectively, compared to the frequently used LSTM technique for wind speed forecasting. The outcomes of this study indicate that this novel approach to wind speed forecasting holds substantial potential for real-world wind farm operational contexts. © 2023 Elsevier Ltd

In recent years, semi-supervised learning has been widely explored and shows excellent data efficiency for 2D data. There is an emerging need to improve data efficiency for 3D tasks due to the scarcity of labeled 3D data. This paper explores how the coherence of different modelities of 3D data (e.g. point cloud, image, and mesh) can be used to improve data efficiency for both 3D classification and retrieval tasks. We propose a novel multimodal semi-supervised learning framework by introducing instance-level consistency constraint and a novel multimodal contrastive prototype (M2CP) loss. The instance-level consistency enforces the network to generate consistent representations for multimodal data of the same object regardless of its modality. The M2CP maintains a multimodal prototype for each class and learns features with small intra-class variations by minimizing the feature distance of each object to its prototype while maximizing the distance to the others. Our proposed framework significantly outperforms all the state-of-the-art counterparts for both classification and retrieval tasks by a large margin on the modelNet10 and ModelNet40 datasets., Comment: BMVC 2021 poster

Cells in biofilms sense and interact with their environment through the extracellular matrix. The physicochemical properties of the matrix, particularly at the biofilm–environment interface, determine how cells respond to changing conditions. In this study we describe the application of atomic force microscopy and confocal imaging to probe in situ the mechanical properties of these interfacial regions and to elucidate how key matrix components can contribute to the physical sensing by the cells. We describe how the Young's modulus of microcolonies differs according to the size and morphology of microcolonies, as well as the flow rate. The Young's modulus increased as a function of microcolony diameter, which was correlated with the production of the polysaccharide Psl at later stages of maturation for hemispherical or mushroom shaped microcolonies. The Young's modulus of the periphery of the biofilm colony was however independent of the hydrodynamic shear. The morphology of the microcolonies also influenced interfacial or peripheral stiffness. Microcolonies with a diffuse morphology had a lower Young's modulus than isolated, circular ones and this phenomenon was due to a deficiency of Psl. In this way, changes in the specific polysaccharide components imbue the biofilm with distinct physical properties that may modulate the way in which bacteria perceive or respond to their environment. Further, the physical properties of the polysaccharides are closely linked to the specific architectures formed by the developing biofilm., Singapore-MIT Alliance in Research and Technology (SMART). BioSystems & Micromechanics IRG, National Science Foundation (U.S.), Singapore. National Research Foundation, Singapore. Ministry of Education. Research Centre of Excellence Programme