Computational Framework Research Articles

Multi-layer process control in selective laser melting: a reinforcement learning approach

AbstractPowder bed fusion (PBF) is an original additive manufacturing technique for creating 3D parts layer-by-layer. While there are numerous benefits to this process, the complex undergoing physical phenomena are challenging to analytically model and interpret. Hence, integrated and control-oriented 3D models are lacking in the current literature. As a result, the state of the art in process control for the powder bed fusion (PBF) process is not as advanced as in other manufacturing processes. Reinforcement learning is a machine learning, data-driven mathematical and computational framework that can be used for process control while addressing this challenge (lack of control-oriented models) effectively. Its flexible formulation and its trial-and-error nature make reinforcement learning suitable for processes where the model is intricate or even unknown. The focus of this research work is selective laser melting, which is a laser-based PBF process. For the first time in the literature we demonstrate the benefits of a reinforcement learning process control framework for multiple layers (complete 3D parts) and we highlight the importance of stability during training. The presented case studies confirm the effectiveness of the proposed control framework, directly addressing heat accumulation issues while demonstrating effective overall process control, hence opening up opportunities for further research and impact in this area.

Integrating dynamical modeling and phylogeographic inference to characterize global influenza circulation

Abstract Global seasonal influenza circulation involves a complex interplay between local (seasonality, demography, host immunity) and global factors (international mobility) shaping recurrent epidemic patterns. No studies so far have reconciled the two spatial levels, evaluating the coupling between national epidemics, considering heterogeneous coverage of epidemiological and virological data, integrating different data sources. We propose a novel combined approach based on a dynamical model of global influenza spread (GLEAM), integrating high-resolution demographic and mobility data, and a generalized linear model of phylogeographic diffusion that accounts for time-varying migration rates. Seasonal migration fluxes across countries simulated with GLEAM are tested as phylogeographic predictors to provide model validation and calibration based on genetic data. Seasonal fluxes obtained with a specific transmissibility peak time and recurrent travel outperformed the raw air-transportation predictor, previously considered as optimal indicator of global influenza migration. Influenza A subtypes supported autumn-winter reproductive number as high as 2.25 and an average immunity duration of 2 years. Similar dynamics were preferred by influenza B lineages, with a lower autumn-winter reproductive number. Comparing simulated epidemic profiles against FluNet data offered comparatively limited resolution power. The multiscale approach enables model selection yielding a novel computational framework for describing global influenza dynamics at different scales - local transmission and national epidemics vs. international coupling through mobility and imported cases. Our findings have important implications to improve preparedness against seasonal influenza epidemics. The approach can be generalized to other epidemic contexts, such as emerging disease outbreaks to improve the flexibility and predictive power of modeling.

Crosslinking degree variations enable programming and controlling soft fracture via sideways cracking

Large deformations of soft materials are customarily associated with strong constitutive and geometrical nonlinearities that originate new modes of fracture. Some isotropic materials can develop strong fracture anisotropy, which manifests as modifications of the crack path. Sideways cracking occurs when the crack deviates to propagate in the loading direction, rather than perpendicular to it. This fracture mode results from higher resistance to propagation perpendicular to the principal stretch direction. It has been argued that such fracture anisotropy is related to deformation-induced anisotropy resulting from the microstructural stretching of polymer chains and, in strain-crystallizing elastomers, strain-induced crystallization mechanisms. However, the precise variation of the fracture behavior with the degree of crosslinking remains to be understood. Leveraging experiments and computational simulations, here we show that the tendency of a crack to propagate sideways in the two component Elastosil P7670 increases with the degree of crosslinking. We explore the mixing ratio for the synthesis of the elastomer that establishes the transition from forward to sideways fracturing. To assist the investigations, we construct a novel phase-field model for fracture where the critical energy release rate is directly related to the crosslinking degree. Our results demonstrate that fracture anisotropy can be modulated during the synthesis of the polymer. Then, we propose a roadmap with composite soft structures with low and highly crosslinked phases that allow for control over fracture, arresting and/or directing the fracture. The smart combination of the phases enables soft structures with enhanced fracture tolerance and reduced stiffness. By extending our computational framework as a virtual testbed, we capture the fracture performance of the composite samples and enable predictions based on more intricate composite unit cells. Overall, our work offers promising avenues for enhancing the fracture toughness of soft polymers.

AI-Powered Multimodal Modeling of Personalized Hemodynamics in Aortic Stenosis.

Aortic stenosis (AS) is the most common valvular heart disease in developed countries. High-fidelity preclinical models can improve AS management by enabling therapeutic innovation, early diagnosis, and tailored treatment planning. However, their use is currently limited by complex workflows necessitating lengthy expert-driven manual operations. Here, we propose an AI-powered computational framework for accelerated and democratized patient-specific modeling of AS hemodynamics from computed tomography (CT). First, we demonstrate that the automated meshing algorithms can generate task-ready geometries for both computational and benchtop simulations with higher accuracy and 100 times faster than existing approaches. Then, we showthat the approach can be integrated with fluid-structure interaction and soft robotics models to accurately recapitulate a broad spectrum of clinical hemodynamic measurements of diverse AS patients. The efficiency and reliability of these algorithms make them an ideal complementary tool for personalized high-fidelity modeling of AS biomechanics, hemodynamics, and treatmentplanning.

Control of Cellular Differentiation Trajectories for Cancer Reversion.

Cellular differentiation is controlled by intricate layers of gene regulation, involving the modulation of gene expression by various transcriptional regulators. Due to the complexity of gene regulation, identifying master regulators across the differentiation trajectory has been a longstanding challenge. To tackle this problem, a computational framework, single-cell Boolean network inference and control (BENEIN), is presented. Applying BENEIN to human large intestinal single-cell transcriptome data, MYB, HDAC2, and FOXA2 are identified as the master regulators whose inhibition induces enterocyte differentiation. It is found that simultaneous knockdown of these master regulators can revert colorectal cancer cells into normal-like enterocytes by synergistically inducing differentiation and suppressing malignancy, which is validated by in vitro and in vivo experiments.

Systematic Survey on Energy Conservation Using Blockchain for Sustainable Computing Challenges and Roadmaps

ABSTRACTThis article proposes a comprehensive analysis of architectures that use blockchain technology to solve important aspects of computing sustainability, with an emphasis on scalability, resource utilization, transparency, and energy conservation. The research focuses on analyzing various structure which embraces decentralization and consensus principles, to redefine the computer infrastructure environment. The study highlights the technology that aids in improving scalability and resource utilization by decentralizing data storage and processing, relieving the load on centralized servers, and lowering the environmental effect of large‐scale data centers. This study's findings are significant in uncovering best practices and optimizing the environmental impact of blockchain technology by evaluating blockchain indicators such as scalability, resource utilization, transparency, and energy saving within the framework of sustainable computing. The project intends to assist in the development of resource‐efficient solutions, address scalability issues, and promote openness and accountability by evaluating the performance of various blockchain implementations. The findings help to promote the larger goal of connecting technical improvements, notably in blockchain, with global environmental goals. The research reveals that energy conservation is an important aspect of sustainability, and the different frameworks including techniques for optimizing energy consumption based on carbon footprint concerns, decentralization, and consensus methods are intended to prioritize energy‐efficient nodes and promote environmentally friendly practices, resulting in a greener computing ecosystem.

Computational analysis of free-corner-induced failure of thick angle-ply composite laminates established upon a global-local model

The complex three-dimensional (3D) state of interlaminar stresses in the vicinity of a free corner in composite laminates brings about delamination, matrix cracking, and overall failure. This study focuses on devising a straightforward and efficient computational framework for analyzing the free-corner-induced failures in mechanically and thermally-loaded thick angle-ply composite laminates based on an innovative global-local finite element model. In so doing, the global region is dealt with by the first-order shear deformation theory, and the local sub-model region near the free corner is scrutinized using the 3D layer-wise theory. Distinctive risk parameters associated with the meso-level damage modes, including ply failure, delamination, and transverse matrix cracking are well-defined under the corresponding criteria. The precision of the model is ensured through verifications in which satisfactory alignment between the current outcomes and those of other available works in the literature substantiates the efficacy of the proposed approach. Furthermore, the impacts of the configuration parameters such as fiber orientation and thickness of plies upon the risk parameters are fully examined. It is found that, generally, by increasing the angle of fibers between two adjacent layers and diminishing the ply thickness, the composite laminate becomes more critical. The matrix cracking mode is dominated regarding the delamination and is in close agreement with the ply failure analysis. The proposed computational infrastructure, occasioning time and financial savings is valuable and enlightening for designing and manufacturing composite structures.

NiCo identifies extrinsic drivers of cell state modulation by niche covariation analysis

Cell states are modulated by intrinsic driving forces such as gene expression noise and extrinsic signals from the tissue microenvironment. The distinction between intrinsic and extrinsic cell state determinants is essential for understanding the regulation of cell fate in tissues during development, homeostasis and disease. The rapidly growing availability of single-cell resolution spatial transcriptomics makes it possible to meet this challenge. However, available computational methods to infer topological tissue domains, spatially variable genes, or ligand-receptor interactions are limited in their capacity to capture cell state changes driven by crosstalk between individual cell types within the same niche. We present NiCo, a computational framework for integrating single-cell resolution spatial transcriptomics with matched single-cell RNA-sequencing reference data to infer the influence of the spatial niche on the cell state. By applying NiCo to mouse embryogenesis, adult small intestine and liver data, we demonstrate the ability to predict novel niche interactions that govern cell state variation underlying tissue development and homeostasis. In particular, NiCo predicts a feedback mechanism between Kupffer cells and neighboring stellate cells dampening stellate cell activation in the normal liver. NiCo provides a powerful tool to elucidate tissue architecture and to identify drivers of cellular states in local niches.

A computational framework for processing time-series of earth observation data based on discrete convolution: global-scale historical Landsat cloud-free aggregates at 30 m spatial resolution.

Processing large collections of earth observation (EO) time-series, often petabyte-sized, such as NASA's Landsat and ESA's Sentinel missions, can be computationally prohibitive and costly. Despite their name, even the Analysis Ready Data (ARD) versions of such collections can rarely be used as direct input for modeling because of cloud presence and/or prohibitive storage size. Existing solutions for readily using these data are not openly available, are poor in performance, or lack flexibility. Addressing this issue, we developed TSIRF (Time-Series Iteration-free Reconstruction Framework), a computational framework that can be used to apply diverse time-series processing tasks, such as temporal aggregation and time-series reconstruction by simply adjusting the convolution kernel. As the first large-scale application, TSIRF was employed to process the entire Global Land Analysis and Discovery (GLAD) ARD Landsat archive, producing a cloud-free bi-monthly aggregated product. This process, covering seven Landsat bands globally from 1997 to 2022, with more than two trillion pixels and for each one a time-series of 156 samples in the aggregated product, required approximately 28 hours of computation using 1248 Intel® Xeon® Gold 6248R CPUs. The quality of the result was assessed using a benchmark dataset derived from the aggregated product and comparing different imputation strategies. The resulting reconstructed images can be used as input for machine learning models or to map biophysical indices. To further limit the storage size the produced data was saved as 8-bit Cloud-Optimized GeoTIFFs (COG). With the hosting of about 20 TB per band/index for an entire 30 m resolution bi-monthly historical time-series distributed as open data, the product enables seamless, fast, and affordable access to the Landsat archive for environmental monitoring and analysis applications.

Enhancing full-waveform inversion of field GPR data: A source-independent approach with dynamic reference selection via SE-Wave-U-Net

Ground-penetrating radar (GPR) is a widely used nondestructive detection tool. Full-waveform inversion (FWI) offers a reliable algorithm for accurately imaging GPR data. Despite its potential, FWI’s effectiveness is often compromised by unknown excitation sources, which can lead to computational failures, diminished imaging accuracy, or both. Addressing this challenge, this paper introduces a convolution-type objective function designed to mitigate the detrimental effects of these unknown excitation sources on FWI imaging. Critical to the success of this function is the precise selection of reference traces. Further refining the capability for signal separation, we have integrated squeeze-and-excitation (SE) modules into the traditional Wave-U-Net architecture, culminating in the development of the SE-Wave-U-Net framework. This framework is adept at dynamically extracting accurate reference traces from field GPR data, which are essential for accurate FWI imaging. The efficacy, accuracy, and practical applicability of our methods are validated through extensive analysis of numerical, laboratory experiments, and field GPR data. Our algorithm not only establishes a high-precision computational framework for FWI but also enhances the reliability of GPR imaging in complex environments.

Optimizing Urban Traffic Flow through Advanced Tensor Analysis and Multilinear Algebra: A Computational Approach to Enhancing Smart City Dynamics

Growing traffic congestion is a worldwide problem that collides against the aims of environmental sustainability, economic productivity, and the quality of life in cities. This research proposes a new computational framework for traffic management that integrates advanced tensor analysis and methods from multilinear algebra. We have developed and validated a new predictive model that greatly improves the optimization of traffic flows by synthesizing the naturally complex multi-dimensional traffic data analysis. Our results demonstrate that, compared with existing systems, the proposed approach results in higher accuracy of prediction, much improved computational efficiency, and provides scalable and adaptable solutions for application in a wide range of urban habitats. Such research may push the boundaries further on the smart city infrastructures to provide a very well-founded mathematical framework for the dynamics of improved urban mobility through high-level data-oriented information.

Revolutionizing healthcare: a comparative insight into deep learning’s role in medical imaging

Recently, Deep Learning (DL) models have shown promising accuracy in analysis of medical images. Alzeheimer Disease (AD), a prevalent form of dementia, uses Magnetic Resonance Imaging (MRI) scans, which is then analysed via DL models. To address the model computational constraints, Cloud Computing (CC) is integrated to operate with the DL models. Recent articles on DL-based MRI have not discussed datasets specific to different diseases, which makes it difficult to build the specific DL model. Thus, the article systematically explores a tutorial approach, where we first discuss a classification taxonomy of medical imaging datasets. Next, we present a case-study on AD MRI classification using the DL methods. We analyse three distinct models-Convolutional Neural Networks (CNN), Visual Geometry Group 16 (VGG-16), and an ensemble approach-for classification and predictive outcomes. In addition, we designed a novel framework that offers insight into how various layers interact with the dataset. Our architecture comprises an input layer, a cloud-based layer responsible for preprocessing and model execution, and a diagnostic layer that issues alerts after successful classification and prediction. According to our simulations, CNN outperformed other models with a test accuracy of 99.285%, followed by VGG-16 with 85.113%, while the ensemble model lagged with a disappointing test accuracy of 79.192%. Our cloud Computing framework serves as an efficient mechanism for medical image processing while safeguarding patient confidentiality and data privacy.

Detecting global and local hierarchical structures in cell-cell communication using CrossChat

Cell-cell communication (CCC) occurs across different biological scales, ranging from interactions between large groups of cells to interactions between individual cells, forming a hierarchical structure. Globally, CCC may exist between clusters or only subgroups of a cluster with varying size, while locally, a group of cells as sender or receiver may exhibit distinct signaling properties. Current existing methods infer CCC from single-cell RNA-seq or Spatial Transcriptomics only between predefined cell groups, neglecting the existing hierarchical structure within CCC that are determined by signaling molecules, in particular, ligands and receptors. Here, we develop CrossChat, a novel computational framework designed to infer and analyze the hierarchical cell-cell communication structures using two complementary approaches: a global hierarchical structure using a multi-resolution clustering method, and multiple local hierarchical structures using a tree detection method. This framework provides a comprehensive approach to understand the hierarchical relationships within CCC that govern complex tissue functions. By applying our method to two nonspatial scRNA-seq datasets sampled from COVID-19 patients and mouse embryonic skin, and two spatial transcriptomics datasets generated from Stereo-seq of mouse embryo and 10x Visium of mouse wounded skin, we showcase CrossChat’s functionalities for analyzing both global and local hierarchical structures within cell-cell communication.

Deposition of nanoparticles through a cylindrical tube of human respiratory system

This paper proposes a mathematical model for nanoparticle deposition in the respiratory bronchiole of the human lung airways. The length of the respiratory bronchiole of the human lung is 1.2 mm, and the diameter is 0.5 mm. Therefore, to investigate the impact of inhaled nanoparticles within the lung bronchioles, first, we used the Navier–Stokes equation by including the Darcy term together with Newton’s equation of motion. Then we non-dimensionalized the governing equation, and numerical solutions are obtained using the finite difference technique. Under the assumption of laminar pulsatile flow conditions and axial symmetry, the problem is simplified to a two-dimensional computational domain. The computational framework is implemented using MATLAB R2018 through user-defined code. Velocity propagation, the Reynolds number, and the Darcy number effect are performed to examine the flow conditions inside the respiratory bronchioles.

Advancing Real-World Image Dehazing: Perspective, Modules, and Training.

Restoring high-quality images from degraded hazy observations is a fundamental and essential task in the field of computer vision. While deep models have achieved significant success with synthetic data, their effectiveness in real-world scenarios remains uncertain. To improve adaptability in real-world environments, we construct an entirely new computational framework by making efforts from three key aspects: imaging perspective, structural modules, and training strategies. To simulate the often-overlooked multiple degradation attributes found in real-world hazy images, we develop a new hazy imaging model that encapsulates multiple degraded factors, assisting in bridging the domain gap between synthetic and real-world image spaces. In contrast to existing approaches that primarily address the inverse imaging process, we design a new dehazing network following the "localization-and-removal" pipeline. The degradation localization module aims to assist in network capture discriminative haze-related feature information, and the degradation removal module focuses on eliminating dependencies between features by learning a weighting matrix of training samples, thereby avoiding spurious correlations of extracted features in existing deep methods. We also define a new Gaussian perceptual contrastive loss to further constrain the network to update in the direction of the natural dehazing. Regarding multiple full/no-reference image quality indicators and subjective visual effects on challenging RTTS, URHI, and Fattal real hazy datasets, the proposed method has superior performance and is better than the current state-of-the-art methods.

Object-Oriented Zero-Information-Optimised Product of Monte Carlo Integrations

Abstract We propose an object-oriented Monte Carlo framework for computation of definite integrals with a primary focus on generality and simplicity, improving in find points existing software libraries. We introduce an optimised sampling method for a product of Monte Carlo integrations. The method does not require any a-priori knowledge of the integrands or the integration domains, and can be combined with other optimisation methods. Using our framework, we implement a naive uniform sampling method and the optimised one, and empirically compare them on the Monte Carlo computation of volumes of N-dimensional balls. The experiments confirm our theoretical expectations.

Adaptive Hybrid Optimized LSTM Models: A Novel Computational Framework for Algorithmic Financial Trading System

In our modern society, with the development of Internet and information system, pre-programing algorithmic trading strategies for online automatic trading has also flourished, especially in the rapidly fluctuating trading market. Using quantitative trading programs that can automatically trade in response to market conditions has attracted the attention of major financial institutions and governments in the financial market. How to use self-adaptive programs to automatically trade in the ever-changing financial market has become a popular research topic for the development and pursuit of all financial markets in recent years. This research proposes an online self-adaptive trading algorithm that can be applied to financial markets such as stock market, currency markets, cryptocurrency markets, futures markets, etc. In the first part of the algorithm, the simplified swarm optimization will be used to optimize the parameters of the newly proposed flexible grid in this research. Then the data will be imported into the artificial neural network model for training in the latter part, helping the trading model automatically select the appropriate parameters for construction flexible grid corresponding to the market conditions. The greatest contribution of the research is to provide a whole new trading algorithm that can adapt to the dynamic trading market, automatically make suitable adjustments to current trading strategy and place real-time orders. The algorithm controlling both profit and risk, which can seem like a balanced trading algorithm that are robust and profitable.

Computational Intelligence for Modeling the Rheological Properties of the Developed Hydrated Lime-Based Alkali-Activated Materials

The development of alkali-activated materials (AAMs) has gained attention as a sustainable alternative to Portland cement-based concrete. A key factor in AAM performance is its rheological behavior, which influences workability and pumpability. Accurate prediction of these properties is vital for optimizing mix design. This study introduces a computational framework using ensemble and non-ensemble machine learning (ML) techniques to model the yield stress (YS) and plastic viscosity (PV) of AAMs. A comprehensive dataset from previous experiments was used to train models, including artificial neural networks (ANN), gene expression programming (GEP), decision tree, random forest, adaptive boosting, and gradient boosting. The models achieved R² values exceeding 0.95 and correlation coefficients above 0.97. Error metrics, such as MAPE, MAE, MSE, and RMSE, further validated the models’ generalizability. A parametric analysis showed that a 20% increase in hydrated lime (HL) content raised YS by 350Pa, which is similar to the effect of varying fly ash (FA) from 60% to 100%. A 40% increase in FA and HL content increased PV by 4Pa.s and 25Pa.s, respectively. A 3% rise in sodium silicate led to a 400Pa increase in YS, while a similar increase in sodium hydroxide raised YS by 170Pa and PV by 15Pa·s. Overall, this study provides an effective tool for optimizing AAM mix designs by accurately predicting rheological properties and quantifying the impact of key parameters, contributing to the advancement of sustainable cementitious materials in construction.

A multi-core makespan model for parallel scientific workflow execution in cloud computational framework

Researchers have shown a lot of potential in optimizing cloud-based workload-scheduling over the past few years. However, executing scientific workloads inside the cloud is time-consuming and costly, making it inefficient from both a financial and productivity standpoint. As a result, there are many investigations conducted, with the general trend being to speed up the rate of processing and establish a cost-effective system, whereby customers are billed according to their actual use. In addition, energy-consumption is capable of being reduced, especially if the available resources are heterogeneous; however, few investigations have optimized multi-core with analyzing makespan parameters collectively to fulfill the quality of service (QoS) and service level agreement (SLA) of the workload task. In this research, we introduce an optimal scheduling for a heterogeneous distributed cloud computing environment called task aware makespan optimized scheduler (TAMOS) that guarantees requirements across the task levels of scientific workflows. The energy and time required to carry out specific workflows are significantly reduced by using this TAMOS strategy. The TAMOS framework was studied using the scientific workflows namely, inspiral and sipht. When compared to the conventional method of scheduling work, our methodology used less energy and makespan.

A Computational Framework to Optimize the Mechanical Behavior of Synthetic Vascular Grafts

A Computational Framework to Optimize the Mechanical Behavior of Synthetic Vascular Grafts