This article presents the first comprehensive literature review of deep learning (DL) applications in additive manufacturing (AM). It addresses the need for a thorough analysis in this rapidly growing yet scattered fi...
详细信息
This article presents the first comprehensive literature review of deep learning (DL) applications in additive manufacturing (AM). It addresses the need for a thorough analysis in this rapidly growing yet scattered field, aiming to bring together existing knowledge and encourage further development. Our research questions cover three major areas of AM: (i) design for AM, (ii) AM modeling, and (iii) monitoring and control in AM. We use a step-by-step approach following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to select papers from Scopus and Web of Science databases, aligning with our research questions. We include only those papers that implement DL across seven major AM categories - binder jetting, directed energy deposition, material extrusion, material jetting, powder bed fusion, sheet lamination, and vat photopolymerization. Our analysis reveals a trend towards using deep generative models, such as generative adversarial networks, for generative design in AM. It also highlights an increasing effort to incorporate process physics into DL models to improve AM processmodeling and reduce data requirements. Additionally, there is growing interest in using 3D point cloud data for AM process monitoring, alongside traditional 1D and 2D formats. Finally, this article summarizes the current challenges and recommends some of the promising opportunities in this domain for further investigation with a special focus on (i) generalizing DL models for a wide range of geometry types, (ii) managing uncertainties both in AM data and DL models, (iii) overcoming limited, imbalanced, and noisy AM data issues by incorporating deep generative models, and (iv) unveiling the potential of interpretable DL for AM.
This paper is devoted to the problem of magnetohydrodynamic stability (MHDS) in the energy-intensive process of primary aluminum production by electrolysis. Improving MHDS control is important because of the high cost...
详细信息
This paper is devoted to the problem of magnetohydrodynamic stability (MHDS) in the energy-intensive process of primary aluminum production by electrolysis. Improving MHDS control is important because of the high costs and reduced efficiency caused by the instability of magnetic and current fields. In this work, a methodological analysis of modern theoretical and numerical methods for studying MHDS was carried out, and approaches to optimizing magnetic fields and control algorithms aimed at stabilizing the process and reducing energy costs were considered. This review identified key challenges and proposed promising directions, including the application of computational methods and artificial intelligence to monitor and control electrolysis in real time. In this paper, it was revealed that wave MHD instability at the metal-electrolyte phase boundary is a key physical obstacle to further reducing specific energy costs and increasing energy stability. The novelty of this paper lies in an integrated approach that combines modeling and practical recommendations. The purpose of this study is to systematically summarize scientific data, analyze the key physical factors affecting the energy stability of electrolyzers, and determine promising directions for their solution. The results of this study can be used to improve the energy efficiency and environmental friendliness of aluminum production.
Continuous flow synthesis offers major advantages in the production of specialty products, e.g. zeolites, such as tight temperature control and lower variation in product quality. Here we consider the hydrothermal syn...
详细信息
Continuous flow synthesis offers major advantages in the production of specialty products, e.g. zeolites, such as tight temperature control and lower variation in product quality. Here we consider the hydrothermal synthesis of NaX zeolites in a continuous oscillatory baffled reactor (COBR). A process model is derived from physicochemical relationships to analyze and optimize the operation of a pilot scale continuous oscillatory baffled reactor for zeolite synthesis. The process model is validated using plant data. Furthermore, the uncertainty of the model predictions is quantified. Based on this analysis, robust optimization is used to compute robust optimal operation points of the COBR. These optimal operation points are validated by the application to a real pilot plant.
This study addresses the underexplored challenge of inherent dynamics in industrial processes through an innovative attention-based latent variable modeling method. Utilizing attention mechanisms, the method articulat...
详细信息
ISBN:
(纸本)9798331540845;9789887581598
This study addresses the underexplored challenge of inherent dynamics in industrial processes through an innovative attention-based latent variable modeling method. Utilizing attention mechanisms, the method articulates time-variant dynamical relationships among samples. The framework extends attention-based dynamical inner principal component analysis to extract latent dynamical features, integrating them with static features obtained through static principal component analysis. This results in comprehensive monitoring statistics for online applications. Numerical simulations and real-world application in an industrial ethylene oxychlorination process demonstrate the proposed method's efficacy. Comparative analysis highlights its advantages and superior performance over existing methods. This innovative approach provides more accurate insights into complex industrial processes, promising advancements in data-driven modeling within the field.
Intelligent microfluidics in nanoparticle synthesis embodies a comprehensive synergistic approach that merges numerical modeling, artificial intelligence, and experimental analysis, striving for controllability over a...
详细信息
Intelligent microfluidics in nanoparticle synthesis embodies a comprehensive synergistic approach that merges numerical modeling, artificial intelligence, and experimental analysis, striving for controllability over an energy-efficient microfluidic device designed for nanoparticle synthesis with desired physical properties. This study delves into a microfluidic mass transfer system, employing an innovative methodology that combines data-driven modeling, machine learning-based comparative multiobjective optimization, and experimental analysis to model a micromixing system. A surrogate data-driven model is employed to the microfluidic mass transfer system, considering four critical geometrical parameters and inlet Reynolds as design variables. The model provides insights into mixer's functionality. It is observed that at lower Reynolds numbers, increasing NoT increases the mixing efficiency by more than 20%. Moreover, altering SNDi value leads to a significant 80% reduction in pressure drop. Identifying the optimal system from numerous design parameters is challenging but accomplished through machine learning. Two distinct machine learning algorithms were integrated with mathematical surrogate modeling to optimize the mixer for three objectives. RSM-Differential Evolution significantly outperforms RSM-NSGA-ii in enhancing mixing characteristics and reducing the mechanical energy consumption by over 85%. Additionally, improvement in energy dissipation and effective energy efficiency of microsystem was made, alongside a comparable enhancement of mixing index and management of pressure drop. Fabrication of two optimal designs confirms an over 80% drop in pressure and an increase in mixing efficiency by over 20% at low Reynolds, outperforming conventional microfluidic mixers. The intelligent micromixer allows precise control over nanoparticle synthesis by adjusting microtransfer design parameters. This controlled process is crucial for tissue engineering hydrogel synthesi
Handling processdata characterized by strong nonlinearity, high dimensionality, cross-correlations, and auto-correlations presents a considerable hurdle for data-driven soft sensor modeling. While traditional slow fe...
详细信息
ISBN:
(纸本)9798350387780;9798350387797
Handling processdata characterized by strong nonlinearity, high dimensionality, cross-correlations, and auto-correlations presents a considerable hurdle for data-driven soft sensor modeling. While traditional slow feature analysis (SFA) adeptly captures slow and static features from linear data, it may falter in capturing the nonlinear, high-dimensional, and dynamic features inherent in time series data. Conversely, although Long Short-Term Memory (LSTM) networks are designed to address long-term dependencies within sequences, they encounter challenges, particularly with very lengthy data sequences, in effectively capturing these dependencies. Consequently, relying solely on SFA or LSTM may prove inadequate for addressing the complexities of time series data in industrial processes. To confront these challenges, this study proposes an innovative approach termed Slow LSTM (SLSTM), which amalgamates the hidden layers of LSTM and SFA to enhance feature extraction. These extracted features are subsequently fed into a fully connected layer for prediction. The efficacy of this proposed method is validated through comparisons with various methods, including SFA-FC, LSTM, and RNN.
According to the command-and-controlprocess, the conceptual model of tactical equipment support command and controlprocess is established, such as use case diagrams, class diagrams, sequence diagrams, and activity d...
详细信息
Aiming at overcoming the difficulties of actual test and demonstration in commercial space launch missions and the lack of visualized analysis tools, this paper studies the digital modeling technology for commercial s...
详细信息
Extracted event data from information systems often contain a variety of process executions making the data complex and difficult to comprehend. Unlike current research which only identifies the variability over time,...
详细信息
ISBN:
(纸本)9783031610066;9783031610073
Extracted event data from information systems often contain a variety of process executions making the data complex and difficult to comprehend. Unlike current research which only identifies the variability over time, we focus on other dimensions that may play a role in the performance of the process. This research addresses the challenge of effectively segmenting cases within operational processes based on continuous features, such as duration of cases, and evaluated risk score of cases, which are often overlooked in traditional processanalysis. We present a novel approach employing a sliding window technique combined with the earth mover's distance to detect changes in control flow behavior over continuous dimensions. This approach enables case segmentation, hierarchical merging of similar segments, and pairwise comparison of them, providing a comprehensive perspective on process behavior. We validate our methodology through a real-life case study in collaboration with UWV, the Dutch employee insurance agency, demonstrating its practical applicability. This research contributes to the field by aiding organizations in improving process efficiency, pinpointing abnormal behaviors, and providing valuable inputs for process comparison, and outcome prediction.
This research investigates the hard part turning of DC53 tool steel, which is engineered for better mechanical properties compared to AISI D2 tool steel, using Xcel cubic boron nitride. The machining input parameters ...
详细信息
This research investigates the hard part turning of DC53 tool steel, which is engineered for better mechanical properties compared to AISI D2 tool steel, using Xcel cubic boron nitride. The machining input parameters such as workpiece hardness (different heat treatments), cutting speed, feed rate, and depth of cut are used to thoroughly evaluate process science across conflicting machinability attributes such as cutting tool life, machined workpiece surface roughness, volume of material removed, machine tool power consumption, and tool-workpiece zone temperature. A full factorial design of experiments with two levels, resulting in 16 experiments, is performed with statistical parametric significance analysis to better controlprocess variability. Multiple artificial neural network (ANN) architectures are generated to accurately model the non-linearity of the process for better prediction of key characteristics. The optimized architectures are used as prediction models to a non-sorting genetic algorithm (NSGA-ii) to determine the optimal compromise among all conflicting responses. The significance analysis highlighted that heat treatment is the most influential variable on machinability, with a significance of 74.63% on tool life, 59.03% on roughness, 66.45% on material removed, 38.03% on power consumption, and 29.60% on interaction-zone temperature. The confidence of all ANN architectures is achieved above 0.97 R2 to accurately incorporate parametric relations with physical mechanisms. The compromise against conflicting machinability attributes identified by NSGA-ii optimization results in a 92.05% increase in tool life, a 91.83% increase in volume removed, a 33.33% decrease in roughness, a 26.73% decline in power consumption, and a 9.61% reduction in machining temperature. The process variability is thoroughly analyzed using statistical and physical analyses and computational intelligence, which will guide machinists in better decision-making. (c) 2025 Author(s). A
暂无评论